hash stringlengths 32 32 | doc_id stringlengths 7 13 | section stringlengths 3 121 | content stringlengths 0 2.2M |
|---|---|---|---|
c9db735a545115f099d996de740d405e | 101 035 | 5.12 Item (l) | The protection switching function (MSP) foreseen in ITU-T Recommendation G.783 [10] is not suitable, if e.g. an improvement of the transmission quality against multipath activity in a radio link is required. As a consequence, in these cases, a radio link should have its own twin path or multi-line hitless protection switching system. |
c9db735a545115f099d996de740d405e | 101 035 | 5.12.1 1 + 1 protection | 1 + 1 protection switching section can be either a regenerator section or a multiplex section. |
c9db735a545115f099d996de740d405e | 101 035 | 5.12.2 n + m protection | A functional block diagram for an n + 1 configuration is shown in figure 9. ETSI TR 101 035 V1.1.3 (1998-05) 22 Radio protection switching (See Item N) STM-1 interfaces SOH access Radio protection switching (See Item N) STM-1 interfaces SOH access Standby radio channel Normal radio channel 1 Normal radio channel n ch 1 ch n ch 1 ch n ch 1 ch n ch 1 ch n MS Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Radio terminal SOH access (A1/A2 FA) Figure 9: n + 1 Switching arrangement Multiplex Section Adaptation (MSA) functions and therefore pointer activities are located only at the terminals. As suggested in ETS 300 635 [16] different solutions are possible so that the exact position of RPS function should not be a subject of standardization. Because K1 and K2 bytes are used for network protection, a data communication channel for the switching control signals should be established in a media specific byte as mentioned in item j). From figure 9 it can be seen that SDH DRRS are frame aligned (A1/A2 FA), on each hop, by means of A1 and A2 frame words of the STM-1 SOH. During multipath phenomena the different operating channels may require the stand-by channel on the basis of BER alarms. When a switch operation starts the signal is transmitted in parallel on the operating channels and on the stand-by channel. Since the various STM-1 signals of the operating channels and of the stand-by channel may not be synchronized in frequency (if the network is still not completely synchronized) and in phase, during the switching operation, the alignment may be lost on each hop of the stand-by channel. This event increases the switching operation time, which, on the other hand, should be as fast as possible (typically less than 10 ms) in order to counteract multipath dynamics and improve the system quality. This implies the necessity of synchronization of all STM-1 signals both in frequency and in phase obtained by a section adaptation (MSA) function on the STM-1 signals in the radio terminal station (radio equipment and switching system). Since a MSA terminates and starts a multiplex section, the consequence is that a n + m multi-line hitless switching section should be a multiplex section. Anyway no termination of the MS may be possible if the number of regenerator sections between radio terminals with RPS is limited so that, the total time of detect/restore the A1/A2 frame loss/alignment of the repeater chain will not affect the total switching time over the required minimum to ensure efficient hitless protection. See ETS 300 635 [16]. |
c9db735a545115f099d996de740d405e | 101 035 | 5.13 Item (m) | See ETS 300 635 [16] and ETS 300 785 [12]. |
c9db735a545115f099d996de740d405e | 101 035 | 5.14 Item (n) | For further study. ETSI TR 101 035 V1.1.3 (1998-05) 23 |
c9db735a545115f099d996de740d405e | 101 035 | 5.15 Item (o) | See appendix 2 of ITU-R Recommendation F.750 [1]. |
c9db735a545115f099d996de740d405e | 101 035 | 5.16 Item (p) | To utilize the section overhead, the bytes mentioned in item j) could be accessed either inside or outside the protection switching. In figures 10 and 11 the access of the overhead is shown. RST MST RST MST RST MST RST MST Radio Switching Protection (RPS) TX side RX side STM-1 physical physical interface interface SOH access Radio Switching Protection Radio (RPS) STM-1 (*) (*) (*) (*) ROHA (SPI) (RSPI) (*) Monitored and processed bytes: A1, A2, D1÷D3, D4÷D12 E1, E2, NU, B1, B2 C1 (see note) F1 (see note) NOTE: Confirmation is required that these bytes are available for system use. Figure 10: Radio system MST (including RST) functions ETSI TR 101 035 V1.1.3 (1998-05) 24 Radio Switching Protection (RPS) TX side RX side STM-1 physical physical interface interface SOH access Radio Switching Protection Radio (RPS) STM-1 (*) (*) (*) (*) RST (MST) (SPI in Radio ROHA Terminals) or Radio physical interface (RSPI in Repeaters) (RSPI) Radio (see note 2) RST (MST) (see note 2) RST RST (MST) (MST) (see note 2) (see note 2) (*) Monitored and processed bytes: A1, A2, D1÷D3, D4÷D120 E1, E2, NU, B1, B2, C1 F1 (see note 1), NU rows 1÷3 and 5÷9, B1, C1 (see note 1) NOTE 1: Confirmation is required that these bytes are available for radio system use. NOTE 2: If rows 5÷9 are accessed, even parity of B2 byte should be maintained. Figure 11: Radio system RST functions ETSI TR 101 035 V1.1.3 (1998-05) 25 History Document history Edition 1 July 1995 Publication as TM-TR 004 V1.1.2 June 1997 Publication V1.1.3 May 1998 Publication ISBN 2-7437-2172-3 Dépôt légal : Mai 1998 |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 1 Scope | The present document identifies ETSI specifications for TETRA V+D and DMO Release 1.1. Release 1.1 specifications were functionally frozen after the 14th EP TETRA meeting in March 2001. NOTE 1: Functionally frozen means that no further functionality/features may be incorporated into the set of specifications, and that only corrective Change Requests (CRs) are to be accepted and agreed. NOTE 2: It can be expected that corrective CRs will be introduced into the Release 1.1 specifications throughout year 2001. NOTE 3: Some of the CRs that will be produced during 2001 will be for the next release as they add new functionalities/features. NOTE 4: Only published documents are included into this release 1.1. |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 2 References | For the purposes of this Technical Report (TR) the following reference applies: [1] Terms of Reference of ETSI Project TETRA: http://portal.etsi.org/tetra/TETRA_ToR.asp. |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 3 Abbreviations | For the purposes of the present document, the following abbreviations apply: CR Change Request DMO Direct Mode Operation PAMR Public Access Mobile Radio PMR Private Mobile Radio SIM Subscriber Identity Module TETRA TErrestial Trunked RAdio V+D Voice plus Data |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4 Specifications of TETRA Release 1.1 | The following clauses contain ETSI deliverables included in TETRA V+D and DMO Release 1.1. |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4.1 Voice plus Data (V+D) | ETSI ETS 300 392-1 (1996): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 1: General network design". NOTE: Only those clauses, which are referred to in other standards, will be maintained. ETSI EN 300 392-2 (V2.3.2): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 2: Air Interface (AI)". ETSI ETS 300 392-3-1 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 3: Interworking at the Inter-System Interface (ISI); Sub-part 1: General design". ETSI EN 300 392-3-2 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 3: Interworking at the Inter-System Interface (ISI); Sub-part 2: Additional Network Feature Individual Call (ANF-ISIIC)". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 7 ETSI TS 100 392-3-2 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 3: Interworking at the Inter-System Interface (ISI); Sub-part 2: Additional Network Feature Individual Call (ANF-ISIIC)". ETSI ETS 300 392-3-3 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 3: Interworking at the Inter-System Interface (ISI); Sub-part 3: Additional Network Feature Group Call (ANF-ISIGC)". ETSI EN 300 392-3-4 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 3: Interworking at the Inter-System Interface (ISI); Sub-part 4: Additional Network Feature Short Data Service (ANF-ISISDS)". ETSI ETS 300 392-3-5 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 3: Interworking at the Inter-System Interface (ISI); Sub-part 5: Additional Network Feature for Mobility Management (ANF-ISIMM)". ETSI ETS 300 392-4-1 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 4: Gateways basic operation; Sub-part 1: Public Switched Telephone Network (PSTN)". ETSI ETS 300 392-4-2 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 4: Gateways basic operation; Sub-part 2: Integrated Services Digital Network (ISDN) gateway". ETSI ETS 300 392-4-3 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 4: Gateways basic operation; Sub-part 3: Data networks gateway". ETSI TS 100 392-7 (V2.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 7: Security". ETSI EN 300 392-7 (V2.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 7: Security". ETSI ETS 300 392-10-1 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 1: Call identification". ETSI ETS 300 392-10-2 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 2: Call report". ETSI ETS 300 392-10-3 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 3: Talking Party Identification (TPI)". ETSI ETS 300 392-10-5 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 5: List Search Call (LSC)". ETSI ETS 300 392-10-6 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 6: Call Authorized by Dispatcher (CAD)". ETSI ETS 300 392-10-7 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 7: Short number addressing". ETSI ETS 300 392-10-8 (1996): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 8: Area selection". ETSI ETS 300 392-10-9 (1998): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 9: Access priority". ETSI ETS 300 392-10-10 (1996): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 10: Priority Call (PC)". ETSI ETS 300 392-10-11 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 11: Call Waiting (CW)". ETSI ETS 300 392-10-12 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 12: Call Hold (CH)". ETSI ETS 300 392-10-13 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 13: Call completion to busy subscriber". ETSI ETS 300 392-10-16 (1996): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 16: Pre-emptive priority call". ETSI ETS 300 392-10-20 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 20: Discreet Listening (DL)". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 8 ETSI ETS 300 392-10-22 (1996): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 22: Dynamic group number assignment". ETSI ETS 300 392-10-23 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 23: Call completion on no reply". ETSI ETS 300 392-10-24 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 10: Supplementary services stage 1; Sub-part 24: Call Retention (CRT)". ETSI ETS 300 392-11-1 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 1: Call Identification (CI)". ETSI ETS 300 392-11-2 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 2: Call Report (CR)". ETSI ETS 300 392-11-3 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 3: Talking Party Identification (TPI)". ETSI ETS 300 392-11-5 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 5: List Search Call (LSC)". ETSI ETS 300 392-11-6 (1998): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 6: Call Authorized by Dispatcher (CAD)". ETSI ETS 300 392-11-7 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 7: Short Number Addressing (SNA)". ETSI EN 300 392-11-8 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 8: Area Selection (AS)". ETSI TS 100 392-11-8 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 8: Area Selection (AS)". ETSI ETS 300 392-11-9 (1998): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 9: Access Priority (AP)". ETSI EN 300 392-11-10 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 10: Priority Call (PC)". ETSI ETS 300 392-11-11 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 11: Call Waiting (CW)". ETSI EN 300 392-11-12 (2001): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 12: Call Hold (CH)". ETSI ETS 300 392-11-13 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 13: Call Completion to Busy Subscriber (CCBS)". ETSI ETS 300 392-11-16 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 16: Pre-emptive Priority Call (PPC)". ETSI ETS 300 392-11-20 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 20: Discreet Listening (DL)". ETSI ETS 300 392-11-22 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 22: Dynamic Group Number Assignment (DGNA)". ETSI ETS 300 392-11-23 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 23: Call Completion on No Reply (CCNR)". ETSI ETS 300 392-11-24 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 11: Supplementary services stage 2; Sub-part 24: Call Retention (CRT)". ETSI ETS 300 392-12-1 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 1: Call Identification (CI)". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 9 ETSI ETS 300 392-12-2 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 2: Call Report (CR)". ETSI ETS 300 392-12-3 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 3: Talking Party Identification (TPI)". ETSI ETS 300 392-12-5 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 5: List Search Call (LSC)". ETSI ETS 300 392-12-6 (1998): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 6: Call Authorized by Dispatcher (CAD)". ETSI ETS 300 392-12-7 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 7: Short Number Addressing (SNA)". ETSI EN 300 392-12-8 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 8: Area Selection (AS)". ETSI TS 100 392-12-8 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 8: Area Selection (AS)". ETSI ETS 300 392-12-9 (1998): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 9: Access Priority (AP)". ETSI EN 300 392-12-10 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 10: Priority Call (PC)". ETSI ETS 300 392-12-11 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 11: Call Waiting (CW)". ETSI EN 300 392-12-12 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 12: Call Hold (CH)". ETSI ETS 300 392-12-13 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 13: Call Completion to Busy Subscriber (CCBS)". ETSI ETS 300 392-12-16 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 16: Pre-emptive Priority Call (PPC)". ETSI ETS 300 392-12-20 (1999): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 20: Discreet Listening (DL)". ETSI ETS 300 392-12-22 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 22: Dynamic Group Number Assignment (DGNA)". ETSI ETS 300 392-12-23 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 23: Call Completion on No Reply (CCNR)". ETSI ETS 300 392-12-24 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 12: Supplementary services stage 3; Sub-part 24: Call Retention (CRT)". ETSI TS 100 392-15 (V1.1.1): "Terrestial Trunked Radio (TETRA) Voice plus DATA (V+D); Part 15: TETRA frequency bands, duplex spacings and channel numbering". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 10 |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4.2 Conformance testing | ETSI EN 300 394-1 (V2.3.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 1: Radio". ETSI TS 100 394-1 (V2.3.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 1: Radio". ETSI ETS 300 394-2-1 (1998): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 2: Protocol testing specification for Voice plus Data (V+D); Sub-part 1: Test suite structure and test purposes". ETSI ETS 300 394-2-2 (1998): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 2: Protocol testing specification for Voice plus Data (V+D); Sub-part 2: Abstract Test Suite (ATS) for Network (NWK) layer". ETSI ETS 300 394-2-3 (1998): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 2: Protocol testing specification for Voice plus Data (V+D); Sub-part 3: Abstract Test Suite (ATS) for Logical Link Control (LLC)". ETSI ETS 300 394-2-4 (1998): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 2: Protocol testing specification for Voice plus Data (V+D); Sub-part 4: Abstract Test Suite (ATS) for Medium Access Control (MAC)". ETSI ETS 300 394-4-1 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 1: Test Suite Structure and Test Purposes (TSS&TP) for Mobile Station to Mobile Station (MS-MS) Air Interface (AI)". ETSI ETS 300 394-4-2 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 2: Abstract Test Suite (ATS) for Mobile Station to Mobile Station (MS-MS) Air Interface (AI)". ETSI EN 300 394-4-3 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 3: Test Suite Structure and Test Purposes (TSS&TP) for Mobile Station (MS) Repeater type 1". ETSI TS 100 394-4-3 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 3: Test Suite Structure and Test Purposes (TSS&TP) for Mobile Station (MS) Repeater type 1". ETSI EN 300 394-4-4 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 4: Test Suite Structure and Test Purposes (TSS&TP) for Direct Mode Repeater (DM-REP) type 1". ETSI TS 100 394-4-4 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 4: Test Suite Structure and Test Purposes (TSS&TP) for Direct Mode Repeater (DM-REP) type 1". ETSI EN 300 394-4-5 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 5: Abstract Test Suite (ATS) for Mobile Station (MS) Repeater type 1". ETSI TS 100 394-4-5 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 5: Abstract Test Suite (ATS) for Mobile Station (MS) Repeater type 1". ETSI EN 300 394-4-6 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 6: Abstract Test Suite (ATS) for Direct Mode Repeater (DM-REP) type 1". ETSI TS 100 394-4-6 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 6: Abstract Test Suite (ATS) for Direct Mode Repeater (DM-REP) type 1". ETSI ETS 300 394-4-7 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 7: Test Suite Structure and Test Purposes (TSS&TP) for Mobile Station to GateWay (MS-GW) Air Interface (AI)". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 11 ETSI ETS 300 394-4-8 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 8: Test Suite Structure and Test Purposes (TSS&TP) for Direct Mode Gateway (DM-GATE)". ETSI ETS 300 394-4-9 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 9: Abstract Test Suite (ATS) for Mobile Station (MS) Gateway". ETSI ETS 300 394-4-10 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 10: Abstract Test Suite (ATS) for Direct Mode Gateway (DM-GATE)". ETSI EN 300 394-4-11 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 11: Test Suite Structure and Test Purposes (TSS&TP) for Mobile Station Repeater type 2". ETSI TS 100 394-4-11 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 11: Test Suite Structure and Test Purposes (TSS&TP) for Mobile Station Repeater type 2". ETSI EN 300 394-4-12 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 12: Test Suite Structure and Test Purposes (TSS&TP) for Repeater type 2". ETSI TS 100 394-4-12 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 12: Test Suite Structure and Test Purposes (TSS&TP) for Repeater type 2". ETSI EN 300 394-4-13 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 13: Abstract Test Suite (ATS) for Mobile station Repeater type 2". ETSI TS 100 394-4-13 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 13: Abstract Test Suite (ATS) for Mobile station Repeater type 2". ETSI EN 300 394-4-14 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 14: Abstract Test Suite (ATS) for Repeater type 2". ETSI TS 100 394-4-14 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 4: Protocol testing specification for Direct Mode Operation (DMO); Sub-part 14: Abstract Test Suite (ATS) for Repeater type 2". ETSI ETS 300 394-5-1 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 5: Security; Sub-part 1: Protocol Implementation Conformance Statement (PICS) proforma specification". ETSI ETS 300 394-5-2 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 5: Security; Sub-part 2: Protocol testing specification for TETRA security". ETSI ETS 300 394-5-3 (1999): "Terrestrial Trunked Radio (TETRA); Conformance testing specification; Part 5: Security; Sub-part 3: Abstract Test Suite (ATS)". |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4.3 Speech codec | ETSI ETS 300 395-1 (1997): "Terrestrial Trunked Radio (TETRA); Speech codec for full-rate traffic channel; Part 1: General description of speech functions". ETSI ETS 300 395-2 (1998): "Terrestrial Trunked Radio (TETRA); Speech codec for full-rate traffic channel; Part 2: TETRA codec". ETSI ETS 300 395-3 (1997): "Terrestrial Trunked Radio (TETRA); Speech codec for full-rate traffic channel; Part 3: Specific operating features". