hash stringlengths 32 32 | doc_id stringlengths 7 13 | section stringlengths 3 121 | content stringlengths 0 2.2M |
|---|---|---|---|
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.5 Alert matching | |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.5.1 Introduction | Alert matching shall be performed either when the received FIG 0/15 OE flag is 0 and the Phase field is set to "Trigger", or when the OE flag is 1 (all OE signalling is Trigger phase signalling). Various criteria exist to determine if a signalled alert message will be played out: only when all the matching criteria are positive shall the alert message be played. The order of the matching process is arbitrary but can be optimized. |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.5.2 Receivability matching | If the OE flag is set to 0 then the alert message is carried in the tuned ensemble. The Id field identifies the subchannel that carries the audio of the alert message. If the subchannel is present in the FIG 0/1 MCI, receivability matching is positive, if not, it is negative. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 23 NOTE: A receiver cannot rely on stored configuration information when it has been in sleep mode because a reconfiguration may have occurred: the count field in FIG 0/7 provides for simple determination of a changed configuration. If the OE flag is set to 1 then the alert message is carried in another ensemble. Single tuner receivers shall determine whether the ensemble referenced by the EId in the Id field exists within tuning memory: if it is, receivability matching is positive, if not, it is negative. Receivers with multiple tuners shall confirm that the ensemble referenced by the EId in the Id field is receivable: if it is, receivability matching is positive, if not, it is negative. |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.5.3 Stage matching | Stage matching depends on the value of the Stage field and the value of user settings (if provided) as shown in table 1. Table 1: Stage matching Stage User setting for this EId + IId Match for receivers in Audio mode Match for receivers in Monitor mode (see note) Level 1 Start n/a Positive Positive Level 1 Update Incident dismiss (EId + IId) = false Positive Positive Incident dismiss (EId + IId) = true Negative Negative Level 1 Repeat Repeat dismiss (EId + IId) = false and Incident dismiss (EId + IId) = false Positive Positive Repeat dismiss (EId + IId) = true or Incident dismiss (EId + IId) = true Negative Negative Level 1 Critical n/a Positive Positive Level 2 Start n/a Positive Negative Level 2 Update Incident dismiss (EId + IId) = false Positive Negative Incident dismiss (EId + IId) = true Negative Negative Level 2 Repeat Repeat dismiss (EId + IId) = false and Incident dismiss (EId + IId) = false Positive Negative Repeat dismiss (EId + IId) = true or Incident dismiss (EId + IId) = true Negative Negative Test n/a Negative Negative NOTE: If a user setting to instruct the receiver to evaluate Level 2 alerts as Level 1 alerts is active then the Level 1 row shall be evaluated. |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.5.4 Location matching | If the alert signalling does not include any location codes, then the alert area is the entire ensemble coverage area. In this case, the location match is automatically positive. If the alert signalling includes location codes, the receiver compares the first location code in the alert set to the location code of its current position. For a fixed or portable receiver this is the location code captured at initialization, for a mobile receiver it is the location code computed from the GNSS received data: the same receiver location code shall be used for all location codes in the alert set. If the zone and digits common to both codes (left-aligned) are identical, the location match is positive and location matching terminates. If the zone and digits common to both codes (left-aligned) do not match, the next location code in the alert set is compared, and so on until either the location match is positive, or all location codes have been tested and a location match has not been found: in this case the location match is negative. Receivers can determine whether all FIG 0/15 instances have been evaluated for an alert set by means of the C/N flag and the NFF field. EXAMPLE: An alert set contains the location codes: Z1:91F, Z1:92C, Z1:953, Z1:960. The receiver location code is Z1:92CB81. The zone and digits of the receiver location code truncated to the same number of digits as the received location codes provide a match for Z1:92C, so the location match is positive on the second comparison and the final two location codes are not compared. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 24 |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.6 Alert mode | |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.6.1 Preparation | After an alert match is positive, the receiver shall store its functional status in memory to allow for a seamless return to that status after the alert playback is terminated. |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.6.2 Tuned ensemble alerts | The receiver shall play the audio from the subchannel indicated by the Id field. To determine whether the audio is coded as DAB audio or DAB+ audio it is necessary to find the SubChId in the FIG 0/2 MCI. If the alert audio carries PAD applications (for example, dynamic label or SlideShow), or the service includes data service components, these should be presented according to the capabilities of the receiver. The service label of the alert service shall be displayed. The alert audio shall be played within 5 s of the alert match decision. The receiver shall not delay audio playback if the service label is not yet available. NOTE: Although the service label will often be available from tuning memory, the alert service may have been introduced by an ensemble reconfiguration. The alert shall continue to be played whilst the Trigger or Sustain phase signalling with the same SubChId is received, unless the alert is terminated by the user (see clause 7.6.4). |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.6.3 Other ensemble alerts | The receiver shall attempt to retune to the ensemble indicated by the Id field using the information in its tuning memory. If the ensemble cannot be tuned, the receiver shall return to the previous functional state. NOTE: Receivability matching has already been performed, so if the ensemble cannot be tuned, the receiver should review the state of its tuning memory. Once tuned, the FIC shall be decoded and the MCI and FIG 0/15 information evaluated. Any Trigger phase or Sustain phase FIG 0/15 instance with the OE flag set to 0 provides the identity of the subchannel carrying the alert audio in the Id field. If no FIG 0/15 is received, or the subchannel does not exist in the FIG 0/1 MCI, the receiver shall return to the previous functional state. The receiver then follows the process for the tuned ensemble alert (clause 7.6.2). |
786b8d411b7b8adcb03f93d3f4b741b4 | 104 089 | 7.6.4 Terminating an alert | Receivers may offer various user functions to terminate the playback of an alert before it is completed. The user functions may be: • to terminate the currently playing alert only; • to terminate the currently playing alert and engage the "dismiss repeats" function to prevent future playback of repeated messages of the currently playing incident (identified by the EId + IId); • to terminate the currently playing alert and engage the "dismiss incident" function to prevent future playback of repeated and updated messages of the currently playing incident (identified by the EId + IId). During alert playback, the user may operate a user function to terminate the alert playback. The end of the alert message is usually signalled using FIG 0/15 with OE set to 0 and the Phase field set to "End", but in exceptional circumstances, when another alert message in the same ensemble immediately follows, the "End" phase is replaced by the "Trigger" phase of the new alert: the receiver shall terminate the alert playback. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 25 Whether alert playback is terminated by the user or by reaching the end of the alert message, audio playback shall be stopped and the receiver shall evaluate FIG 0/15 and determine if any other alerts are being signalled. If no such alerts are being signalled, the receiver shall return to the stored prior functional state; if such alerts are signalled, they shall be evaluated appropriately according to the type of receiver and if a match is found shall be played; otherwise the receiver shall return to the stored prior functional state. However, it shall be ensured that the receiver does not restart playing an alert that the user has terminated. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 26 Annex A (normative): DAB location code presentation format A.1 Introduction The coding of DAB location codes in FIG 0/15 is optimized for signalling efficiency, but it is not very human friendly. This annex describes the way that location codes are to be presented and used in the consumer setting, for example, for a user to be able to enter the location code at receiver set-up. The presentation format is designed to be suitable for entry on receivers with basic user controls, such as up/down/select functionality. Therefore, the number of symbols is restricted and the symbols are grouped into blocks. A.2 Conversion process The DAB location code at maximum resolution is a 30-bit binary coded integer. The most significant 6 bits represent the zone and the least significant 24 bits represent the six digits of the location code. The modulo-61 division of the 30-bit integer produces a 6-bit checksum. The checksum is appended to create a 36-bit integer. This 36-bit integer is separated into three blocks of four octal digits. A.3 Presentation format The presentation format consists of three groups of four symbols, separated with the hyphen character. The symbols used are the digits 1 to 8. Each octal digit (value range 0 to 7) is converted to a presentation symbol by adding 1 to give the symbols "1" to "8". EXAMPLE 1: The location code for BBC Broadcasting House is Z10:B736BB. In