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 12 ETSI ETS 300 395-4 (2000): "Terrestrial Trunked Radio (TETRA); Speech codec for full-rate traffic channel; Part 4: Codec conformance testing". |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4.4 Direct Mode Operation (DMO) | ETSI ETS 300 396-1 (1998): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 1: General network design". ETSI ETS 300 396-2 (1998): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 2: Radio aspects". ETSI ETS 300 396-3 (1998): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 3: Mobile Station to Mobile Station (MS-MS) Air Interface (AI) protocol". ETSI EN 300 396-4 (V1.2.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 4: Type 1 repeater air interface". ETSI TS 100 396-4 (V1.2.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 4: Type 1 repeater air interface". ETSI ETS 300 396-5 (2000): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 5: Gateway air interface". ETSI ETS 300 396-6 (1998): "Terrestrial Trunked Radio (TETRA); Direct Mode Operation (DMO); Part 6: Security". ETSI EN 300 396-7 (V1.2.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 7: Type 2 repeater air interface". ETSI TS 100 396-7 (V1.2.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 7: Type 2 repeater air interface". ETSI ETS 300 396-8-1 (2000): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 8: Protocol Implementation Conformance Statement (PICS) proforma specification; Sub-part 1: Mobile Station to Mobile Station (MS-MS) Air Interface (AI)". ETSI EN 300 396-8-2 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 8: Protocol Implementation Conformance Statement (PICS) proforma specification; Sub-part 2: Type 1 repeater Air Interface (AI)". ETSI TS 100 396-8-2 (V1.1.1, 2000): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 8: Protocol Implementation Conformance Statement (PICS) proforma specification; Sub-part 2: Type 1 repeater Air Interface (AI)". ETSI ETS 300 396-8-3 (1999): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 8: Protocol Implementation Conformance Statement (PICS) proforma specification; Sub-part 3: Gateway Air Interface (AI)". ETSI EN 300 396-8-4 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 8: Protocol Implementation Conformance Statement (PICS) proforma specification; Sub-part 4: Type 2 repeater Air Interface (AI)". ETSI TS 100 396-8-4 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 8: Protocol Implementation Conformance Statement (PICS) proforma specification; Sub-part 4: Type 2 repeater Air Interface (AI)". ETSI TS 100 396-10 (V1.1.1): "Terrestrial Trunked Radio (TETRA); Technical requirements for Direct Mode Operation (DMO); Part 10: Managed Direct Mode Operation (M-DMO)". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 13 |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4.5 Subscriber Identity Module (SIM) | ETSI TR 101 494 (V1.1.1): "Terrestrial Trunked Radio (TETRA); SIM; Review". |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4.6 Type approval | ETSI TBR 035 (Citation in the OJ 07/04/2000): "Terrestrial Trunked Radio (TETRA); Emergency access". ETSI EN 301 435-1 (V1.2.4): "Terrestial Trunked Radio (TETRA) Attachment requirements for TETRA terminal equipment; Part 1: Civil access". ETSI EN 301 435-2 (V1.2.4): "Terrestrial Trunked Radio (TETRA); Attachment requirements for TETRA terminal equipment; Part 2: Emergency access". |
2458c21fac9272175e9c4c20a3086f72 | 100 392-17-1 | 4.7 Other specifications | ETSI TS 101 789-1 (V1.1.1): "Terrestial Trunked Radio (TETRA); TMO Repeaters; Part 1: Requirements, test methods and limits". ETSI EN 301 040 (V2.0.0): "Terrestrial Trunked Radio (TETRA); Security; Lawful Interception (LI) interface". ETSI EN 301 489-18 (V1.1.1): "ElectroMagnetic compatibility and Radio spectrum Matters (ERM); ElectroMagnetic Compatibility (EMC) standard for radio equipment and services; Part 18: Specific conditions for Terrestrial Trunked Radio (TETRA) equipment". ETSI ETR 300-1 (1997): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Designers' guide; Part 1: Overview, technical description and radio aspects". ETSI ETR 300-2 (1997): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Designers' guide; Part 2: Radio channels, network protocols and service performance". ETSI ETR 300-3 (2000): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Designers' guide; Part 3: Direct Mode Operation (DMO)". ETSI ETR 300-4 (1997): "Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Designers' guide; Part 4: Network management". ETSI ETSI TR 100 392-17-1 V1.1.3 (2005-02) 14 History Document history V1.1.1 May 2001 Publication as TS 100 392-17 V1.1.2 October 2004 Publication V1.1.3 February 2005 Publication |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 1 Scope | The present document is to define a specific resource management protocol to support the transfer ABT-DT capability and to consider some specific cases of its possible application. The material contained in the present document is strictly informative to give guidance for implementation in networks. No additional ATM transfer capabilities as compared to ITU-T Recommendation I.371 [1] are introduced. The present document specifies the functions supported by the ABT-DT protocol, the protocol messages and the protocol procedures. In order to achieve an effective resource management capability, these functions shall be complemented by additional functions (e.g. connection policing, resource allocation) that are network operator specific and are not covered by the present document. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 2 References | References may be made to: a) specific versions of publications (identified by date of publication, edition number, version number, etc.), in which case, subsequent revisions to the referenced document do not apply; or b) all versions up to and including the identified version (identified by "up to and including" before the version identity); or c) all versions subsequent to and including the identified version (identified by "onwards" following the version identity); or d) publications without mention of a specific version, in which case the latest version applies. A non-specific reference to an ETS shall also be taken to refer to later versions published as an EN with the same number. [1] ITU-T Recommendation I.371: "Traffic Control and Congestion Control in B-ISDN"; |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 3 Abbreviations | For the purposes of the present document the following abbreviations apply: ABT ATM Block Transfer ABT-DT ATM Block Transfer with Delayed Transmission ATM Asynchronous Transfer Mode BCR Block Cell Rate B-VPN Broadband Virtual Private Network FRM Fast Resource Management IBT Intrinsic Burst Tolerance OAM Operation And Maintenance QoS Quality of Service RM Resource Management SCR Sustainable Cell Rate TR 101 123 V1.1.1 (1997-11) 7 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 4 General principles | The protocol specified in the present document is an application of Fast Resource Management (FRM) techniques, specifically designed to support the ABT/DT transfer capability. FRM techniques operate on the time scale of the round-trip propagation delay of an ATM connection and allow the modification of the ATM resources allocated to a connection depending on the actual need of the connection and/or on the network status. A basic characteristic of FRM procedure is that the command messages required by the specific protocol are transported "in band" into the data flow by means of specific Resource Management (RM) cells. This is intended to improve the protocol response time. The general format of RM cells is specified in ITU-T Recommendation I.371 [1] as well as the specific RM cell format used by the ABT/DT protocol. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5 ABT/DT Protocol description | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.1 Introduction | This subclause presents the specification of a protocol for supporting the ATM Block Transfer with Delayed Transmission (ABT/DT) capability. The high level description of ABT/DT as well as the definition of the ATM block are given in ITU-T Recommendation I.371 [1]. The basic principle behind ABT/DT is that the information is transmitted in block of cells (ATM Blocks) and that the bandwidth available for transmission, i.e., the Block Cell Rate (BCR), is negotiated individually for each ATM Block. ATM Blocks are delineated by specific messages of the ABT/DT protocol, as will be specified later on in the present document. In the present specification, only point-to-point bi-directional communications are considered and the protocol is limited to a single network; point-to-multipoint communications are for further study. Even though the ABT/DT transfer capability may usually involve bi-directional data transmission, for the sake of simplicity the description provided by the present document refers only to a single direction of data transmission, the generalization to the bi-directional case being straightforward. With this restriction the reference configuration for an ABT/DT connection is shown in figure 1. The terminal originating the data is said to be the Source terminal and the terminal receiving the data is said to be the Destination terminal. Between the two terminals there is a mono-directional data flow (from the source toward the destination) and a bi-directional control flow realized by means of RM cells. The direction of the control flow going from the source toward the destination is said to be the Forward direction; the reverse direction is said to be the Backward direction. Furthermore, the network node closest to the source is said to be the Ingress node; the one closest to the destination is said to be the Egress node. Sourc e Des tination U s er Netw ork Interfac e Ingress N ode Egress node Network ABT/DT U nit (UP C) N ode n ABT /DT U nit (UPC) U s er Netw ork Interface Figure 1: Reference configuration for ABT/DT Each direction of transmission is realized by means of an ABT/DT connection, whose traffic contract specifies: 1) -The peak cell rate value for each cell flow of the connection, namely: - the maximum cell rate Λmax(0+1) of the user generated CLP=0+1 cell flow (including user OAM but no RM cells); TR 101 123 V1.1.1 (1997-11) 8 - the maximum cell rate Λmax(OAM) of the user OAM cell flow. Note that in the signalling message this peak cell rate is declared as a proportion of the aggregate peak cell rate Λmax(0+1). Only the first parameter is mandatory; the second one is optional. Values of these two parameters are upperbounds for the respective negotiable BCR values of the different cell flows. 2) The CDV tolerance associated with each maximum cell rate, namely: - τmax(0+1) of the CLP=0+1 cell flow (mandatory); - τmax(OAM) of the OAM cell flow (optional and possibly implicit). CDV Tolerances may be implicitly or explicitly declared (note 1). 3) The maximum frequency of BCR negotiations by means of: - the peak cell rate 1/TRM of the user request RM cells. 4) The CDV tolerance τRM associated with the peak cell rate of the user request RM cell flow. 5) The committed bandwidth of the CLP=0+1 cell flow by means of a Sustainable Cell Rate (SCR) and an Intrinsic Burst Tolerance (IBT); the SCR may be set equal to 0. 6) The requested QoS class. NOTE 1: The CDV tolerance τ associated with the CLP=0+1 cell flow may depend on the peak cell rate λ of this cell flow, namely τ is a function τ(λ) of λ. In that case, the function τ(λ) is specified on a subscription basis and τmax(0+1)=τ(Λmax(0+1)) . The CDV tolerance associated with OAM traffic should satisfy the general requirements given in Appendix II to ITU-T Recommendation I.371 [1]. An ABT/DT connection is subject to Connection Admission Control (CAC) on the basis of the declared parameters (peak and sustainable cell rates). At the establishment of an ABT/DT connection, the connectivity between the source and the destination is assured but no resources are allocated; the connection is then said dormant. Resources are allocated only via BCR negotiation by means of RM cells and the connection is then said active. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.2 Notation and conventions | The following conventions are used in the present document in order to identify the different cell flows generated by a user terminal: - OAM cell flow: includes only end to end OAM cells generated by the user; - RM cell flow: includes any kind of RM cells generated by the user; - user data CLP=0+1 cell flow: includes all user data cells, regardless the value of the CLP bit; - aggregated CLP=0+1 cell flow: includes all user data cells and OAM cells, but not RM cells. Furthermore, the following symbols will be used in the rest of the present document: - Λ(0+1): a generic value of the BCR for the aggregated CLP=0+1 cell flow; - Λ(OAM): a generic value of the BCR for the OAM cell flow; - Λcurrent(0+1): value of the BCR for the aggregated CLP=0+1 cell flow currently allocated to a connection at the Ingress node; - ΛCurrent(OAM): value of the BCR for the OAM cell flow currently allocated to a connection at the Ingress node; - λCurrent(0+1): value of the BCR for the aggregated CLP=0+1 cell flow currently allocated to a connection at the Egress node; - λCurrent(OAM): value of the BCR for the OAM cell flow currently allocated to a connection at the Egress node; TR 101 123 V1.1.1 (1997-11) 9 - ΛΝ(0+1): value of the BCR for the aggregated CLP=0+1 cell flow currently allocated to a connection at the generic node N; - ΛN(OAM): value of the BCR for the OAM cell flow currently allocated to a connection at the generic node N; - Λ'Current(0+1): value of the BCR for the aggregated CLP=0+1 cell flow currently reserved but not yet allocated to a connection at the Ingress node; - Λ'Current(OAM): value of the BCR for the OAM cell flow currently reserved but not yet allocated to a connection at the Ingress node; - λ'Current(0+1): value of the BCR for the aggregated CLP=0+1 cell flow currently reserved but not yet allocated to a connection at the Egress node; - λ'Current(OAM): value of the BCR for the OAM cell flow currently reserved but not yet allocated to a connection at the Egress node; - ΛReq(0+1): value of the BCR for the aggregated CLP=0+1 cell flow requested by the source; - ΛReq(OAM): value of the BCR for the OAM cell flow requested by the source; - λReq(0+1): value of the BCR for the aggregated CLP=0+1 cell flow requested by the destination; - λReq(OAM): value of the BCR for the OAM cell flow requested by the destination; - Λalloc(0+1): value of the BCR for the aggregated CLP=0+1 cell flow that has been allocated following a negotiation request; - ΛAlloc(OAM): value of the BCR for the OAM cell flow that has been allocated following a negotiation request; |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.3 Functions supported by the ABT/DT protocol | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.3.1 BCR negotiation functions | The BCR negotiation functions are the basic functions supported by the ABT-DT protocol. By means of these functions the user terminals may negotiate with the network new values for the BCR of both the aggregated CLP=0+1 cell flow (Λ(0+1)) and the OAM cell flow (Λ(OAM)). The negotiation may be intended at achieving either a BCR decrease or a BCR increase. The BCR negotiation can be initiated either by the source terminal (forward BCR negotiation) or by the destination terminal (backward BCR negotiation). In the case of concurrent requests, one initiated by the source and the other initiated by the destination, the latter takes the precedence. Finally, the BCR negotiations can be carried out in two different modes: the Rigid mode and the Elastic mode. The rigid mode is characterized by the fact that the BCR value specified by the user in its request can not be modified (i.e., lowered) by the network, so that the request is accepted only if all the network elements along the path of the connection are able to allocate the requested amount of resources. The elastic mode is characterized by the fact that the BCR value specified by the user in its request can be modified (i.e., lowered) by the network, so that a request can be accepted also if some of the network elements along the path of the connection are able to allocate only an amount of resources lower than the one requested by the user. By convention, BCR negotiations intended at achieving a BCR decrease are carried out only in the rigid mode; therefore, there are compulsively six types of BCR negotiation functions: 1) Rigid Forward BCR decrease: allow the source to negotiate a lower BCR for the aggregated CLP=0+1 cell flow and/or for the OAM cell flow; 2) Rigid Forward BCR increase: allow the source to negotiate a higher BCR for the aggregated CLP=0+1 cell flow and/or for the OAM cell flow; TR 101 123 V1.1.1 (1997-11) 10 3) Rigid Backward BCR decrease: allow the destination to negotiate a lower BCR for the aggregated CLP=0+1 cell flow and/or for the OAM cell flow. 4) Rigid Backward BCR increase: allow the destination to negotiate a higher BCR for the aggregated CLP=0+1 cell flow and/or for the OAM cell flow. 5) Elastic Forward BCR increase: allow the source to negotiate a higher BCR for the aggregated CLP=0+1 cell flow and/or for the OAM cell flow; 6) Elastic Backward BCR increase: allow the destination to negotiate a higher BCR for the aggregated CLP=0+1 cell flow and/or for the OAM cell flow. The Rigid Backward BCR decrease function and the Rigid Backward BCR increase function are realized through the same protocol procedure (see subclause 5.6.1.3). A BCR decrease is always in the rigid mode. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.3.2 User maintenance functions | User maintenance functions allow a user terminal to know the BCR value currently allocated within the network and to make it equal all along the connection. These functions can be used if some discrepancy is observed between the BCR values known at the source and at the destination and the BCR value allocated within the network. The use of these functions can not implicate the allocation of new resources to a connection, but only the alignment of the amount of allocated resources to the minimum among the values currently allocated by the different network elements along the path of the connection. User maintenance functions can be initiated either by the source or by the destination and are always carried out in the elastic mode. Therefore, there are two types of user maintenance functions: 1) Forward Status Enquiry: is a user maintenance function initiated by the source; 2) Backward Status Enquiry: is a user maintenance function initiated by the destination; |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.3.3 Traffic control functions | Traffic Control functions are initiated by the network under certain conditions (e.g., block non conformance, presence of congestion etc.) to force a modification (usually a reduction) of the BCR allocated to a connection. These functions are similar to BCR negotiation functions, apart from the fact that they are initiated by a network node instead of a user terminal. They can be initiated either by the ingress node (Forward Traffic Control) or by the egress node (Backward Traffic Control) and can be carried out either in the Rigid mode or in the Elastic mode. Traffic Control functions have priority over BCR Negotiation functions initiated by either user. Specifically, if a BCR Negotiation is in progress while a traffic control procedure is running, then some of the messages corresponding to the BCR Negotiation may be discarded at the Ingress node or at the Egress node. Moreover, upon reception of Traffic Control message at a network node, any BCR reservation performed by either user is cancelled. If a BCR negotiation is initiated by the source (respectively the destination) while a Traffic Control message issued at network ingress (respectively egress), the source (respectively the destination) should receive a Not_Ready message. Finally, Traffic Control messages generated at the Ingress node have priority over those generated at the Egress node. As above, some messages relative to the traffic control procedure at network egress may be discarded. Rigid Traffic Control functions are typically used by the network to verify if the BCR currently allocated to a connection is still available and, if this is not the case, to change it at a lower value. Elastic Traffic Control functions are used by the network to perform an elastic re-negotiation of the BCR allocated to a connection and to notify the new allocated value to the source. The new value may either be lower or higher than the previous one. Summarizing, there are four types of Traffic Control Functions: 1) Rigid Forward Traffic Control: is a Rigid Traffic Control function initiated by the ingress node; 2) Rigid Backward Traffic Control: is a Rigid Traffic Control function initiated by the egress node; 3) Elastic Forward Traffic Control: is an Elastic Traffic Control function initiated by the ingress node; TR 101 123 V1.1.1 (1997-11) 11 4) Elastic Backward Traffic Control: is an Elastic Traffic Control function initiated by the egress node. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.3.4 Network maintenance functions | Network Maintenance functions are used by the network to force to some specific value the BCR allocated to a connection in each network node along the path of the connection or to retry a transaction. These functions are typically used to recover from anomalous conditions and are triggered by the expiration of some of the internal timers of the protocol. These are for instance Time-Out Handling functions. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.3.4.1 Time-out handling functions | Time-Out Handling functions are Network Maintenance functions triggered by the expiration of internal timers of the protocol due, for instance, to the loss of some messages during a BCR negotiation procedure or a Traffic Control procedure. This ETR defines two Time-Out Handling functions: 1) Recovering from Ingress Time-Out: this function is triggered by the expiration of W_Ingress timer at the network ingress (see subclause 5.6.4.1.1); 2) Recovering from Egress Time-Out: this function is triggered by the expiration of W_Egress timer at the network egress (see subclause 5.6.4.1.1). Additional Time-Out Handling functions are for further studies. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.3.4.2 Other Network Maintenance functions | Additional Network Maintenance functions are for further studies. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.4 RM cell format | The RM cell format used for ABT/DT is specified in ITU-T Recommendation I.371 [1]. Within the present document the different fields of the RM cells are referred to according to the following nomenclature: - M.Type: indicates the Req/Ack bit of the Message Type field and discriminates between Request cells (M.Type=0) and Acknowledgement cells (M.Type=1); - M.Dir: indicates the Direction bit of the Message Type field and discriminates between cells flowing in the forward direction (M.Dir=0) and cells flowing in the backward direction (M.Dir=1); - M.Elastic: indicates the Elastic/Rigid bit of the Message Type field and discriminates between the elastic mode (M.Elastic=0) and the Rigid mode (M.Elastic=1); - M.CI: indicates the Congestion Indication bit of the Message Type field and indicates whether the network is congested (M.CI=1) or not (M.CI=0); note that, in the elastic mode, this field can be used to indicate whether the connection has received at least the bandwidth corresponding to its fair share (M.CI=0) or not (M.CI=1); - M.Maintenance: indicates the Maintenance bit of the Message Type field and discriminates between RM cells devoted to user or network maintenance functions (M. Maintenance=1) and RM cells devoted to BCR negotiation and traffic control functions (M.Maintenance=0); - M.Trf: indicates the Traffic Management bit of the Message Type header and discriminates between RM cells devoted to Network Traffic Management functions (M.Trf=1) and RM cells devoted to BCR negotiation or user maintenance functions (M.Trf=0); - M.Rate(0+1): indicates the CLP=0+1 BCR field; - M.Rate(OAM): indicates the User OAM BCR field; - M.Number: indicate the Sequence Number field; when used, this field is incremented by one (modulo 232) in each subsequent RM cell transmitted by a given entity. On the contrary, when not used, it is set equal to 0; - M.Size: indicate the Block Size field; this field is not used in this version of the protocol. TR 101 123 V1.1.1 (1997-11) 12 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5 Protocol messages | This subclause defines the protocol messages used for implementing the functions described in subclause 5.3. Each message is identified by the values of the Message Type binary fields in the RM cells. Messages are classified into three main categories according to their utilization: - BCR Negotiation messages; - User Maintenance messages; - Network Traffic Management messages, and are listed in the form of tables that specify the coding of the Message Type binary fields used by each message. The characters "--" mean that either the value 0 or the value 1 can be used for the corresponding field and that the value can be changed within the network. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.1 BCR negotiation messages | These messages are identified by M.Maintenance=0 and M.Trf=0. The M.Elastic field discriminates between messages related to elastic negotiation functions (M.Elastic=0) and messages related to rigid negotiation functions (M.Elastic=1). The M.Dir field discriminates between messages flowing in the forward direction (M.Dir=0) and messages flowing in the backward direction (M.Dir=1). The M.Type field discriminates between request messages (M.Type=0) and Acknowledgement messages (M.Type=1). The M.CI field does not identifies any specific message, but it is used to indicate whether congestion has been encountered in the network (M.CI=1) or not. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.1.1 Rigid BCR negotiation | The four messages listed in table 1 are devoted to support the Rigid BCR Negotiation functions defined in subclause 5.3.1. Two of them are Bandwidth Request messages (M.Type=0), whereas the remaining two messages are Bandwidth Acknowledgement messages (M.Type=1). Table 1: Messages for Rigid BCR Negotiation Primitive’s name M.Type M.Dir M.Elastic M.CI M.Maint. M.Trf R_Forward_User_Req 0 0 1 -- 0 0 R_Backward_User_Req 0 1 1 -- 0 0 R_Backward_Net_Ack 1 1 1 -- 0 0 R_Forward_User_Ack 1 0 1 -- 0 0 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.1.2 Elastic BCR negotiation | The four messages listed in table 2 are devoted to support the Elastic BCR Negotiation functions defined in subclause 5.3.1.2. Two of them are Bandwidth Request messages (M.Type=0), whereas the remaining two messages are Bandwidth Acknowledgement messages (M.Type=1). Table 2: Messages for Elastic BCR Negotiation Primitive’s name M.Type M.Dir M.Elastic M.CI M.Maint. M.Trf E_Forward_User_Req 0 0 0 -- 0 0 E_Backward_User_Req 0 1 0 -- 0 0 E_Backward_Net_Ack 1 1 0 -- 0 0 E_Forward_User_Ack 1 0 0 -- 0 0 TR 101 123 V1.1.1 (1997-11) 13 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.2 User maintenance messages | These messages are identified by M.Maintenance=1 and M.Trf=0. The M.Elastic field is always set to 0. The M.Dir field discriminates between messages flowing in the forward direction (M.Dir=0) and messages flowing in the backward direction (M.Dir=1). The M.Type field discriminates between request messages (M.Type=0) and Acknowledgement or Not Ready messages (M.Type=1). The M.CI field is always set to 0. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.2.1 Forward and backward status enquiry | The four messages listed in table 3 are devoted to support the Forward and Backward Status Enquiry functions defined in subclause 5.3.2. Two of them are Request messages (M.Type=0), whereas the remaining two messages are Acknowledgement messages (M.Type=1). Table 3: Messages for forward and backward status enquiry functions Primitive’s name M.Type M.Dir M.Elastic M.CI M.Maint. M.Trf M_Forward_User_Req 0 0 0 0 1 0 M_Backward_User_Req 0 1 0 0 1 0 M_Backward_Net_Ack 1 1 0 0 1 0 M_Forward_User_Ack 1 0 0 0 1 0 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.2.2 Not ready | The two messages listed in table 4 are the Not Ready messages generated by the network ingress or the network egress nodes when they are unable to accept the BCR negotiation request forwarded by the source or by the destination. The M.Elastic field should be set equal to 0. Table 4: Not ready messages Primitive’s name M.Type M.Dir M.Elastic M.CI M.Maint. M.Trf M_Forward_Notready_Ack 1 0 0 1 1 0 M_Backward_Notready_Ack 1 1 0 1 1 0 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.3 Traffic control messages | These messages are identified by M.Maintenance=0 and M.Trf=1. The M.Elastic field discriminates between messages related to the Elastic mode (M.Elastic=0) end messages related to the rigid mode (M.Elastic=1). The M.Dir field discriminates between messages flowing in the forward direction (M.Dir=0) and messages flowing in the backward direction (M.Dir=1). The M.Type field discriminates between request messages (M.Type=0) and Acknowledgement or Not Ready messages (M.Type=1). The M.CI field does not identify any specific message, but it is used to indicate whether congestion has been encountered in the network (M.CI=1) or not (M.CI=0). |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.3.1 Rigid traffic control | The four messages listed in table 5 are devoted to support the Rigid Traffic Control functions defined in subclause 5.3.3. Two of them are Request messages (M.Type=0), whereas the remaining two messages are Acknowledgement messages (M.Type=1). TR 101 123 V1.1.1 (1997-11) 14 Table 5: Messages for rigid traffic control functions Primitive’s name M.Type M.Dir M.Elastic M.CI M.Maint. M.Trf R_Traffic_Forward_Req 0 0 1 -- 0 1 R_Traffic_Backward_Req 0 1 1 -- 0 1 R_Traffic_Backward_Ack 1 1 1 -- 0 1 R_Traffic_Forward_Ack 1 0 1 -- 0 1 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.3.2 Elastic traffic control | The four messages listed in table 6 are devoted to support the Elastic Traffic Control functions defined in subclause 5.3.3. Two of them are Request messages (M.Type=0), whereas the remaining two messages are Acknowledgement messages (M.Type=1). Table 6: Messages for elastic traffic control functions Primitive’s name M.Type M.Dir M.Elastic M.CI M.Maint. M.Trf E_Traffic_Forward_Req 0 0 0 -- 0 1 E_Traffic_Backward_Req 0 1 0 -- 0 1 E_Traffic_Backward_Ack 1 1 0 -- 0 1 E_Traffic_Forward_Ack 1 0 0 -- 0 1 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.4 Network maintenance messages | These messages are identified by M.Maintenance=1 and M.Trf=1. The M.Elastic field is takes the value of the M.Elastic field of the transaction which is pending. The M.Dir field discriminates between messages flowing in the forward direction (M.Dir=0) and messages flowing in the backward direction (M.Dir=1). The M.Type field discriminates between request messages (M.Type=0) and Acknowledgement or Not Ready messages (M.Type=1). The M.CI field does not identifies any specific message, but it is used to indicate whether congestion has been encountered in the network (M.CI=1) or not. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.5.4.1 Time-out handling | The four messages listed in table 7 are devoted to support the Time-Out Handling functions defined in subclause 5.3.4.1. Two of them are Request messages (M.Type=0), whereas the remaining two messages are Acknowledgement messages (M.Type=1). Table 7: Messages for time-out handling functions Primitive’s name M.Type M.Dir M.Elastic M.CI M.Maint. M.Trf N_Forward_Req 0 0 -- -- 1 1 N_Backward_Req 0 1 -- -- 1 1 N_Backward_Ack 1 1 -- -- 1 1 N_Forward_Ack 1 0 -- -- 1 1 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6 Normal protocol procedures | This subclause describes the protocol procedures executed under normal conditions, i.e., in the assumption that no collisions happen between different procedures. The problem of collisions will be dealt with in subclause 5.7. TR 101 123 V1.1.1 (1997-11) 15 5.6.1 BCR negotiation procedures |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.1 General principles | Any BCR negotiation (BCR increase or decrease) by either user of a bi-directional ABT/DT communication is achieved by sending a Bandwidth Request message into the network and by waiting for a Bandwidth Acknowledgement message from the network, except in the case of a BCR decrease initiated by the source itself, which is immediately taken into account by the network. If positive, this Bandwidth Acknowledgement message should be acknowledged by the source by sending another Bandwidth Acknowledgement message. The source can start the transmission of a new ATM block only after having sent this Bandwidth Acknowledgement message. ATM blocks are delineated by one of the following messages: - Bandwidth Request messages indicating a BCR decrease initiated by the source; - Bandwidth Acknowledgement messages sent by the source. A Bandwidth Request message when sent by a given entity (either the source or the destination) should convey the respective requested BCR values for the CLP=0+1 and OAM cell flows. The BCR values conveyed by a Bandwidth Request message shall be lower than the maximum values negotiated at connection set up ΛMax(0+1) and ΛMax(OAM); therefore, the fields M.Rate(0+1) and M.Rate(OAM) of these messages are checked and enforced at the Ingress node (Forward Bandwidth Request messages) or at the Egress node (Backward Bandwidth Request messages) as follows: { } { } M.Rate 0 1 M.Rate 0 1 0 +1 M.Rate OAM M.Rate OAM OAM M.Rate 0 1 ( ): min ( ), ( ) , ( ): min ( ), ( ), ( ) . max max + = + = + Λ Λ (1) A Bandwidth Acknowledgement message or a Bandwidth Request message when received by the destination should convey the respective allocated BCR values ΛAlloc(0+1) and ΛAlloc(OAM) for the CLP=0+1 and OAM cell flows. The fields M.Rate(0+1) and M.Rate(OAM) of Bandwidth Acknowledgement messages sent by the source are checked at the ingress node and forced according to: { } { } M. Rate 0 1 0 1 M. Rate(0 + 1) M. Rate OAM OAM M. Rate(OAM), M. Rate(0 + 1) Current Current ( ): min ' ( ), , ( ): min ' ( ), . + = + = Λ Λ (2) Upon the reception of a Bandwidth Request message from the user (either the source or the destination) each network node establishes whether the request is a Bandwidth Increase Request or a Bandwidth Decrease Request by looking at the fields M.Rate(0+1) and M.Rate(OAM), if M.Rate(0+1) ≤ ΛCurrent(0+1) (3) then the request is a Bandwidth Decrease Request; otherwise it is a Bandwidth Increase Request. By convention, Bandwidth Decrease Requests shall be forwarded in the rigid mode only; it is an implementation option how to deal with Bandwidth Decrease Requests forwarded in the elastic mode; for instance, these requests can be: − discarded at the Ingress or Egress node; − changed into rigid requests by forcing the M.Elastic field to 0 at the Ingress or Egress node. Only one BCR negotiation initiated by a given entity (the source or the destination) may be in progress within the network in a given direction. If the same entity initiates a BCR negotiation in a given direction while another one is still in progress within the network, then the network discards the new Bandwidth Request message and sends a Not Ready message back to the originating entity. TR 101 123 V1.1.1 (1997-11) 16 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.2 Rigid BCR negotiation procedures | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.2.1 Rigid forward BCR decrease | A Rigid Forward BCR Decrease procedure is initiated by the source by sending a R_Forward_User_Req message (see table 1) which satisfies (3). Immediately after issuing such a message, the source should adapt its transmission rate in order to conform to the new BCR values (and associated CDV tolerances), since the message delineates a new ATM block. The flow of messages for the Rigid Forward BCR Increase procedure is shown in figure 3. R_Forw ard_User_Req Network Beginning of the ATM block at the U NI Source U N I Ingress N ode Egress N ode U N I D estination R_Forw ard_User_Req Figure 2: Rigid Forward BCR Decrease procedure with no collision At the Ingress node, the requested BCR values are enforced as in (1) and then checked to see if inequality (3) is satisfied. If this is the case, then ΛCurrent(0+1) and ΛCurrent(OAM) are updated to the new BCR values and the message is forwarded to the network nodes. Each node receiving the message checks that inequality (3) is verified and immediately updates the bandwidth allocated to the connection. . A BCR decrease is not acknowledged by the network. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.2.2 Rigid forward BCR increase | A Rigid Forward BCR Increase procedure is initiated by the source by sending a R_Forward_User_Req (see table 1) message with: M.Rate(0+1) > Λcurrent(0+1) In this case, the request message does not delineate a new ATM block and the source shall wait for the acknowledgement from the network before updating its transmission rate. The flow of messages for the Rigid Forward BCR Increase procedure is shown in figure 3. TR 101 123 V1.1.1 (1997-11) 17 Network Be ginning of th e A TM bloc k a t th e U N I S ource U N I Ingress N ode E gress N ode U N I D e stination R_Forward_User_Req R_ Fo rward_ User_Req (if M.CI=0) R_Backward_N et_A ck R_Forward_U ser_ Ack (if M.CI = 0) R_Forwa rd_User_A ck Figure 3: Rigid Forward BCR Increase procedure with no collision At the Ingress node the requested BCR values are enforced according to (1) and checked to see whether inequality (3) is satisfied. Since the message initiates a BCR Increase procedure inequality (3) will not be satisfied. The Ingress node then processes the message and accepts or denies the bandwidth increase request according to a specific resource allocation policy. If the requested resources are available, then they are reserved (but not yet allocated); otherwise, the request is denied and the M.CI field is set equal to 1. In both cases the message is forwarded into the network. Each node receiving the message checks first the M.CI field: if the M.CI field is equal to 1 the message is forwarded to the next node and no other actions are taken (the bandwidth increase request has already been rejected). Otherwise the message is processed to check the availability of the requested resources as in the Ingress node and it is then forwarded to the next node. A R_Forward_User_Req message reaching the Egress node with M.CI=1 is discarded after having generated the transmission of a R_Backward_Net_Ack message with M.CI=1 back to the source. The R_Backward_Net_Ack message travelling into the network cancels the bandwidth reservation in all the nodes that had been able to reserve the requested resources and it is then forwarded to the source that does not has any other action to take. The reception of this message terminates the BCR negotiation phase. A R_Forward_User_Req message reaching the Egress node with M.CI=0 is processed as usual. If the requested resources are not available the message is discarded and a R_Backward_Net_Ack message with M.CI=1 is sent back to the source as in the previous case. Otherwise, the message is forwarded to the destination and a R_Backward_Net_Ack message with M.CI=0 is sent back to the source. This message carries into the M.Rate(0+1) and M.Rate(OAM) fields the BCR values that have been reserved for the connection. Each node receiving this message forwards it to the following node (in the backward direction) without the need to take any special action (it may optionally read the reserved BCR values and update its internal values). Once the message is received by the source, it triggers the transmission of a R_Forward_User_Ack message with M.CI=0. This message should carry values lower than or equal to the reserved BCR and delineates a new ATM block. Immediately after its transmission the source is allowed to update its transmission rate. At the Ingress node the M.Rate(0+1) and M.Rate(OAM) fields of the R_Forward_User_Ack message are enforced according to (2); then the reserved resources are actually allocated to the connection and the message is forwarded into the network up to the destination. Each node receiving the message allocates the reserved resources. The reception of this message terminates the BCR negotiation phase. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.2.3 Rigid backward BCR increase or decrease | A Rigid Backward BCR Increase or Decrease procedure is initiated by the destination by sending a R_Backward_User_Req message. TR 101 123 V1.1.1 (1997-11) 18 The flow of messages for this procedure is shown in figure 4. Source Destination Ingress Node Egress Node UNI UNI R_Backward_User_Req R_Forward_User_Ack (If M.CI = 1) R_Backward_User_Req (If M.CI = 0) R_Forward_User_Ack (If M.CI = 0) Beginning of the ATM Block at the UNI Figure 4: Rigid Backward BCR Increase or Decrease procedure with no collision At the Egress Node the requested BCR values are enforced according to (1) and checked to see whether inequality (3) is satisfied. If inequality (3) is satisfied then the request is accepted. If inequality (3) is not satisfied, then the request is processed according to a specific resource allocation policy to check whether the requested resources are available. If this is the case the request is accepted, otherwise it is denied. For any accepted request, the requested resources are reserved (but not yet allocated) and the R_Backward_User_Req message is forwarded to the next node toward the source with M.CI=0. For any denied request no resources are reserved and the R_Backward_User_Req message is forwarded to the next node toward the source with M.CI=1. Each node receiving the message checks first the M.CI field: if M.CI=1 the message is forwarded to the next node and no other actions are taken (the bandwidth negotiation request has already been rejected). Otherwise the message is processed to check whether the request can be accepted or not as in the Ingress node and it is then forwarded to the next node. An R_Backward_User_Req message reaching the Ingress node with M.CI=1 is not forwarded to the source but generates the transmission of an R_Forward_User_Ack message with M.CI=1 back to the destination. The R_Forward_User_Ack message travelling into the network cancels the bandwidth reservation in all the nodes that had been able to reserve the requested resources and it is then forwarded to the destination that does not has any other action to take. The reception of this message terminates the BCR negotiation phase. An R_Backward_User_Req message reaching the Ingress node with M.CI=0 is processed as usual. If the request can not be accepted the message is not forwarded to the source but an R_Forward_User_Ack message with M.CI=1 is sent back to the destination as in the previous case. Otherwise, the message is forwarded to the source with M.CI=0. The source is then required to send a R_Forward_User_Ack message with M.CI=0 toward the destination. This message delineates a new ATM block and the source shall immediately update its transmission rate after having sent it. The R_Forward_User_Ack message sent by the source shall carry into the M.Rate(0+1) and M.Rate(OAM) fields BCR values lower or equal to the ones that have been reserved for the connection; these values are enforced at the Ingress node according to (2). Each node receiving this message allocates the reserved resources to the connection and forwards the message downstream. The reception of this message terminates the BCR negotiation phase. Once the message has reached the Egress node, it is forwarded to the destination and the procedure terminates. TR 101 123 V1.1.1 (1997-11) 19 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.3 Elastic BCR negotiation procedures | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.3.1 Elastic forward BCR increase | The Elastic Forward BCR Increase procedure is initiated by the source sending an E_Forward_User_Req message. The flow of messages for this procedure is shown in figure 5. Network Beg inning of the A TM block at the U N I Source U N I Ingress N od e Eg ress N o de UN I D estina tion E _Forward_User_Req E_Forw ard_User_Req E_Backw ard_Netr_Ack E _Forward_User_Ack E _Forw ard_User_Ack Figure 5: Elastic Forward BCR increase procedure with no collision The Elastic Forward BCR Increase procedure differs from the corresponding rigid procedure in the following points: 1) the set of messages used by this procedure is the one in table 2 instead of the one in table 1 that are used in the rigid procedure; 2) in the elastic mode the source requests are never denied; each network node updates the M.Rate(0+1) and M.Rate(OAM) fields to indicate the amount of available resources. In particular, the network node may allocate resources between elastic requests in order to achieve some fair share of the available bandwidth. The M.CI field may then be used (i.e., set equal to 1) to indicate that the BCR reserved for the CLP=0+1 cell flow is less than the BCR that would have been obtained under fair share conditions. This is a network option and the definition of fairness is definitely network operator specific. Setting the M.CI field is simply intended to indicate to the source that it could obtain more resources in a next BCR negotiation and to incite downstream network nodes to provision some resources for the connection considered; 3) the E_Forward_User_Req message is always forwarded to the destination, regardless the value of the M.CI field; 4) each node shall either update its internal values of the reserved bandwidth upon the reception of an E_Backward_Net_Ack message (according to the M.Rate(0+1) and M.Rate(OAM) fields) or allocate resources on the basis of the BCR values carried by the E_Forward_User_Ack rather than on the basis of the reserved values. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.1.3.2 Elastic backward BCR increase | The Elastic Backward BCR Increase procedure is initiated by the destination sending an E_Backward_User_Req message. The flow of messages for this procedure is shown in figure 6. TR 101 123 V1.1.1 (1997-11) 20 Network Beg inning of the A TM block at the U N I Source U N I Ingress N od e Eg ress N o de UN I D estina tion E _B ackw ard_User_Req E _Backward_User_R eq E_Forw ard_User_Ack E _Forw ard_User_Ack Figure 6: Elastic Backward BCR increase procedure with no collision The Elastic Backward BCR Increase procedure differs from the corresponding rigid procedure in the following points: 1) the set of messages used by this procedure is the one in table 2 instead of the one in table 1 that are used in the rigid procedure; 2) in the elastic mode the destination requests are never denied; each network node updates the M.Rate(0+1) and M.Rate(OAM) fields to indicate the amount of available resources. In particular, the network node may allocate resources between elastic requests in order to achieve some fair share of the available bandwidth. The M.CI field may then be used (i.e., set equal to 1) to indicate that the BCR reserved for the CLP=0+1 cell flow is less than the BCR that would have been obtained under fair share conditions. This is a network option and the definition of fairness is definitely network operator specific. Setting the M.CI field is simply intended to indicate to the destination that it could obtain more resources in a next BCR negotiation and to incite downstream network nodes to provide some resources for the connection considered; 3) at the reception of an E_Backward_User_Req message the Ingress node never sends an E_Forward_User_Ack message back to the destination, regardless the value of the M.CI field in the received message, but forwards the an E_Backward_User_Req message to the source; 4) each node shall allocate resources on the basis of the BCR values carried by the E_Forward_User_Ack message rather than on the basis of the reserved values. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.2 User maintenance procedures | User Maintenance procedures are identical to Elastic BCR Negotiation procedures apart from the following points: 1) the set of messages used by this procedure is the one in table 3 instead of the one in table 2; 2) maintenance messages do not imply any negotiation of the BCR values allocated to the connection. Upon reception of a Maintenance Request message in a network node, the resource allocation algorithm is disabled and for each cell flow, the corresponding BCR value in the Request message is updated to the minimum of the value contained in Request message and of that allocated in the network node. In other words, the BCR values are updated as: { } { } M Rate M Rate M Rate M Rate N N . ( ): min . ( ), ( ) , . ( ): min . ( ), ( ) . 0 1 0 1 0 + 1 OAM OAM OAM + = + = Λ Λ (4) 3) Maintenance messages shall never have the M.CI field set to 1. TR 101 123 V1.1.1 (1997-11) 21 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.2.1 Forward status enquiry procedures | The Forward Status Enquiry procedure is identical to the Elastic Forward BCR Increase procedure apart from the points mentioned above. The flow of messages for this procedure is identical to the one shown in figure 5, provided that the corresponding messages of table 3 are substituted to the one shown in that figure. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.2.2 Backward status enquiry procedures | The Backward Status Enquiry procedure is identical to the Elastic Backward BCR Increase procedure apart from the points mentioned above. The flow of messages for this procedure is identical to the one shown in figure 6, provided that the corresponding messages of table 3 are substituted to the one shown in that figure. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.3 Traffic control procedures | Traffic Control procedures are similar to BCR negotiation procedures, with the difference that they are initiated by the Ingress node or the Egress node instead of by the source or the destination. In addition, Traffic Control messages shall have the M.Number field properly set, since Traffic Control procedures make use of this field. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.3.1 Rigid traffic control procedures | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.3.1.1 Rigid forward traffic control | A Rigid Forward Traffic Control procedure is initiated by the Ingress node by sending into the forward direction an R_Traffic_Forward_Req. The flow of messages for this procedure is shown in figure 7. Source Destination Ingress Node Egress Node UNI UNI Beginning of the ATM Block at the UNI R_Traffic_Forward_Req R_Traffic_Backward_Ack R_Traffic_Forward_Ack R_Traffic_Forward_Req R_Traffic_Forward_Ack Figure 7: Rigid Forward Traffic Control procedure with no collision The R_Traffic_Forward_Req message sent by the Ingress node shall have M.Rate (0+1)= Λcurrent(0+1), M.Rate(OAM)= ΛCurrent(OAM), and M.Number set to the proper progressive number. Upon reception of this message in a network node, the requested resources are examined. If they are not available according to the resource allocation algorithm, the M.CI field is set and a possible BCR for the CLP=0+1 cell flow is computed. This new value is overwritten in the M.Rate(0+1) field. TR 101 123 V1.1.1 (1997-11) 22 Once received at the Egress node, the request message is forwarded to the destination. In addition, the Egress node shall send back a R_Traffic_Backward_Ack message having the same values of the fields M.CI, M.Rate(0+1) and M.Rate(OAM). Upon reception of this message at the Ingress node, the Ingress node computes the new BCR values to be proposed to the source, writes them in the M.Rate(0+1) and M.Rate(OAM) fields and forwards the message to the source. If M.CI=1, these BCR values depend on the policing strategy adopted by the network. For instance, they may be equal to 0, to minimum BCR values, or to the possible BCR values computed in the network nodes. In the contrary, if M.CI=0 the BCR values are exactly the original values Λcurrent(0+1) and Λcurrent(OAM). The R_Traffic_Backward_Ack message should itself be acknowledged by the source by sending an R_Traffic_Forward_Ack message which is the leading RM cell of a new ATM block. The BCR values acknowledged by the source should be less than or equal to the BCR values proposed by the network and are enforced at the Ingress node according to (2). Upon the reception of this message each network node modifies the resources allocated to the connection according to the BCR values carried in the message and forwards the message to the downstream node. Once the message is received by the destination the procedure terminates. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.3.1.2 Rigid backward traffic control | A Rigid Backward Traffic Control procedure is initiated by the Egress node by sending into the backward direction an R_Traffic_Backward_Req. The flow of messages for this procedure is shown in figure 8. Source Destination Ingress Node Egress Node UNI UNI Beginning of the ATM Block at the UNI R_Traffic_Backward_Req R_Traffic_Forward_Ack R_Traffic_Forward_Ack R_Traffic_Backward_Req Figure 8: Rigid Backward Traffic Control procedure with no collision The R_Traffic_Backward_Req message sent by the Egress node shall have M.Rate (0+1)= λcurrent(0+1), M.Rate(OAM)= λCurrent(OAM), and M.Number set to the proper progressive number. This request message is handled as the corresponding request message in the Rigid Forward Traffic Control procedure described in subclause 5.6.3.1.1 (computation of possible BCR values in each network node). In particular, if the requested resources are not available, the M.CI field is set. Upon reception of this R_Traffic_Backward_Req message at the ingress node, if M.CI=0 the BCR ΛCurrent(0+1) and ΛCurrent(OAM) are updated according to the values of the M.Rate(0+1) and M.Rate(OAM) fields and the message is forwarded to the source. Note that for setting the M.CI field, the Ingress node should take into account the BCR parameters of the UPC. If M.CI=1, the R_Traffic_Backward_Req message is forwarded to the source with some proposed BCR values which depend on the network policing strategy. Upon reception of the R_Traffic_Backward_Req message the source shall send a R_Traffic_Forward_Ack message with the same value of the M.CI field. This message delineates a new ATM block. The BCR values conveyed by this message should be less than or equal to the values suggested by the Ingress node and are enforced at the Ingress node according to (2). Upon reception of this message each network node modifies the resources allocated to the connection according to the BCR values carried in the TR 101 123 V1.1.1 (1997-11) 23 message and forward the message to the downstream node. Once the message is received by the destination the procedure terminates. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.3.2 Elastic traffic control procedures | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.3.2.1 Elastic forward traffic control | An Elastic Forward Traffic Control procedure is initiated by the Ingress node by sending into the forward direction a R_Traffic_Forward_Req. The flow of messages for this procedure is shown in figure 9. Network Beg inning of the A TM block at the U N I Source U N I Ingress N od e Eg ress N o de UN I D estina tion E _Traffic_Forward_Req E _Traffic_Backward_Ack E _Traffic_Forward_Ack E _Traffic_Forw ard_Ack E _Traffic_Forward_Req Figure 9: Elastic Forward Traffic Control procedure with no collision In the elastic mode, the BCR values contained in the R_Traffic_Forward_Req message transmitted from the Ingress node may be greater than the BCR values currently allocated but shall be less than or equal to the values Λmax(0+1) and Λmax(OAM) negotiated at connection set up. Upon reception of this message, some resources are reserved in the network nodes by applying the resource allocation algorithm as for a usual BCR negotiation in the elastic mode (also the M.CI field is handled in the same way). Once the request message has been processed at the Egress node, it is forwarded to the destination for information. In addition, the Egress node shall send back toward the source an E_Traffic_Backward_Ack message with the same values of the M.CI, M.Rate(0+1) and M.Rate(OAM) fields. Each node, receiving this message, may update its internal values of the reserved bandwidth accordingly. Upon reception of this message at the Ingress node, this message is forwarded to the source with the proposed BCR values for the CLP=0+1 and OAM cell flows. The source shall acknowledge this message by sending an E_Traffic_Forward_Ack message that delineates a new ATM block. The acknowledged BCR values should be less than or equal to those proposed by the network and are enforced at the Ingress node according to (2). At the reception of this message each network node allocates the requested resources to the connection and forwards the message to the downstream node. Once the message is received by the destination the procedure terminates. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.3.2.2 Elastic backward traffic control | An Elastic Backward Traffic Control procedure is initiated by the Egress node by sending into the backward direction an R_Traffic_Backward_Req. The flow of messages for this procedure is shown in figure 10. TR 101 123 V1.1.1 (1997-11) 24 Network Beg inning of the A TM block at the U N I Source U N I Ingress N od e Eg ress N o de UN I D estina tion E _Traffic_Backward _Req E _Traffic_Forw ard_Ack E _Traffic_B ackward_Req E _Traffic_Forward_Ack Figure 10: Elastic Backward Traffic Control procedure with no collision In the elastic mode, the BCR values contained in the E_Traffic_Backward_Req message transmitted from the Egress node may be greater than the BCR values currently allocated at network egress but shall be less than or equal to the values Λmax(0+1) and Λmax(OAM) negotiated at connection set up. Upon reception of this message, some resources are reserved in the network nodes by applying the resource allocation algorithm as for a usual BCR negotiation in the elastic mode (also the M.CI field is handled in the same way). Once this request message has been processed at the Ingress node, it is forwarded to the source with the proposed BCR values for the CLP=0+1 and OAM cell flows. The source shall acknowledge this message by sending an E_Traffic_Forward_Ack that delineates a new ATM block. The acknowledged BCR values should be less than or equal to the BCR values proposed by the network and are enforced at the Ingress node according to (2). At the reception of this message each network node allocates the requested resources to the connection and forwards the message to the downstream node. Once the message is received by the destination the procedure terminates. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.4 Network maintenance procedures | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.4.1 Time-out handling procedures | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.4.1.1 Timers associated with ABT/DT procedures | Some protocol messages (either Bandwidth Request or Bandwidth Acknowledgement messages) may be lost within the network. In that case, after initiating a given ABT/DT procedure, it may happen that a given entity (the source, the destination, a network node, etc.) does not get any response from the network or either user. To overcome the problems due to the loss of messages or to the absence of response from an entity, some timers are introduced as follows: 1) W_Source: after initiating a BCR negotiation, the source should get a response from the network within a time interval at most W_Source time unit long. Timer W_Source is activated each time the source initiates an ABT/DT procedure (i.e., immediately after having sent a R_Forward_User_Req message, or an E_Forward_User_Req message or a M_Forward_User_Req message); the timer is deactivated at the reception of one of the following messages: - the backward acknowledgement message corresponding to the forward request message sent by the source (i.e. a R_Backward_Net_Ack message or an E_Backward_Net_Ack message or a M_Backward_Net_Ack message); this happens in case of normal completion of the procedure; - an M_Backward_Notready_Ack; this happens in the case of collision of the request at the Ingress node (see subclause 5.7.1). Deactivation of the timer W_Source upon reception of this message is an implementation option that does not affect the behaviour of the protocol; - an R_Backward_User_Req message or an E_Backward_User_Req message; this happens in case of collision within the network with a backward BCR negotiation request; TR 101 123 V1.1.1 (1997-11) 25 - an R_Traffic_Backward_Req message or an E_Traffic_Backward_Req message; this happens in case of collision within the network with a backward Traffic Control request; - an N_Backward_Ack message; this happens in case of expiration of timer W_Ingress (see subclause 5.6.4.1.2). Actions taken by the source in case timer W_Source expires are implementation specific. 2) W_Destination: similarly to the source, after initiating a BCR negotiation, the destination should get a response from the network within a time interval at most W_Destination time unit long. Timer W_Destination is activated each time the destination initiates an ABT/DT procedure (i.e., immediately after having sent a R_Backward_User_Req message, or an E_Backward_User_Req message or a M_Backward_User_Req message); the timer is deactivated upon reception of one of the following messages: − the forward acknowledgement message corresponding to the backward request message sent by the source (i.e. a R_Forward_User_Ack message or an E_Forward_User_Ack message or a M_Forward_User_Ack message); this happens in the case of a normal completion of the procedure; - an M_Forward_Notready_Ack; this happens in the case of collision of the request at the Egress node (see subclause 5.7.2). Deactivation of the timer W_Destination upon reception of this message is an implementation option that does not affect the behaviour of the protocol; - an R_Traffic_Forward_Req message or an E_Traffic_Forward_Req message; this happens in the case of collision within the network with a forward Traffic Control request; - an N_Forward_Ack message; this happens in case of expiration of timer W_Egress (see subclause 5.6.4.1.2). Actions taken by the destination in case timer W_Destination expires are implementation specific. 3) W_Ingress: after sending a Request message (either a Bandwidth Request or a Traffic Control Request) into the network, the Ingress node should get an answer within a time interval at most W_Ingress time units long. Timer W_Ingress is activated at the Ingress node immediately after the transmission of an R_Forward_User_Req message, or an E_Forward_User_Req message or a R_Traffic_Forward_Req message or an E_Traffic_Forward_Req message or an N_Forward_Req message; the timer is deactivated upon reception of one of the following messages: - an R_Backward_Net_Ack message or an _E_Backward_Net_Ack message; this happens if the timer was previously activated by a R_Forward_User_Req message or an E_Forward_User_Req message and no collisions occur neither with a backward BCR negotiation request nor with a backward Traffic Control request; - an R_Backward_User_Req message or an E_Backward_User_Req message; this happens if the timer was previously activated by an R_Forward_User_Req message or an E_Forward_User_Req message and a collision occurs with a backward BCR negotiation request but not with a backward Traffic Control request; - an R_Traffic_Backward_Req message or an E_Traffic_Backward_Req message; this happens if the timer was previously activated by an R_Forward_User_Req message or an E_Forward_User_Req message and a collision occurs with a backward Traffic Control request initiated by the destination; - an R_Traffic_Backward_Ack message or an E_Traffic_Backward_Ack message; this happens if the timer was previously activated by a forward Traffic Control request; - a N_Backward_Ack message; this happens if the timer was previously activated by a N_Forward_Req message. In case timer W_Ingress expires the ingress node initiates an Ingress Time out Recovering procedure as described in subclause 5.6.4.1.2. 4) W_Egress: similarly to what happens at the ingress node, after sending a Request message (either a Bandwidth or a Traffic Control Request) into the network the Egress node should get an answer within a time interval at most W_Egress time units long. The same principle holds after the transmission of a backward acknowledgement cell. Summarizing, timer W_Egress is activated at the Egress node immediately after the transmission of one of the following messages: TR 101 123 V1.1.1 (1997-11) 26 - an R_Backward_User_Req message or an E_Backward_User_Req message; - an R_Traffic_Backward_Req message or an E_Traffic_Backward_Req message; - an N_Backward_Req message; - an R_Backward_Net_Ack message with M.CI=0; - an E_Backward_Net_Ack message; - an M_Backward_Net_Ack message; - an R_Traffic_Backward_Ack message or an E_Traffic_Backward_Ack message. The timer is deactivated at the reception of one of the following messages: - an R_Forward_User_Ack message or an E_Forward_User_Ack message; this happens if the timer was previously activated by a backward BCR request message or by a backward BCR acknowledgement message; - an R_Traffic_Forward_Req message or an E_Traffic_Forward_Req message; this happens if the timer was previously activated by a backward BCR request message or by a backward Traffic Control request message and a collision occurs with a forward Traffic Control request; - an R_Traffic_Forward_Ack message or an E_Traffic_Forward_Ack message; this happens if the timer was previously activated by a backward Traffic Control request and no collisions occur with forward Traffic Control Request; - an N_Forward_Ack message; this occurs if the timer was previously activated by an N_Backward_Req message. In the case timer W_Egress expires the egress node initiates an Egress Time out Recovering procedure as described in subclause 5.6.4.1.3. Note that Traffic Control messages should have correct Sequence Number in order to be taken into account at the Egress or Ingress node. Traffic Control messages with incorrect Sequence Numbers are ignored. The meaning of the above timers is illustrated in figure 11 for the case of a Rigid Forward BCR Negotiation and in figure 12 for the case of a Rigid Backward BCR Negotiation. Source Destination Ingress Node Egress Node UNI UNI R_Forward_User_Req R_Forward_User_Req (If M.CI=0) R_Backward_Net_Ack R_Forward_User_Ack (If M.CI=0) W_Ingress W_Egress W_Source Figure 11: Timer associated with a Rigid Forward BCR negotiation TR 101 123 V1.1.1 (1997-11) 27 Source Destination Ingress Node Egress Node UNI UNI R_Backward_User_Req R_Forward_User_Ack (If M.CI = 1) R_Backward_User_Req (If M.CI = 0) R_Forward_User_Ack (If M.CI = 0) Beginning of the ATM Block at the UNI W_Egress W_Destination Figure 12: Timer associated with a Rigid Backward BCR negotiation |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.4.1.2 Recovering from ingress Time-Out | Let us consider Rigid BCR negotiations processed at network ingress. The regular case (i.e. no RM cell loss and no collision with a BCR negotiation by the destination) is depicted in figure 11. If one of the relevant messages does not arrive at network ingress within a time interval at most W_Ingress long, the network retries the transaction. Specifically, an N_Forward_Req message with an appropriate Sequence Number is sent into the network with the BCR ΛCurrent(0+1) and ΛCurrent(OAM) currently allocated at network ingress if no BCR values are reserved or Λ'Current(0+1) and Λ’Current(OAM) if some BCR values have been reserved; the Elastic bit is set equal to that of the transaction which is pending. Timer W_Ingress is reinitialised after the transmission of this message. Such a message has priority over any BCR negotiation messages sent by the destination or Traffic Control message coming from the network egress. Upon reception of this message in a network node, a possible BCR for the CLP=0+1 cell flow is computed and reserved according to the value of the Elastic bit as for any other transaction; the elastic bit is handled accordingly. At network egress, the N_Forward_Req message is forwarded to the destination for information and a N_Backward_Ack message is sent back to network ingress; timer W_Egress is deactivated, if necessary, and (re)initialized. Upon reception of this N_Backward_Ack message at network ingress, timer W_Ingress is deactivated and the message is forwarded to the source to indicate that some BCR negotiation message have been lost within the network and that the BCR values ΛAlloc(0+1) and ΛAlloc(OAM) are to be allocated within the network. The source should acknowledge the N_Backward_Ack message by sending back the N_Forkward_Ack message by itself. The BCR ΛAlloc(0+1) and ΛAlloc(OAM) are allocated to the connection upon reception of this message . Timer W_Egress is deactivated upon reception of one of the above messages, which is forwarded to the destination. The actions taken by the network when the timer W_Ingress expires are shown in figure 13. TR 101 123 V1.1.1 (1997-11) 28 Source Destination Ingress Node Egress Node UNI UNI R_Forward_User_Req R_Backward_Net_Ack W_Ingress W_Egress W_Source N_Forward_Req N_Backward_Ack N_Forward_Ack Figure 13: Recovering from W_Ingress Time-Out |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.4.1.3 Recovering from Egress Time-Out | Besides timer W_Ingress, it may happen that timer W_Egress expires. In the case of a forward BCR negotiation this timer is activated when the R_Backward_Net_Ack (or E_Backward_Net_Ack) message is sent back to the source, as shown in figure 11. Timer W_Egress is activated also in the case of a backward BCR negotiation when the R_Backward_User_Req (or E_Backward_User_Req) message is sent into the network, as shown in figure 12. In all these cases, the appropriate Acknowledgement message Bandwidth Acknowledgement RM cell is expected at network egress within a given time interval. If this message is not received before timer W_Egress expire, special actions intended to retry the incomplete transaction are taken at network egress. Specifically, an N_Backward_Req message with an appropriate Sequence Number is sent on the backward direction into the network. The BCR values conveyed by this message are the values λCurrent(0+1) and λCurrent(OAM) currently allocated to the CLP=0+1 and OAM cell flow at network egress if no BCR values have been reserved or λ'Current(0+1) and λ'Current(OAM) if some BCR values have been reserved; the elastic bit in this message is set equal to that of the transaction, which is pending. TR 101 123 V1.1.1 (1997-11) 29 Source Destination Ingress Node Egress Node UNI UNI R_Forward_User_Req R_Backward_Net_Ack W_Source R_Forward_User_Ack N_Backward_Req W_Egress N_Forward_Ack Figure 14: Recovering from W_Egress Time-Out Upon reception of this message in a network node, a possible BCR for the CLP=0+1 cell flow is computed by taking into account the Elastic bit as for any other transaction and the Elastic bit is set accordingly. At network ingress, upon reception of the N_Backward_Req message, the message is forwarded to the source and no other actions are taken by the ingress node. The source should acknowledge the N_Backward_Req message by itself by sending a N_Forward_Ack message, which is then forwarded to the network egress. Upon reception of the N_Forward_Ack message in a network node, the BCR ΛAlloc(0+1) is updated to the value conveyed in the message. The traffic management process in the case of expiration of timer W_Egress is shown in figure 14 for the case in which timer W_Egress has been triggered by the transmission of a R_Backward_Net_Ack message (rigid forward BCR negotiation procedure). The other cases of expiration of timer W_Egress are dealt with in the same way. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.6.4.1.4 Further considerations | The above procedures related the expiration of timers W_Ingress and W_Egress may be reiterated several times. Beyond a certain number of iterations, the network may declare the connection as to be non compliant and release the connection. The definition of non compliant connection is network operator specific. If the source ( the destination) has not received any response from the network within a time interval at most W_Source ( W_Destination) time unit long after having initiated a BCR negotiation, the source ( the destination) might reattempt the BCR negotiation. TR 101 123 V1.1.1 (1997-11) 30 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7 Behaviour in the case of collision of different procedures | Collisions may happen at ingress node, egress node or within the network; the procedures followed in these three cases are different. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7.1 Collision at the ingress node | Collision at the ingress node occurs when the source tries to initiate a new forward BCR negotiation (or a forward Status Enquiry procedure) when another one is still in progress within the network or when the network has initiated a Forward Traffic Control procedure. In both the cases the new negotiation request is denied and a M_Backward_Notready_Ack message is sent to the source, as shown in figure 15. Source Destination Ingress Node Egress Node UNI UNI R_Forward_User_Req or E_Forward_User_Req or M_Forward_User_Req M_Backward_Notready_Ack Figure 15: Collision at the ingress node |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7.2 Collision at the egress node | Collision at the egress node occurs when the destination tries to initiate a new backward BCR negotiation (or a backward Status Enquiry procedure) when another one is still in progress within the network or when the network has initiated a Backward Traffic Control procedure. In both cases the new negotiation request is denied and a M_Forward_Notready_Ack message is sent to the source, as shown in figure 16. Source Destination Ingress Node Egress Node UNI UNI R_Backward_User_Req or E_Backward_User_Req or M_Backward_User_Req M_Forward_Notready_Ack Figure 16: Collision at the Egress node |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7.3 Collision within the network | Since BCR negotiations in a given direction may be initiated by either user, two BCR negotiations may collide within the network. Specifically, two BCR negotiations collide if the Forward BCR Negotiation Request message issued by the TR 101 123 V1.1.1 (1997-11) 31 source arrives at a network node, which has received a Backward BCR Negotiation Request message from the destination and not the corresponding Acknowledgement message. To solve this problem, BCR negotiations initiated by the destination have priority over BCR negotiation initiated by the source (backward priority principle). Furthermore, Traffic Control messages (with M.Maintenance=0 and M.Trf=1) issued either at network egress or ingress in the case of potential congestion have priority over any BCR negotiation initiated by the source or the destination. Finally, those traffic control messages issued at network ingress have priority over those generated at network egress. Summarizing, the following priority order applies (decreasing priority): 1) forward Traffic Control procedures; 2) backward Traffic Control procedures; 3) backward BCR negotiation procedures; 4) forward BCR negotiation procedures. |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7.3.1 Collision between forward and backward BCR negotiation | Let consider first the case in which a Forward BCR Negotiation collides with a Backward BCR Negotiation but no Traffic Control procedures are initiated by the network. In that case the Backward BCR Negotiation procedure takes priority over the Forward BCR Negotiation procedure and will be completed in the normal way (see subclauses 5.6.1.2.3 and 5.6.1.3.2) whereas the Forward BCR Negotiation will be aborted. Upon reception at a network node of a Backward Bandwidth Request message due to the destination, any BCR reservation by the source is cancelled. On the other hand any Forward Bandwidth Request message issued by the source and arriving at a network node while some resources have been reserved by a Backward Bandwidth Request message issued by the destination is ignored by the network node and the M.CI field is set, if necessary. No acknowledgement to this request is sent at network egress to the source. (In replacement of editor’s note).The only difference with the no collision case is that in a rigid backward BCR negotiation the R_Backward_User_Req is forwarded to the source even if M.CI=1. This is intended to alert the source that a collision happened in the network. The source has to acknowledge this message only if M.CI=0 or M.Elastic=0. The messages exchanged across the UNI are shown in figures 17 and 18 in case of Rigid and Elastic Backward BCR negotiations respectively. tttR Source Destination Ingress Node Egress Node UNI UNI R_Backward_User_Req R_Forward_User_Req or E_Forward_User_Req R_Backward_User_Req R_Forward_User_Ack if M.CI=1 R_Forward_User_Ack if M.CI=0 Beginning of the ATM Block at the UNI Figure 17: BCR negotiation initiated by the destination in the rigid mode with collision TR 101 123 V1.1.1 (1997-11) 32 Source Destination Ingress Node Egress Node UNI UNI E_Backward_User_Req R_Forward_User_Req or E_Forward_User_Req E_Backward_User_Req E_Forward_User_Ack Beginning of the ATM Block at the UNI Figure 18: BCR negotiation initiated by the destination in the elastic mode with collision |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7.3.2 