binary: Zone 10 = 001010; B736BB = 1011 0111 0011 0110 1011 1011. As a 30-bit binary integer: 001010101101110011011010111011. In decimal: 179 779 259. The modulo 61 checksum (in decimal): 59. The modulo 61 checksum (in binary): 111011. The 36-bit integer: 001010101101110011011010111011111011. The three block, 4-digit octal representation: 1255 6332 7373. The presentation code: 2366-7443-8484. EXAMPLE 2: The location code for Svalbard Museum is Z0:152FF1. In binary: Zone 0 = 000000; 152FF1 = 0001 0101 0010 1111 1111 0001. As a 30-bit binary integer: 000000000101010010111111110001. In decimal: 13 885 529. The modulo 61 checksum (in decimal): 47. The modulo 61 checksum (in binary): 101111. The 36-bit integer: 000000000101010010111111110001101111. The three block, 4-digit octal representation: 0005 2277 6157. The presentation code: 1116-3388-7268. NOTE: The generation of the location codes in these examples is given in annex F. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 27 A.4 URI definition To enable "smart" devices to use the presentation format, a URI is defined as follows: • DLI://<presentation code> Table A.1 shows the examples above in URI format with illustrative QR codes. Table A.1: Example of URI coding and QR code presentation of examples above Location URI encoded location string QR code for location BBC Broadcasting House Z10:B736BB DLI://2366-7443-8484 Svalbard Museum Z0:152FF1 DLI://1116-3388-7268 A.5 Possible webpage or app implementation example To provide users with the location code to input into their receiver, a map-based implementation could be devised. Figure A.1 shows an example. NOTE: Map data is available under the Open Database License, see Copyright and License | OpenStreetMap. Figure A.1: Example of map-based webpage to provide presentation format location code ETSI ETSI TS 104 089 V1.1.1 (2024-09) 28 Annex B (informative): Example of EWS signalling Figure B.1 illustrates the coverage areas of three ensembles, A, B and C which all participate in an EWS. Figure B.1: Coverage areas of ensembles A, B and C, and incidents 1 and 2 Ensembles A and C will each make synchronized alerts at the beginning of the next minute - ensemble A will make an alert concerning incident 1 and ensemble C will make an alert concerning incident 2. The coverage areas of the ensembles overlap, and so each ensemble signals all the alerts provided by these three ensembles. Figure B.2 shows the progression of the FIG 0/15 signalling made in each ensemble at the start of the alerts. Before the alert begins, all ensembles are signalling the Heartbeat once per second to identify themselves as EWS ensembles. Ensemble A and ensemble C each carry a synchronized alert message and so signal all phases of tuned ensemble signalling, beginning with the Pre-trigger phase at seconds count = 55 using the special Sec field value of 63. This signalling is received off-air and fed to the FIC generators of the other ensembles to construct their OE signalling. Starting at seconds count = 0, ensembles A and C signal Trigger phase signalling for their own alert followed by OE signalling for the alert in the other ensemble, and ensemble B signals OE signalling for both alerts. In each case, the two alert sets may have one to four FIG 0/15 instances to provide the alert area, and the two alert sets form the alert group. The signalling of the alert group is repeated continuously until seconds count = 4 (inclusive). Due to the length of the alert group, the trigger signalling spills into the beginning of seconds count = 5. Ensembles A and C then begin signalling Sustain phase at a rate of once per second until the end of the alert message, and ensemble B signals the Heartbeat. Figure B.3 shows a later period where the alert concerning incident 1 has ended. In ensemble C, the alert concerning incident 2 comes to its end during seconds count = 42, with End phase signalling provided rapidly for two seconds. An update is made concerning incident 1 in ensemble A at seconds count = 44 and the ensemble signals Pre-trigger phase signalling from seconds count = 39 until seconds count = 41 (inclusive) with the Sec field set to 44. Ensemble A signals trigger phase continuously from seconds count = 45 for 5 seconds and then signals trigger phase at once per second for 1 additional second due to the length of the start jingle. Ensembles B and C provide OE signalling from seconds count = 44: in ensemble C the OE Trigger phase signalling overlaps with the End phase signalling. The OE signalling in ensembles B and C continue longer than the trigger phase signalling in ensemble A dues to processing delays between off-air reception of ensemble A and FIC encoding. Incident 1 Incident 2 Ensemble A Ensemble C Ensemble B ETSI ETSI TS 104 089 V1.1.1 (2024-09) 29 Figure B.2: FIG 0/15 signalling in ensembles A, B, and C: start of synchronized alerts for incidents1 and 2 programme 53 54 55 56 57 58 59 00 01 02 03 04 05 06 07 alert seconds count audio FIG 0/15 insertion – ensemble A Heartbeat Pre-trigger Trigger Other ensemble Sustain FIG 0/15 type FIG 0/15 insertion – ensemble B FIG 0/15 insertion – ensemble C 1 2 Incident