Collision between backward BCR negotiation and Forward | Traffic Control procedures When a Backward BCR Negotiation collides with a Forward Traffic Control procedure the latter takes priority and is completed in the normal way (see subclauses 5.6.3.1.1 and 5.6.3.2.1) whereas the Backward BCR Negotiation is aborted. Upon reception at a network node of a Forward Traffic Control Request message, any BCR reservation made by the destination is cancelled. On the other hand any Backward Bandwidth Request message issued by the destination and arriving at a network node while some resources have been reserved by a Forward Bandwidth Request message is ignored by the network node and the M.CI field is set, if necessary. No acknowledgement to this request is sent to the destination. The messages exchanged across the UNI are shown in figures 19 and 20 in case of Rigid and Elastic Forward Traffic Control, respectively. TR 101 123 V1.1.1 (1997-11) 33 Source Destination Ingress Node Egress Node UNI UNI R_Backward_User_Req or E_Backward_User_Req Beginning of the ATM Block at the UNI R_Traffic_Forward_Req R_Traffic_Backward_Ack R_Traffic_Forward_Ack Figure 19: Forward Traffic Control procedure in the rigid mode with collision with a Backward BCR Negotiation procedure Source Destination Ingress Node Egress Node UNI UNI R_Backward_User_Req or E_Backward_User_Req Beginning of the ATM Block at the UNI E_Traffic_Forward_Req E_Traffic_Backward_Ack E_Traffic_Forward_Ack Figure 20: Forward Traffic Control procedure in the elastic mode with collision with a Backward BCR negotiation procedure |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7.3.3 Collision between forward BCR negotiation and backward | traffic control procedures When a Forward BCR Negotiation collides with a Backward Traffic Control procedure the latter takes priority and is completed in the normal way (see subclauses 5.6.3.1.2 and 5.6.3.2.2) whereas the Backward BCR Negotiation is aborted. Upon reception at a network node of a Backward Traffic Control Request message, any BCR reservation made by the source is cancelled. On the other hand any Forward Bandwidth Request message issued by the source and arriving at a network node while some resources have been reserved by a Backward Bandwidth Request message is ignored by the network node and the M.CI field is set, if necessary. No acknowledgement to this request is sent to the source. The messages exchanged across the UNI are shown in figures 21 and 22 in the case of Rigid and Elastic Backward Traffic Control, respectively. TR 101 123 V1.1.1 (1997-11) 34 Source Destination Ingress Node Egress Node UNI UNI R_Forward_User_Req or E_Forward_User_Req Beginning of the ATM Block at the UNI R_Traffic_Backward_Req R_Traffic_Forward_Ack Figure 21: Backward Traffic Control procedure in the rigid mode with collision with a Forward BCR negotiation procedure Source Destination Ingress Node Egress Node UNI UNI R_Forward_User_Req or E_Forward_User_Req Beginning of the ATM Block at the UNI E_Traffic_Backward_Req E_Traffic_Forward_Ack Figure 22: Backward Traffic Control procedure in the elastic mode with collision with a Forward BCR negotiation procedure TR 101 123 V1.1.1 (1997-11) 35 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 5.7.3.4 Collision between forward and backward traffic control | procedures (In replacement of editor’s note). When a Forward Traffic Control collides with a Backward Traffic Control procedure the former takes priority and is completed in the normal way (see subclauses 5.6.3.1.1 and 5.6.3.2.1) whereas the Backward Traffic Control is aborted. Upon reception at a network node of a Forward Traffic Control Request message, any BCR reservation made by the Backward Traffic Control is cancelled. On the other hand any Backward Traffic Control message arriving at a network node while some resources have been reserved by a Forward Traffic Control message is ignored by the network node and the M.CI field is set, if necessary. No acknowledgement to this request is sent to the source. The messages exchanged across the UNI are shown in figures 23 and 24 in the case of Rigid and Elastic Forward Traffic Control, respectively. Source Destination Ingress Node Egress Node UNI UNI R_Traffic_Backward_Ack R_Traffic_Forward_Req Beginning of the ATM Block at the UNI R/E_Traffic_Backward_Req R_Traffic_Forward_Ack Figure 23: Forward Traffic Control procedure in the rigid mode with collision with a Backward Traffic Control negotiation procedure Source Destination Ingress Node Egress Node UNI UNI E_Traffic_Backward_Ack E_Traffic_Forward_Req Beginning of the ATM Block at the UNI R/E_Traffic_Backward_Req E_Traffic_Forward_Ack Figure 24: Forward Traffic Control procedure in the elastic mode with collision with a Backward Traffic Control negotiation procedure TR 101 123 V1.1.1 (1997-11) 36 |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 6 Cases of possible applications | |
82c9666b291bb991c5a3eb56cd7cb09f | 101 123 | 6.1 Support of A-VPN | A possible application of the ABT-DT protocol described in the previous subclauses is the support of Broadband Virtual Private Networks (B-VPN), specially in the context of a semi-permanent ATM environment. In the following this application will be addressed by the term "A-VPN". In this application, the idea of disassociating connectivity from bandwidth allocation is still a basic concept: the customer leases from the public network a mesh of DBR VP "links", which topologically connects the various sites of its private network across the public one; in addiction, a full mesh of VCCs connects each pair of locations belonging to the AVPN, but no resources are associated to them, the VCCs being only a logical concatenation of VP links. The total bandwidth of the VP link is shared by the VCCs, and it is even possible for a single VCC to consume the total bandwidth of one VP link. The concept of the A-VPN is captured by figure 25. If no controls are applied to the VCCs overload on a "link" might arise every time two or more VCCs which share some "link", currently require a greater bandwidth than that of the VPC "link". Moreover, as inside the ATM nodes the A- VPN traffic, in general, shares resources with the traffic of other customers, the degradation of performance due to overload also affects the QoS of connections not belonging to the A-VPN responsible for it. In the above scenario overload occurs due to "structural" unawareness of the ATM network about the behaviour of the A-VPN users and the current load of the VCCs. VCCs cannot be policed by ATM, as the only traffic contract between ATM operator and A-VPN user refers to VPC "links", while the VCCs are completely under the user control only; any VCC can, in principle, reach the total capacity of the smaller VPC "link" connecting one of the end point of the VCC to the first ATM node; so ATM network results practically unprotected. The above overload problems can efficiently be solved , by the utilization of ABT-DT VCCs: the ABT-DT protocol, indeed, will allow the interworking between the "private" network management, responsible for the operation of the A- VPN, and the public network management, which should guarantee network performance objective both for the private A-VPN and for the rest of the services contemporarily supported over the public ATM network, avoiding undue interference between the traffic of one service and the others. The ABT-DT protocol, in fact, provides a way to control the actual allocation of resources to the various VCCs, avoiding inconsistency which can overload some of the VP links and some of the nodes. Before an ABT-DT VCC can use a certain amount of resources, it should ask the public ATM network for its availability in the VP links involved. The bandwidth of the VCC can only be enlarged if the answer is positive. Resources remain allocated to a VCC until it decides to release them, for use by other VCCs belonging to the AVPN. By means of the ABT-DT protocol the network keeps track of the current usage of the resources of the VP links, and is allowed to deny VCC usage if some of the VP links become overloaded. Furthermore, the network is aware of the amount of bandwidth allocated to each VCC and can enforce it at the network ingress, provided that the parameters of the UPC mechanism can be dynamically changed according to the results of the bandwidth negotiation procedures. In this approach, resources of the VP links are dynamically allocated and released by its VCC, the requests being driven by the traffic fluctuations at the VCC level. TR 101 123 V1.1.1 (1997-11) 37 Figure 25: A-VPN concept It is the user’s responsibility, where the word "user" refers to the management of the AVPN, to correctly dimension his private network in terms of number, topology and bandwidth of the VP links, and to define the routeing of the VCC among the different CNPs through the AVPN links. The public network should be provided with units able to run the ABT-DT protocol. It is to be noted that, since each AVPN is supported over dedicated VPCs, the segregation of the AVPN traffic only requires that the resources allocated to these VPCs are not exceeded and this result can be achieved with the basic BCR negotiation procedures provided by the ABT-DT protocol; therefore, in order to support AVPN, a simplified version of the ABT-DT protocol could be sufficient. In particular, since the competition to access the available resources is limited to the VCCs belonging to the same customer, the VCCs do not need to have any guaranteed QoS at the block level so that the Traffic Control procedures as well as the Network Maintenance procedures foreseen by the protocol do not need to be implemented. TR 101 123 V1.1.1 (1997-11) 38 Annex A: Reference behaviours A.1 Source reference behaviour Any BCR modification by the source is initiated by sending a Bandwidth Request RM cell on the forward direction with DIR=0 and the requested BCR. The BCR request RM cells (forward or backward requests) generated by the source should be conforming to GCRA(TRM,τRM); any non conforming RM cell may be ignored by the network. Moreover, the source should not initiate a BCR modification in the forward direction while another one is still in progress within the network (i.e., the source has not received any response from the network); otherwise, the source receives a specific maintenance message (Not_Ready message) from the network. For a given allocated BCR value λ for the CLP=0+1 cell flow, the source should deliver a cell flow conforming to the allocated BCR λ and the corresponding CDV tolerance τ(λ). The OAM cell flow should be conforming to the allocated BCR value and the associated CDV tolerance satisfying the general requirements given in Appendix II to ETS 300 301. The source may receive: 1) a Bandwidth Acknowledgement RM cell with Traffic Management=0 corresponding to a BCR request initiated by the source. If CI=0 or Elastic=0, the source should acknowledge this message by sending in the forward direction a Bandwidth Acknowledgement RM cell with BCR values less than or equal to those communicated by the network; 2) a Bandwidth Request RM cell with Traffic Management=0/1, indicating a BCR negotiation initiated by the destination, a traffic control procedure at network egress, or the loss of some RM cells. This message may be received even if the source has not initiated any BCR negotiation. If Traffic Management=1 or (Traffic Management=0 and (CI=0 or Elastic=0)), this message should be acknowledged by sending a Bandwidth Acknowledgement RM cell with BCR values as above; 3) a Bandwidth Acknowledgement RM cell with Traffic Management=1 indicating a traffic control procedure at network ingress or the loss of some RM cells. This message may be received even if the source has not initiated any BCR negotiation. The source should acknowledge this message by sending in the forward direction a Bandwidth Acknowledgement RM cell with correct BCR values. The state of the source is characterized by two values IDLE and HUNT. The source is in state IDLE when no BCR negotiation is in progress. The source is in state HUNT when a BCR negotiation has been initiated by the source. The source should get a response from the network within a time interval at most W_Source time unit long. The pseudo code of the source behaviour is given in table A.1. The source process is related to the ingress process via the DoS (downstream) and UpS (Upstream) routes. The following notation is used in the SDL diagrams: - T stands for true and F for false in Boolean expressions; - binary flags are coded by Boolean instead of 0 and 1. Specifically 1 is coded by T except for the Elastic bit where it is coded by F. Messages are represented by two SDL signals: 1) FMsg(kind,bandwidth, ci, el, mt, trf, seq) for messages with the Dir bit set to 0; 2) BMsg(kind,bandwidth, ci, el, mt, trf, seq) for messages with the Dir bit set to 1. where: - kind: message type (Ack or Req); - rate : vector of BCR values for the CLP=0+1 and OAM cell flows; - ci : congestion indication (Boolean); - el : elasticity indication (Boolean); - mt: maintenance indication (Boolean); - trf: traffic management indication (Boolean); TR 101 123 V1.1.1 (1997-11) 39 - seq: sequence number (integer). The BCR currently allocated at the source to the different cell flows of the connection are denoted by ΛSource(0+1) and ΛSource(OAM). The BCR acknowledged by the source should be less than or equal to the BCR communicated by the network and are computed by the procedure adapt(M) as a function of the message received from the network. Moreover, the following procedures are defined: - the check_number(M) procedure checks whether a traffic message M has a correct sequence number with respect to the value of the seq_number variable; the result is either true or false; - the change_number(M) procedure updates the sequence number by incrementing its value by 1 modulo 2 32. Table A.1: Pseudo code of the source behaviour TR 101 123 V1.1.1 (1997-11) 40 PROCESS Source; SYNONYM TSource duration = EXTERNAL; DCL ΛSource,Λ Bandwidth; DCL seq,sn Integer; DCL kind Kind; DCL ci,ci0,el,el0,mt0,mt,tf Boolean; TIMER WSource; PROCEDURE check_number REFERENCED; PROCEDURE Change_number REFERENCED; PROCEDURE Adapt REFERENCED; START ; TASK sn:=0; NEXTSTATE Idle; STATE Idle; INPUT Change_BCR(Λ); COMMENT 'Rigid BCR negotiation'; TASK el := F, mt := F; Modify: CALL Change_number; OUTPUT FMsg(Req,Λ,F,el,mt,F,sn) VIA UpS; DECISION Λ <= ΛSource; (T): CALL adapt; NEXTSTATE Idle; (F): SET(NOW+TSource,WSource); NEXTSTATE Hunt; ENDDECISION; INPUT Change_BCR_EL(Λ); COMMENT 'Elastic BCR negotiation'; TASK el := T, mt := F; JOIN Modify; INPUT Status_Enquiry; COMMENT 'status enquiry'; TASK Λ := Null, el := T, mt := T; CALL Change_number; OUTPUT FMsg(Req,Λ,F,el,mt,F,sn) VIA UpS; SET(NOW+TSource,WSource); NEXTSTATE Hunt; INPUT BMsg(kind,Λ,ci,el0,mt0,tf,seq); DECISION kind; (Req): JOIN SrcReq; (Ack): DECISION tf; (F): (T): JOIN SrcTrf; ENDDECISION; ENDDECISION; NEXTSTATE Idle; STATE Hunt; SAVE Change_BCR; SAVE Change_BCR_EL; SAVE Status_Enquiry; INPUT BMsg(kind,Λ,ci,el0,mt0,tf,seq); DECISION kind; (Req): RESET(WSource); SrcReq: COMMENT 'Request from the egress or the destination'; DECISION tf; (F): COMMENT 'User BCR negotiation'; DECISION ci and not(el0); (T): COMMENT 'that failed'; (F): COMMENT 'that succeeded'; CALL Adapt; OUTPUT FMsg(Ack,Λ,ci,el0,mt0,tf,seq) VIA UpS; ENDDECISION; (T): JOIN SrcTrf; ENDDECISION; (Ack): COMMENT 'Treatment of acknowledgements'; DECISION tf; (F): DECISION mt0 and ci; (T): NEXTSTATE -; (F): COMMENT 'Checks the sequence number before reseting the timer'; CALL check_number(ci0); DECISION ci0; (T): RESET(WSource); (F): NEXTSTATE -; ENDDECISION; COMMENT 'In any case, adapt to its contents'; DECISION not(ci) or el0; (T): CALL Adapt; OUTPUT FMsg(Ack,Λ,ci,el0,mt0,tf,seq) VIA UpS; (F): ENDDECISION; ENDDECISION; (T): RESET(WSource); SrcTrf: COMMENT 'Traffic messages (Req or Ack) are accepted as is'; CALL Adapt; OUTPUT FMsg(Ack,Λ,ci,el0,mt0,tf,seq) VIA UpS; ENDDECISION; ENDDECISION; NEXTSTATE Idle; INPUT WSource; COMMENT 'Retry on failure'; CALL Change_number; SET(NOW+TSource,WSource); OUTPUT FMsg(Req,Λ,F,el,mt,F,sn) VIA UpS; NEXTSTATE -; ENDPROCESSSource TR 101 123 V1.1.1 (1997-11) 41 A.2 Destination reference behaviour The destination may receive messages from the network due to BCR negotiations initiated by the source or the network (traffic control procedures). Moreover, the destination may initiate a BCR negotiation applicable to the traffic generated by the source. For this purpose, the destination should send a Bandwidth Request RM cell with DIR=1 and the requested BCR values. In response to this request, the destination may receive: 1) a Bandwidth Acknowledgement RM cell with Traffic Management=0, in response to the BCR request initiated by the destination; 2) a Bandwidth Request RM cell with Traffic Management=1 and Maintenance=1 due to a traffic control procedure at network ingress (RM cell loss); 3) a Bandwidth Acknowledgement RM cell with Traffic Management=1 and Maintenance=0 due to a traffic control procedure at network egress; or 4) a Bandwidth Acknowledgement RM cell with Traffic Management=1 and Maintenance=1 due to the loss of some RM cells. The destination has to accept a BCR negotiation. When the destination submits BCR request RM cells (forward or backward requests) to the network, they have to be conforming to the GCRA(T'RM,τ'RM); any non conforming RM cell may be ignored by the network. The pseudo code of the destination is given in table A.2. The destination process is related to the network via the DoS (downstream) and UpS (upstream) routes. The following procedures are defined: - the check_number(M) procedure checks whether a traffic message M has a correct sequence number with respect to the value of the seq_number variable; the result is either true or false; - the change_number(M) procedure updates the sequence number by incrementing its value by 1 modulo 2 32. Table A.2: Pseudo code of the destination behaviour TR 101 123 V1.1.1 (1997-11) 42 PROCESS Destination; DCL Λ,ΛDestination Bandwidth; DCL sn,seq Integer; DCL kind Kind; DCL ci,ci0,el,el0,mt,mt0,tf Boolean; DCL TDest duration; TIMER WDest; PROCEDURE check_number REFERENCED; PROCEDURE Change_number REFERENCED; START ; TASK sn := 1; TASK TDest := Tusr; NEXTSTATE Idle; STATE Idle; INPUT Change_BCR(Λ); COMMENT 'Rigid BCR negotiation'; TASK el := F, mt := F; CALL Change_number; OUTPUT BMsg(Req,Λ,F,el,mt,F,sn) VIA DoS; SET(NOW+TDest,WDest); NEXTSTATE Hunt; INPUT Change_BCR_EL(Λ); COMMENT 'Elastic BCR negotiation'; TASK el := T, mt := F; CALL Change_number; OUTPUT BMsg(Req,Λ,F,el,mt,F,sn) VIA DoS; SET(NOW+TDest,WDest); NEXTSTATE Hunt; INPUT Status_Enquiry; COMMENT 'status enquiry'; TASK Λ := Null, el := T, mt := T; CALL Change_number; OUTPUT BMsg(Req,Λ,F,el,mt,F,sn) VIA DoS; SET(NOW+TDest,WDest); NEXTSTATE Hunt; INPUT FMsg(kind,Λ,ci,el0,mt0,tf,seq); DECISION kind; (Req): DestReqForward: DECISION not(el0) and Λ <= ΛDestination and not(tf) and not(mt0); (T): COMMENT 'BCR decrease'; TASK ΛDestination := Λ; NEXTSTATE Idle; (F): COMMENT 'other request: wait for acknowledgement'; SET(NOW+TDest,WDest); NEXTSTATE Waiting; ENDDECISION; (Ack): DECISION tf; (T): COMMENT 'Only kind of Ack that may arrive (traffic from egress)'; TASK ΛDestination := Λ; (F): ENDDECISION; ENDDECISION; NEXTSTATE -; STATE Waiting; SAVE Change_BCR; SAVE Change_BCR_el; SAVE Status_Enquiry; INPUT FMsg(kind,Λ,ci,el0,mt0,tf,seq); COMMENT 'Blocked until an Ack arrives'; DECISION kind; (Req): NEXTSTATE -; (Ack): RESET(WDest); TASK ΛDestination := Λ; NEXTSTATE Idle; ENDDECISION; INPUT WDest; COMMENT 'Timeout - back to normal state'; NEXTSTATE Idle; STATE Hunt; SAVE Change_BCR; SAVE Change_BCR_el; SAVE Status_Enquiry; INPUT WDest; COMMENT 