number End ETSI ETSI TS 104 089 V1.1.1 (2024-09) 30 Figure B.3: FIG 0/15 signalling in ensembles A, B, and C: end of alert for incident 2 and start of a new alert for incident 1 programme 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 alert seconds count audio FIG 0/15 insertion – ensemble A Heartbeat Pre-trigger Trigger Other ensemble Sustain FIG 0/15 type FIG 0/15 insertion – ensemble B FIG 0/15 insertion – ensemble C 1 2 Incident number End End and OE Trigger provided in parallel OE Triggers continue longer than TE Trigger ETSI ETSI TS 104 089 V1.1.1 (2024-09) 31 Annex C (informative): Example of coding an alert area Figure C.1 shows an approximation of the official service area of the Cardiff and Newport ensemble in the United Kingdom (red area). It also shows an approximation of the overspill area (pink area) where good reception is a consequence of geography and transmitter antenna patterns. The ensemble area is in Zone 10, which extends from (36, -36) to (72, 0). The first division of Zone 10 produces areas of 9° × 9° and the second of 2,25° × 2,25°. The area shown in figure C.1 extends from (49,5, -4,5) to (51,75, -2,25) and is identified by location code Z10:B6. It is divided into 16 sub-areas, which are labelled. Figure C.1: Coverage of Cardiff-Newport ensemble within the Z10:B6 spherical rectangle (Map source: © 2024 Google) It can be seen that for this location (about 51° N) that the spherical rectangles have an aspect ratio of about 2:3 (cos 51° = 0,6293). ETSI ETSI TS 104 089 V1.1.1 (2024-09) 32 When an emergency situation arises in Cardiff, the authorities wish to provide a warning to those affected, whilst not disturbing others, especially since the overspill area is some distance away. Therefore, location codes need to be provided to identify the valid area of the alert. It can be seen that Cardiff is located in sub-area 2. Figure C.2 provides an enlargement of this sub-area. Figure C.2: Cardiff alert area within the Z10:B62 spherical rectangle (Map source: © 2024 Google) It can be seen that the alert area is contained within four sub-areas of the Z10:B62 spherical rectangle. Figure C.3 provides an enlarged view of the alert area. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 33 Figure C.3: Cardiff alert area detail (Map source: © 2024 Google) The alert area consists of 17 spherical rectangles with 5-digit location codes (in this location, each spherical rectangle with a 5-digit code is approximately 4 km north-south by 2,5 km east-west). Since the area is within four 4-digit location codes, the area can be coded successfully by using sub-codes. The alert area is coded with four location codes (LC) as follows: • LC1: Zone = 10; SCF = 1; Num Digits = 3; Digit 1 = B; Other digits = 624; Padding = 0; Sub-codes = CC00. • LC2: Zone = 10; SCF = 1; Num Digits = 3; Digit 1 = B; Other digits = 625; Padding = 0; Sub-codes = F730. • LC3: Zone = 10; SCF = 0; Num Digits = 4; Digit 1 = B; Other digits = 6283. • LC4: Zone = 10; SCF = 1; Num Digits = 3; Digit 1 = B; Other digits = 629; Padding = 0; Sub-codes = 0007. LC3 is coded without the use of sub-codes as there is only one sub-area inside the alert area. The location codes use a total of 22 bytes (3 × 6 + 4 bytes). The four location codes therefore fit into a single FIG 0/15. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 34 Annex D (informative): Alert area translation method D.1 Introduction The alert area is the geographical area in which users are presented with the alert message by the matching algorithm in the EWS receiver. The alert area may be the entire coverage area of the DAB ensemble, or it may be an area defined by use of DAB location codes. In the latter case, the maximum number of location codes that can be signalled is limited in order to ensure that the alert area is signalled in a timely way. The raw alert information will typically be provided by a national authority and the source format most commonly uses the CAP standard [i.2]. The CAP standard specifies alert areas as polygons or circles, using co-ordinate systems, or as geocodes which have national encoding formats such as SAME or SHN. For many national formats official public domain area definitions exist, so that a highly accurate geographic polygon can be obtained from an administrative entity identifier by way of dereferencing. A DAB EWS has a coverage area within which suitable receivers can respond to signalled alerts. The coverage area of the DAB EWS is the superposition of all the DAB ensembles that participate in the EWS. Every incident will be relevant to a particular geographical area. This alert area may be deemed to be sufficiently close to one of the coverage areas of a participating EWS ensemble that the entire ensemble area is appropriate as the alert area, or it may be that the alert area does not correspond to any EWS ensemble coverage area, in which case the alert area will be described by the use of location codes. The general recommendation is that the ensemble that carries the alert message has a coverage area that is greater than the full extent of the alert area. A DAB EWS may be centrally managed, in