'Timeout - retries the negotiation'; CALL Change_number; SET(NOW+TDest,WDest); OUTPUT BMsg(Req,Λ,F,el,mt,F,sn) VIA DoS; NEXTSTATE -; INPUT FMsg(kind,Λ,ci,el0,mt0,tf,seq); DECISION kind; (Req): RESET(WDest); JOINT DestReqForward; (Ack): DECISION mt0 and ci; (T): COMMENT 'Not ready'; (F): COMMENT 'Updates according to the Acknowledge'; TASK ΛDestination := Λ; CALL check_number(ci0); DECISION ci0 or tf; (T): RESET(WDest); NEXTSTATE Idle; (F): ENDDECISION; ENDDECISION; NEXTSTATE -; ENDDECISION; ENDPROCESSDestination TR 101 123 V1.1.1 (1997-11) 43 A.3 Network element reference behaviour The ingress and egress network nodes should perform some UPC/NPC functions. For instance, they may control RM cell flows by discarding other RM cells than ABT/DT RM cells and by enforcing the peak cell rate of any ABT/DT bandwidth request RM cell flow, by means of the GCRA(TRM,τRM) and GCRA(T'RM,τ'RM). They should also process the conforming ABT/DT cells to run the ABT/DT procedure. Specifically, special functions located in the ingress and egress nodes should be able to guarantee the correct exchange of ABT/DT cells. For this purpose, such a function should be able to handle ABT/DT RM cells propagating on the forward as well as backward directions. A.3.1 Ingress node An ingress node is modelled by a four state process: - NORMAL: there is no BCR negotiation in progress; - TRAFFIC: a traffic control procedure has been initiated and is still in progress; - ACKNOWLEDGING: an RM cell has been sent to the source and the process is waiting for a response. WAITING: some resource are booked and the process is waiting for an acknowledgement RM cell. A node maintain two couples of BCR values, namely: - the BCR Λ currently allocated to the cell flows at network ingress; - the BCR Λ’ reserved for the cell flows at network ingress. The book(M,U) procedure updates the reserved and allocated BCR values maintained in an ABT/DT unit. For instance the book(M,U) procedure at network ingress can be coded as: PROCEDURE book; FPAR M,U : bandwidth; START TASK Λ (0+1) := M(0+1); TASK Λ(OAM) := min(M(0+1),M(OAM)); TASK Λ’(0+1) := U(0+1); TASK Λ’(OAM) := min(U(0+1),U(OAM)); RETUN ENDPROCEDURE To handle traffic control message, an implicit variable seq_number is defined. It stores the sequence number of the last relevant traffic control message. With regard to the sequence number, the following procedures are defined: - the check_number(M) procedure checks whether a traffic message M has a correct sequence number with respect to the value of the seq_number variable; the result is either true or false; - the change_number(M) procedure updates the sequence number by incrementing its value by 1 modulo 232; - the attribute_msg(M) procedure assigns the value of the seq_number to the sequence number of message M. The ingress ABT/DT process is linked to the source process by the UpS (upstream) route and to the first intermediate process by the DoS (downstream) route. The pseudo code of the ingress node is given in table A.3. Table A.3: Pseudo code of the ingress node TR 101 123 V1.1.1 (1997-11) 44 PROCESS Ingress; SYNONYM lmax Bandwidth = EXTERNAL; SYNONYM lpcr Bandwidth = EXTERNAL; SYNONYM Tingress Duration = EXTERNAL; DCL Λ,Λ’,rate Bandwidth; DCL seq,seq_counter,ignored_seq Integer; DCL kind Kind; DCL ci,el,mt,tf,auxiliary,el_memory Boolean; TIMER Wingress; PROCEDURE Book REFERENCED; PROCEDURE change_number REFERENCED; PROCEDURE attribute_msg REFERENCED; PROCEDURE check_number REFERENCED; START; TASK Λ := Null, Λ’ := Null; NEXTSTATE Normal; STATE Normal; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req) : DECISION tf; (T): COMMENT 'Traffic request'; COMMENT 'On a UNI node : error from the source'; TASK tf := F; JOIN IngressNormalForward; (F): COMMENT 'User request'; IngressNormalForward: TASK rate := min(lmax,rate); COMMENT 'Status enquiry or BCR increase'; DECISION (mt or rate > Λ); (T): DECISION mt; (T): TASK rate := Λ; (F): ENDDECISION; /* mt */ CALL Book(Λ,rate); TASK el_memory := el; SET (NOW+Tingress, Wingress); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Waiting; (F): COMMENT 'BCR decreased (not acknowledged)'; TASK el := F; CALL Book(rate,Null); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; ENDDECISION; /* mt or rate \(>\) Λ */ ENDDECISION; /* tf */ (Ack) : COMMENT 'No ACK in this state'; NEXTSTATE Normal; ENDDECISION; /* kind */ INPUT BMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req) : IngressBackward: DECISION tf; (T): COMMENT 'Traffic message from the egress'; IngressTrafficBackward: CALL Book(Λ,rate); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Acknowledging; (F): COMMENT 'BCR negotiation from the destination'; DECISION el or not(ci); (T): COMMENT 'successful'; CALL Book(Λ,rate); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; (F): COMMENT 'failed'; CALL Book(Λ,Null); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; COMMENT 'On the UNI, it is looped back to the destination'; TASK rate := Λ; TASK kind := Ack; OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; ENDDECISION; /* el or mt */ ENDDECISION; /* tf */ NEXTSTATE Acknowledging; (Ack): NEXTSTATE -; ENDDECISION; /* kind */ INPUT POL(el); COMMENT 'Policing action from the ingress'; IngressPolicing: DECISION el; (T): TASK rate := lpcr; (F): TASK rate := Λ; ENDDECISION; /* el */ TASK el_memory:=el; SET(NOW+Tingress,Wingress); TASK kind := Req, mt := F, tf := T, ci := F; CALL Book(Λ,rate); CALL change_number; CALL attribute_msg; OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Traffic; STATE Acknowledging; SAVE POL; INPUT FMsg(kind,rate,ci,el,mt,tf,ignored_seq); DECISION kind; (Req) : NotReadyIngress: COMMENT 'A Not Ready is sent'; OUTPUT BMsg(Ack,rate,T,T,T,F,0) VIA DoS; NEXTSTATE -; (Ack) : COMMENT 'The expected acknowledgement from the source'; TASK rate := min(rate,Λ’); CALL Book(rate,Null); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; ENDDECISION; /* kind */ INPUT BMsg(kind,rate,ci,el,mt,tf,seq); COMMENT 'Only traffic messages from the egress are considered'; TR 101 123 V1.1.1 (1997-11) 45 DECISION kind; (Req) : DECISION tf; (T): JOIN IngressTrafficBackward; (F): ENDDECISION; (Ack) : ENDDECISION; NEXTSTATE -; STATE Waiting; INPUT POL(el); COMMENT 'Same as in state Normal'; JOIN IngressPolicing; INPUT Wingress; COMMENT 'The timer has expired, an error recovery message is sent'; TimerIngress: SET(NOW+Tingress,Wingress); CALL change_number; CALL attribute_msg; OUTPUT FMsg(Req,Λ’,F,el_memory,T,T,seq) VIA UpS; NEXTSTATE Traffic; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); JOIN NotReadyIngress; INPUT BMsg(kind,rate,ci,el,mt,tf,seq); RESET(Wingress); DECISION kind; (Req) : COMMENT 'Requests from the destination or egress have priority over requests from the source'; DECISION tf; (T): JOIN IngressTrafficBackward; (F): DECISION el or not(ci); (T): CALL Book(Λ,rate); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; (F): COMMENT 'On failure always wait for an acknow- ledgement from the source to show that its negotiation has failed'; CALL Book(Λ,Λ); OUTPUT BMsg(kind,Λ,ci,el,mt,tf,seq) VIA DoS; ENDDECISION; NEXTSTATE Acknowledging; ENDDECISION; (Ack) : COMMENT 'Acknowledgement to the source BCR modification'; DECISION not(ci) or el; (T): COMMENT 'Successful negotiation'; CALL Book(Λ,rate); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Acknowledging; (F): COMMENT 'Unsuccessful negotiation (No ack expected)'; CALL Book(Λ,Null); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Normal; ENDDECISION; /* not(ci) or el */ ENDDECISION; /* kind */ STATE Traffic; SAVE Pol; INPUT Wingress; JOIN TimerIngress; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); JOIN NotReadyIngress; INPUT BMsg(kind,rate,ci,el,mt,tf,seq); DECISION tf and (kind=Ack); (T): COMMENT 'Acknowledgement to a traffic message'; CALL check_number(auxiliary); DECISION auxiliary; (T): RESET(Wingress); DECISION ci or (el and not(mt)); (T): CALL Book(Λ,rate); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Acknowledging; (F): COMMENT 'if no modification, directly answer without acknowledgement from the source'; CALL Book(Λ,Null); TASK rate := Λ; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; ENDDECISION; /* ci or (el and not(mt)) */ (F): COMMENT 'old message with wrong sequence number'; ENDDECISION; /* auxiliary */ (F): COMMENT 'Other messages are ignored'; ENDDECISION; /* tf and (kind=Ack) */ NEXTSTATE -; ENDPROCESSIngress; TR 101 123 V1.1.1 (1997-11) 46 A.3.2 Intermediate node An intermediate node process has three states : - NORMAL: no BCR negotiation is in progress; - WAITING: a user negotiation has been initiated and is still in progress; - TRAFFIC: a traffic control procedure has been initiated and is still in progress. The possible and fairshare procedures that control allocation respectively in the rigid and elastic case are network operator specific. As for the ingress process, two variables Λ and Λ’ are used. Note that the book procedure should also modify the available bandwidth ∆ in a network node. Moreover, information on OAM traffic is irrelevant. As a consequence the book procedure can be coded as : PROCEDURE book; FPAR M,U : bandwidth; START; TASK Λ := M(0+1); TASK Λ’ := U(0+1); TASK ∆ := ∆ + max (Λ(0+1),Λ’(0+1)) - max(M(0+1),U(0+1)); RETURN; ENDPROCEDURE; As an example, if the BCR negotiation context is not taken into account in the allocation of network resources, the possible procedure may be defined as follows: PROCEDURE possible; FPAR IN λ bandwidth; IN mt Boolean; IN/OUT λr bandwidth; START; TASK λr := min(λ,∆+max(λ,Λ’)); RETURN; ENDPROCEDURE; Under the same assumptions and in the case of unweighted fairness, the fairshare procedure may be defined as : PROCEDURE fairshare; FPAR λ bandwidth; IN/OUT λr bandwidth; IN/OUT ci bandwidth; START; TASK λr := min(C/N, ∆+Λ); TASK λr := min(λr,λ(0+1)); TASK ci := (λr < C/N); RETURN; ENDPROCEDURE; where N is the number of ABT/DT connexions and C is the total amount of bandwidth for ABT/DT connections. The pseudo code of an intermediate node is given in table A.4. Table A.4: Pseudo code of an intermediate node TR 101 123 V1.1.1 (1997-11) 47 PROCESS PInt; DCL Λ,Λ0,Λ’,avlb,rate Bandwidth; DCL seq Integer; DCL kind Kind; DCL ci,ci0,el,mt,tf Boolean; PROCEDURE Book REFERENCED; PROCEDURE possible REFERENCED; PROCEDURE Fairshare REFERENCED; PROCEDURE Forgetting REFERENCED; START; TASK Λ := Null, Λ’ := Null; NEXTSTATE Normal; STATE Normal; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): IntFwdTraffic: DECISION tf; (T): COMMENT 'Traffic request from the ingress'; DECISION el; (T): COMMENT 'with elastic parameters'; CALL Fairshare(rate,Λ0,ci0); TASK ci := ci or ci0, rate := Λ0; (F): COMMENT 'with rigid parameters'; CALL possible(rate,mt,Λ0); DECISION rate > Λ0; (T): TASK ci := T, rate := Λ0; (F): ENDDECISION; ENDDECISION; /* el */ CALL Book(Λ,rate); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Traffic; (F): COMMENT 'Request from the source'; DECISION mt; (T): COMMENT 'Status enquiry'; CALL possible(rate,T,Λ0); TASK rate := min(rate,Λ0); CALL Book(Λ,rate); (F): DECISION el; (T): COMMENT 'Elastic request'; CALL Fairshare(rate,Λ0,ci0); TASK ci := ci or ci0, rate := Λ0; CALL Book(Λ,rate); (F): DECISION ci; (F): DECISION (rate <= Λ); (T): COMMENT 'BCR decrease'; CALL Book(rate, Null); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; (F): COMMENT 'Rigid BCR increase ...'; CALL possible(rate,mt,Λ0); TASK ci := IF rate > Λ0 THEN T ELSE ci FI; DECISION ci; (F): COMMENT '... that succeeds'; CALL Book(Λ,rate); (T): COMMENT '... that fails'; ENDDECISION; ENDDECISION; /* rate ≤ Λ */ (T): COMMENT 'request that has already failed'; ENDDECISION; /* ci */ ENDDECISION; /* el */ ENDDECISION; /* mt */ COMMENT 'in any case, the message is forwarded'; OUTPUT FMsg(Req,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Waiting; ENDDECISION; /* tf */ (Ack): COMMENT 'There should be no acknowledgement in this state'; NEXTSTATE Normal; ENDDECISION; INPUT BMsg(kind,rate,ci,el,mt,tf,seq); IntBwd: DECISION kind; (Req): DECISION tf; (T): COMMENT 'Traffic request from the egress'; DECISION el; (T): COMMENT 'with elastic parameters'; CALL Fairshare(rate,Λ0,ci0); TASK ci := ci or ci0, rate := Λ0; (F): COMMENT 'with rigid parameters'; CALL possible(rate,mt,Λ0); DECISION rate > Λ0; (T): TASK ci := T, rate := Λ0; (F): ENDDECISION; ENDDECISION; CALL Book(Λ,rate); OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Traffic; (F): COMMENT 'Request from the destination'; DECISION mt; (T): COMMENT 'status enquiry'; CALL possible(rate,T,Λ0); TASK rate := min(rate,Λ0); CALL Book(Λ,rate); (F): DECISION el; (T): COMMENT 'elastic BCR modification'; CALL Fairshare(rate,Λ0,ci0); TASK ci := ci or ci0, rate := Λ0; CALL Book(Λ,rate); (F): DECISION ci; (F): COMMENT 'rigid BCR modification'; CALL possible(rate,mt,Λ0); TASK ci := IF rate > Λ0 THEN T ELSE ci FI; DECISION ci0; (F): TR 101 123 V1.1.1 (1997-11) 48 CALL Book(Λ,rate); (T): TASK ci := T; ENDDECISION; (T): COMMENT 'failed request'; ENDDECISION; ENDDECISION; ENDDECISION; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Waiting; ENDDECISION; (Ack): COMMENT 'No acknowledgement in this state'; NEXTSTATE Normal; ENDDECISION; STATE Waiting; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): DECISION tf; (T): COMMENT 'Traffic request have priority'; JOIN IntFwdTraffic; (F): COMMENT 'Request from the source are ignored (collision principle)'; TASK ci := T; OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Waiting; ENDDECISION; (Ack): COMMENT 'Acknowledgement from the source is taken into account'; CALL Book(rate, Null); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; ENDDECISION; INPUT BMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): JOIN IntBwd; (Ack): DECISION ci and not(el) and not(mt) and not(tf); (T): COMMENT 'rigid source BCR increase has failed, everything is cleared'; CALL Book(Λ, Null); OUTPUT BMsg(Ack,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Normal; (F): DECISION el; (T): COMMENT 'if the negotiation was elastic, just keeps what can actually be used'; CALL Book(Λ,rate); (F): COMMENT 'otherwise just pass it to the source'; ENDDECISION; ENDDECISION; OUTPUT BMsg(Ack,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Waiting; ENDDECISION; STATE Traffic; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): DECISION tf; (T): COMMENT 'Traffic message from the source have the highest precedence'; JOIN IntFwdTraffic; (F): COMMENT 'Other messages are ignored'; OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE -; ENDDECISION; (Ack): DECISION tf; (T): COMMENT 'The acknowledgement is taken into account'; CALL Book(rate,Null); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; (F): COMMENT 'Not the acknowledgement expected'; OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE -; ENDDECISION; ENDDECISION; INPUT BMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): DECISION tf and mt; (T): COMMENT 'error recovery from a traffic message emited by the egress : nothing is booked but the computation is done'; CALL possible(rate,mt,Λ0); DECISION rate > Λ0; (T): TASK ci := T, rate := Λ0; (F): ENDDECISION; (F): COMMENT 'other messages are ignored'; ENDDECISION; (Ack): DECISION tf; (T): COMMENT 'Acknowledgement for a traffic message initiated by the ingress. The node shifts to Waiting state so that error recovery can be performed without hack'; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Waiting; (F): ENDDECISION; ENDDECISION; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE -; ENDPROCESS PInt; TR 101 123 V1.1.1 (1997-11) 49 A.3.3 Egress node An egress node has three states as an intermediate node. The procedures used by an egress node are similar to the procedures used in an ingress node. The pseudo code on the egress node is given in table A.5. Table A.5: Pseudo code of the egress node TR 101 123 V1.1.1 (1997-11) 50 PROCESS Egress; SYNONYM lmax Bandwidth = EXTERNAL; SYNONYM lpcr Bandwidth = EXTERNAL; SYNONYM TEgress Duration = EXTERNAL; DCL Λ,Λ’,rate Bandwidth; DCL ign_seq,seq,seqcpt Integer; DCL kind Kind; DCL ci,el,mt,tf ,el_memory,aux Boolean; TIMER WEgress; PROCEDURE Book REFERENCED; PROCEDURE change_number REFERENCED; PROCEDURE attribute_msg REFERENCED; PROCEDURE check_number REFERENCED; PROCEDURE SETNumbering REFERENCED; START ; TASK Λ := Null, Λ’ = Null; NEXTSTATE Normal; STATE Normal; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): DECISION tf; (T): COMMENT 'Traffic request from the ingress'; EgressNetTrafficFwd; CALL Book(Λ,rate); COMMENT 'On a UNI loop it back'; CALL SETNumbering; TASK el_memory := el; SET(NOW+TEgress,WEgress); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; TASK kind := Ack; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Traffic; (F): DECISION el or (not(ci) and rate > Λ); (T): COMMENT 'Successful user BCR increase'; CALL Book(Λ,rate); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; COMMENT 'Loop back on the UNI'; TASK kind := Ack, el_memory := el; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; SET(NOW+TEgress,WEgress); NEXTSTATE Waiting; (F): DECISION rate <= Λ; (T): COMMENT 'User BCR decrease'; CALL Book(rate,Null); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; (F): COMMENT 'Unsuccessful BCR modification'; CALL Book(Λ,Null); COMMENT 'Loop back on the UNI'; CALL Book(Λ,Null); TASK kind := Ack; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Normal; ENDDECISION; /* rate ≤ Λ */ ENDDECISION; /* el or not(ci and rate \(>\) Λ) */ ENDDECISION; /* tf */ (Ack): COMMENT 'No ack in this state’; NEXTSTATE Normal; ENDDECISION; /* kind */ INPUT BMsg(kind,rate,ci,el,mt,tf,seq); EgressBackward: TASK rate := min(rate,lmax); DECISION kind; (Req): EgressBackwardReq: DECISION tf; (T): COMMENT 'Reset the tf bit on a user request'; TASK tf := F; JOIN EgressBackwardReq; (F): COMMENT 'User request initialization'; TASK ci:=F; DECISION mt; (T): COMMENT 'Initialize status enquiry'; TASK rate := Λ, el := T; (F): COMMENT 'Initialize BCR modification'; TASK el:= IF rate <= Λ THEN F ELSE el FI; ENDDECISION; CALL Book(Λ,rate); SET(NOW+TEgress,WEgress); TASK el_memory := el; OUTPUT BMsg(kind,rate,ci,el,mt,tf,seq) VIA DoS; NEXTSTATE Waiting; ENDDECISION; /* tf */ (Ack): ENDDECISION; /* kind */ NEXTSTATE -; INPUT POL(el); COMMENT 'Policing action from the egress'; EgressPolicing: TASK rate := IF el THEN lpcr ELSE Λ FI; TASK el_memory := el; SET(NOW+TEgress,WEgress); CALL Book(Λ,rate); CALL change_number; CALL attribute_msg; OUTPUT BMsg(kind,rate,ci,el,F,tf,seq) VIA DoS; NEXTSTATE Traffic; STATE Waiting; INPUT WEgress; COMMENT 'The TIMER has expired'; TIMEREgress: CALL change_number; CALL attribute_msg; SET(NOW+TEgress,WEgress); OUTPUT BMsg(Req,Λ’,F,el_memory,T,T,seq) VIA DoS; NEXTSTATE Traffic; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): COMMENT 'Only traffic requests are treated'; DECISION tf; (T): RESET(WEgress); JOIN EgressNetTrafficForward; (F): NEXTSTATE -; ENDDECISION; /* tf */ (Ack): TR 101 123 V1.1.1 (1997-11) 51 COMMENT 'Acknowledgement (New block) from the source'; CALL Book(rate,Null); RESET(WEgress); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; ENDDECISION; /* kind */ INPUT BMsg(kind,rate,ci,el,mt,tf,seq); COMMENT 'Only traffic requests from an upward network are treated otherwise sends a Not Ready'; OUTPUT FMsg(Ack,rate,T,T,T,F,seq) VIA UpS; NEXTSTATE -; INPUT POL(el); JOIN EgressPolicing; STATE Traffic; INPUT WEgress; JOIN TIMEREgress; INPUT FMsg(kind,rate,ci,el,mt,tf,seq); DECISION kind; (Req): JOIN EgressForwardRequestInTraffic; (Ack): COMMENT 'Treat Acks from the source with the right sequence number'; DECISION tf; (T): CALL check_number(aux); DECISION aux; (T): RESET(WEgress); CALL Book(rate,Null); OUTPUT FMsg(kind,rate,ci,el,mt,tf,seq) VIA UpS; NEXTSTATE Normal; (F): ENDDECISION; /* aux */ (F): ENDDECISION; /* tf */ NEXTSTATE -; ENDDECISION; /* kind */ INPUT BMsg(kind,rate,ci,el,mt,tf,ign_seq); OUTPUT FMsg(Ack,rate,T,T,T,F,seq) VIA UpS; NEXTSTATE -; ENDPROCESS Egress; TR 101 123 V1.1.1 (1997-11) 52 A.4 Further considerations In the above specification, a protocol for supporting the ABT/DT capability has been presented. Note that the selection of the parameters of the protocol (W_Ingress, W_Egress, TRM, T'RM, etc.) has not been discussed. In fact, the implementation of the timers at network ingress and egress is not necessary if BCR negotiations are performed periodically by the network, in which case either user does not need to initiate any BCR negotiation and has just to acknowledge the Bandwidth Acknowledgement RM cells sent by the network. This is a possible use of the above protocol for which some options are removed. TR 101 123 V1.1.1 (1997-11) 53 History Document history V1.1.1 November 1997 Publication ISBN 2-7437-1825-0 Dépôt légal : Novembre 1997 |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 1 Scope | The present document provides the mapping scheme to be used for the transport of ATM cells over MPEG-2 Transport Stream packets. The purpose of encapsulating ATM cells directly into an MPEG-2 transport stream is to: • interconnect ATM networks via DVB systems or provide ATM services to end-users using DVB systems; • offer differentiated quality of service to end-users by utilizing the ATM mechanisms. The main constraints of this problem are to: • providing an efficient encapsulation mechanism; • maintaining DVB/MPEG-2 compatibility when transporting ATM cells; • respecting ATM Quality-of-Service requirements throughout DVB/MPEG systems. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 2 References | The following documents contain provisions which, through reference in this text, constitute provisions of the present document. • References are either specific (identified by date of publication, edition number, version number, etc.) or non-specific. • For a specific reference, subsequent revisions do not apply. • For a non-specific reference, the latest version applies. • A non-specific reference to an ETS shall also be taken to refer to later versions published as an EN with the same number. [1] AF-PHY-0017.000: "The ATM Forum Technical Committee, UTOPIA Specification, Level 1, Version 2.01". [2] EN 301 192: "Digital Video Broadcasting (DVB); DVB specification for data broadcasting". [3] ITU-T Recommendation G.826: "Error performance parameters and objectives for international, constant bit rate digital paths at or above the primary rate". [4] ITU-T Recommendation I.356: "B-ISDN ATM layer cell transfer performance". [5] ITU-T Recommendation I.371: "Traffic control and congestion control in B-ISDN". [6] ITU-T Recommendation I.432.1: "B-ISDN user-network interface - Physical layer specifications: General characteristics". [7] ITU-T Recommendation I.432.2: "B-ISDN user-network interface - Physical layer specification: 155 520 kbit/s and 622 080 kbit/s operation". [8] ITU-T Recommendation I.610: "B-ISDN Operation and maintenance principles and functions". [9] ISO/IEC 13818-1: "Information technology - Generic coding of moving pictures and associated audio information: Systems". ETSI TR 100 815 V1.1.1 (1999-02) 6 |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 3 Abbreviations and definitions | |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 3.1 Abbreviations | For the purpose of the present document, the following abbreviations apply: ATM Asynchronous Transfer Mode CDV Cell Delay Variation CEC Cell Error Control CRC Cyclic Redundancy Check EDC Error Detection Code HEC Header Error Control LOC Loss Of Cell delineation LOS Loss Of Signal MPEG-TS MPEG Transport Stream OAM Operation, Administration and Maintenance PCR Program Clock Reference PES Packetized Elementary Stream PID Packet Identifier PL Physical Layer QoS Quality of Service RDI Remote Defect Indication REB Remote Errored Blocks SAR Segmentation And Re-assembly TS Transport Stream VBR Variable Bit Rate VC Virtual Channel VP Virtual Path |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 3.2 Definitions | For the purposes of the present document, the following terms and definitions apply: idle cell: a cell which is inserted and extracted by the physical layer in order to adapt the cell flow rate at the boundary between the ATM layer and the physical layer to the available payload capacity of the transmission used. valid cell: a cell whose header has no errors or has been modified by the cell Header Error Control (HEC) verification process. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 4 System description | The standard DVB/MPEG system defines all functionality's for encoding video and audio into MPEG Transport Streams (TSs), which are then multiplexed, along with data, into a single Transport Stream. This multiplex is modulated for transmission over the network. At the receiver side, the demodulator outputs the multiplex to a Transport Stream demultiplexer which extracts individual streams. The goal of this guideline is to give recommendations on carrying native ATM services using the standard DVB/MPEG system. Figure 1 shows an end-to-end DVB/MPEG conceptual block diagram, where additional blocks have been added to insert ATM cells into the system and extract ATM cells at the user side. ETSI TR 100 815 V1.1.1 (1999-02) 7 The 188 bytes (4 bytes header) MPEG transport packet is part of the data link sublayer defined to carry MPEG-2 video, audio, and data streams in DVB systems. A segmentation and re-assembly mechanism is needed to segment each information stream into MPEG-TS packets. Different schemes are presently defined in the MPEG standard ISO/IEC 13818-1 [9] and the DVB data broadcasting standard EN 301 192 [2]. In the MPEG standard ISO/IEC 13818-1 [9], the Packetized Elementary Stream (PES) and the Private Section are defined. For PES, MPEG defines an Adaptation Field mechanism which provides a technique to stuff MPEG-TS packets with stuffing bytes when transmit buffers are empty. For Private Sections, MPEG only defines a minimum structure. DVB specifies a format for the Private Sections in order to carry Multi-Protocol Encapsulation which can transport any type of Network Layer traffic. Also, another scheme defined by DVB is called data piping, consists in putting the raw traffic directly over MPEG-TS packets without any extra overhead. In ATM, the typical data link layer is composed of 53 bytes cells (5 bytes header) and is using its own segmentation and re-assembly mechanism (AAL). Usually, ATM cells are transported directly over a physical medium, which has its own transmission mechanism. Although ATM cells are asynchronous by nature, they are usually transmitted over the physical medium of a synchronous hierarchy network. Transmit and Receive buffers take care of following the constraints defined by ITU-T Recommendation I.356 [4]. Video Video server MPEG-TS MUX MPEG2 encoder ATM ATM switch or multiplexing ATM to MPEG-TS Encapsulation Modulator DVB transmission Demodulator Video / Audio ATM Demux ATM from MPEG-TS De-E ncapsulation Figure 1: Conceptual DVB/MPEG chain carrying ATM 5 Mapping of ATM cells into DVB/MPEG-2 Transport Stream |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.1 Frame format | The basic MPEG-TS packet structure of 188 octets as described in ISO/IEC 13818-1 [9] shall be used. The ATM cells are mapped into the 184 payload octets of the TS packet with the octet structure of the cell aligned with the octet structure of the packet. The data piping mechanism described in EN 301 192 [2] shall be used to transport ATM cells in MPEG-TS packets. The 53 bytes ATM valid cells are consecutively inserted into the 184 bytes payload of the MPEG- TS packets. When valid cells are not available from the ATM layer, the mechanism described in subclause 5.2 shall be used. ETSI TR 100 815 V1.1.1 (1999-02) 8 |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.2 Cell rate adaptation | The cell rate adaptation to the payload capacity of the MPEG-TS packets is performed either (i) by the insertion of idle cells, as described in ITU-T Recommendation I.432-1 [6] (see Figure 2), or (ii) by using the MPEG-TS adaptation fields mechanism, as described in ISO/IEC 13818-1 [9] (see Figure 3). A combination of these two mechanisms is possible. The start of an idle cell at the end of an MPEG-TS packet shall be completed at the beginning of the next packet with the same PID. The MPEG adaptation fields mechanism allows that the next MPEG-TS packet can start immediately with a new valid cell, if available. idle end idle start ATM start ATM ATM ATM ATM ATM packet n+1 packet n Figure 2: Idle cells inserted into the stream Adapt Field ATM ATM ATM ATM ATM ATM packet n+1 packet n ATM Start Figure 3: Adaptation fields inserted into the MPEG-TS header When no valid cell is available from the ATM layer during the transmission of a full MPEG-TS packet, either (i) the MPEG-TS packet is transmitted filled up with idle cells, or (ii) a NULL MPEG-TS packet is sent to the transport multiplexer. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.3 Header Error Control (HEC) generation | The HEC value is generated and inserted in the specific field in compliance with ITU-T Recommendation I.432.1 [6]. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.4 Scrambling of the ATM cell payload | The ATM cell payload (48 bytes) shall be scrambled before mapping into the MPEG-TS. In the reverse operation, following termination of the MPEG-TS, the ATM cell payload will be descrambled before being passed to the ATM layer. A self-synchronizing scrambler with the generator polynomial x 43 + 1, as described in ITU-T Recommendation I.432.1 [6], shall be used. Cell payload field scrambling is required to provide security against false cell delineation and replication of the MPEG-TS packets synchronization. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.5 Cell delineation | The cell delineation shall be performed using the Header Error Control (HEC) mechanism as defined in ITU-T Recommendation I.432.1 [6]. The mechanism for the detection of loss and recovery of cell delineation shall follow the description given in subclause 8.2.2 of ITU-T Recommendation I.432.2 [7]. Loss Of Cell delineation (LOC) causes a LCD defect. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.6 Cell header verification and extraction | The cell header verification shall be performed in compliance with ITU-T Recommendation I.432.1 [6]. All the physical layer cells shall be extracted and only the valid cells are passed to the ATM layer. Idle cells are discarded. ETSI TR 100 815 V1.1.1 (1999-02) 9 |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.7 Physical layer Operation And Maintenance (OAM) | OAM flows defined in ITU-T Recommendation I.610 [8] for ATM shall be transmitted along with traffic flows according to the present document. F1-F3 are physical layer related OAM flows, whereas F4-F5 flows are ATM layer related and are carried using ATM cells with a specific VCI. In order to avoid specific definitions in the MPEG-TS for ATM maintenance, F1 and F3 flows shall be carried in maintenance cells (PL-OAM) using a specific pattern in the header as defined for cell based transmission systems. The ATM cell header and the allocation of OAM functions in the information field of these special ATM cells shall follow the description given in subclause 7.2.2 of ITU-T Recommendation I.432.2 [7]. These cells are not passed to the ATM layer. F2 flow is not provided, but the associated functions are supported by F3 flows. Each MPEG stream shall carry its own F1-F3 flows. Maximum spacing between PL-OAM cells: The spacing between two adjacent PL-OAM cells is 216 cells using the block boundaries as described in subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7]. This interval consists of 8 monitored blocks with each block containing 27 cells. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.7.1 Signal processing for PL-OAM cells in the transmitter | a) Continuity Counter This function shall include for F1 and F3 flows, respectively, a continuos sequence number into the PSN field according to subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7]. b) Error Detection Code (EDC) This function shall calculate an EDC according to subclause 7.2.2.3 ITU-T Recommendation I.432.2 [7]. The result of this calculation is included into the appropriate EDC fields of the actual PL-OAM cell. c) Remote Errored Blocks (REB) This function performs the remote error control of the far end system as described in subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7] by using the contents of the REB field of the PL-OAM cell. NOTE 1: In case of unidirectional transmission, there is no associated receiving path termination and the REB field shall be set to a logical "0". d) Cell Error Control (CEC) This function shall perform a CRC calculation of the actual PL-OAM cell as described in subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7]. e) Transmission path alarm indication This function is only used in the F3 OAM cells in order to alert the equipment in the direction of transmission that a failure has been detected. The function shall follow the specification given in subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7]. f) Remote Defect Indication (RDI) If one of the defects described in subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7] are detected in the downstream path, the appropriate bits of the RS-RDI/TP-RDI field are set in order to alert the upstream equipment in the opposite direction of transmission that a defect has been detected along the downstream path. NOTE 2: In case of unidirectional transmission, there is no associated receiving path termination and the RDI field shall be set to a logical "0". ETSI TR 100 815 V1.1.1 (1999-02) 10 |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 5.7.2 Signal processing for PL-OAM cells in the receiver | a) Continuity Check For each type (F1 and F3) of PL-OAM cell, this function checks the correct spacing between two adjacent PL-OAM cells according to subclause 8.2.1 of ITU-T Recommendation I.432.2 [7]. b) Error Detection This function computes the EDC of the monitored interval according to subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7] and compares the result with the recovered values of the EDC field of the appropriate PL-OAM cell. A difference between the computed value and the recovered value is taken as evidence of one or more errors having occurred in the specified block. c) Cell Error Control (CEC) This function computes a CRC calculation of the actual PL-OAM cell as described in subclause 7.2.2.3 of ITU-T Recommendation I.432.2 [7] and compares its result with the recovered values of the CEC field of the PL-OAM cell. A difference between the computed value and the recovered value is taken as evidence of one or more errors having occurred in the actual PL-OAM cell. d) Remote Defect Indication (RDI) In order to enable single ended maintenance of a bi-directional transmission, the remote defect indication is recovered from the RS-RDI/TP-RDI field. NOTE: In case of unidirectional transmission, the contents of the RS-RDI/TP-RDI field is ignored. e) Loss Of Signal (LOS) The criteria for detection and clearance of LOS are given in subclause 8.2.1 of ITU-T Recommendation I.432.2 [7]. f) Performance monitoring This function shall generate performance parameters according to the specification given in annex D of ITU-T Recommendation G.826 [3]. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 6 Implementation aspects | |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 6.1 ATM to MPEG-TS gateway and MPEG-TS multiplexer | The MPEG-TS multiplexers usually have the possibility to reserve a constant bandwidth for a given stream. They do not have to maintain a fixed amount of time between successive MPEG-TS packets of the same PID, since the PCR mechanism is used at the receiver to recover synchronization. Yet, ATM traffic is asynchronous, and presents additional Quality of Service (QoS) constraints such as Cell Delay Variation (CDV), which shall be taken into account when transmitting it over a non perfectly synchronous MPEG-TS stream. ITU-T Recommendations I.356 [4] and I.371 [5] shall be followed to measure CDV over DVB/MPEG systems, and traffic shaping algorithms at the ATM to MPEG-TS gateway may need to be implemented, especially if a lot of Variable Bit Rate (VBR) connections are set up. Also, if the amount of time between successive MPEG-TS packets varies too much through time, the CDV tolerance may be exceeded. Therefore, it is important to try to respect the most periodic distribution of the MPEG-TS packets carrying ATM cells in the MPEG-TS multiplexer, as shown in Figure 4. ETSI TR 100 815 V1.1.1 (1999-02) 11 VCC 1 VCC 2 VCC 3 ATM to MPEG Encapsulation Point Arrival Time MPEG-TS PID k PID k PID k PID k ATM extraction time at Receiver τ > C D V } Not OK MPEG-TS PID k PID k PID k PID k ATM extraction time at Receiver } OK Figure 4: MPEG-TS packets distribution for ATM traffic support |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 6.2 MPEG-TS demultiplexer | The MPEG-TS demultiplexer may output directly ATM cells on a hardware interface such as the UTOPIA interface [1], in order to interface with other standard Segmentation And Re-assembly (SAR) devices for AAL layer processing. A separate ATM extraction device can also be placed after standard transport demultiplexers in order to extract ATM cells according to the present document. |
36e046c3afff655fa7a23f83eddc349b | 100 815 | 7 Conditional Access | Conditional Access may be implemented at the MPEG-TS level or at the ATM level, or at both. When applied at the MPEG-TS level, the entire stream is scrambled, but individual VP/VCs may not be independently scrambled. When applied at the ATM level, each VP/VC may be individually scrambled. ETSI TR 100 815 V1.1.1 (1999-02) 12 History Document history V1.1.1 February 1999 Publication ISBN 2-7437-2858-2 Dépôt légal : Février 1999 |
012f477a125960debab2c11aca72cb9f | 100 392-17-6 | 1 Scope | The present document identifies ETSI specifications and reports for TETRA V+D and DMO Release 2.2. It also includes ETSI specifications and reports relating to Critical Communications evolution undertaken by TC TCCE as well as relevant specifications produced by other ETSI technical committees. The present document is a revision of ETSI TR 100 392-17-6 V1.1.1 published in 2018 and contains the specifications and reports that have been produced since then. ENs relating to TETRA specifications are often updated by first introducing an ETSI Technical Specification (TS) which will contain new features and change requests. Therefore, in these cases, the latest specification is shown along with the previous version, if one is an EN and the other a TS. In the same way, EN versions of the same specification will also be updated, replacing a TS in some cases. Release 2.2 specifications were functionally frozen after the 52nd TC TCCE meeting in October 2018. Release 2.2 specifications contain corrections and enhancements to Release 2.1 including a major development of the Inter System Interface specifications to develop separate specifications for IP and PSS1 over E.1 transport layers as well as transport layer independent parts. NOTE 1: Functionally frozen means that no further functionality/features may be incorporated into the set of specifications, and that only corrective Change Requests (CRs) are to be accepted and agreed. NOTE 2: It can be expected that corrective CRs will be introduced into the Release 2.2 specifications. NOTE 3: Some of the CRs that will be produced will be for the next Release as they add new functionalities/features. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.