which case the assignment of incidents to ensembles will be a matter of the management system, or it may be managed in a distributed way, whereby ensembles manage their own provision of incidents. In the latter case, there is the possibility that more than one EWS ensemble could choose to carry an alert for the same incident, although the precise content of the alert messages will be different in some ways. The alert source information will usually be provided by a responsible institution and will be the result of an administrative decision and may also represent a sovereign act that is legally protected. As such the process to translate the source alert area into DAB location codes should not reduce the spatial extent. A DAB location code identifies a spherical rectangle; the dimensions of the spherical rectangle are defined by the number of digits that make up the location code - the larger the number of digits, the smaller the area of the spherical rectangle identified. The DAB alert area is therefore the addition of all the spherical rectangles described by a set of location codes needed to include the entire alert area described by the source data. The DAB alert area is necessarily different from the source alert area. The translation procedure ensures that the output set of location codes completely includes the source alert area, except if it can be numerically shown that the source alert area overlaps a location code only by a miniscule amount: in this case the location code can be omitted from the output set of location codes. This annex contains a reference encoding procedure for generating a set of location codes that adheres to the EWS signalling constraints that allow a maximum of four location code fields, each with a maximum length of 25 bytes. D.2 Translation procedure D.2.1 Generating the source data polygon(s) Whatever the format of the source alert data, each alert area needs to be accurately mapped to one or more polygons using the WGS84 coordinate system, as this is the basis of the DAB location coding system. In the process described in this annex, each alert area is assumed to be described by such polygons using WGS84 coordinates. For source data defined in a coordinate system other than WGS84, a coordinate transform is applied to the source data before the translation procedure. The process aims to provide the best fit to the source alert area within the capacity constraints of the signalling whilst ensuring that accuracy requirements are met. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 35 The alert area defined in the alert source information is accurately mapped to one or more polygons, each consisting of a line string of geographic points, each expressed as a pair of polar coordinates defined by WGS84 latitude and longitude. D.2.2 Generating the location code set D.2.2.1 Introduction The translation procedure consists of four steps. The algorithm is defined on an abstract level in natural language as to not limit the implementation beyond the requirements. The "level" describes the number of hexadecimal digits in the location code and thus the size of the area described: level 1 (L1) has one digit and describes the largest area; level 6 (L6) has six digits and describes the smallest area. D.2.2.2 Initial parent level determination From the source data polygon(s), determine the dimensions (in degrees) of the latitude extent and the longitude extent of the alert area: the latitude extent is the difference between the minimum and maximum values of all the geographic points of the polygon(s), similarly the longitude extent. The smallest extent (either latitude or longitude) determines the initial parent level as given by table D.1. Table D.1: Initial parent level from smallest extent Smallest extent, E E > 9° 9° ≥ E > 2,25° 2,25° ≥ E > 0,5625° 0,5625° ≥ E > 0,140625° 0,140625° ≥ E Parent level L1 L2 L3 L4 L5 D.2.2.3 Generating the parent set of location codes The parent set is the set of location codes at the parent level that when taken together completely include the alert area as defined by the source data polygon. If the parent set contains more location codes than the count threshold defined in table D.2, the parent level is reduced by one (that is, it now describes a larger spherical rectangle) and a new parent set at the new parent level is generated. Table D.2: Count thresholds at parent level Parent Level L1 L2 L3 L4 L5 L6 Count Threshold 24 24 20 20 16 - D.2.2.4 Generating the child set of location codes The child level is set at one level greater than the parent level (that is it describes a smaller spherical rectangle). The child set is the set of location codes at the child level that when taken together completely include the alert area as defined by the source data polygon(s). If the child set contains fewer than, or an equal number of location codes to, the count threshold defined in table D.3, the child level is increased by one (that is, it now describes a smaller spherical rectangle) and a new child set at the new child level is generated. Table D.3: Count thresholds at child level Child Level L1 L2 L3 L4 L5 L6 Count Threshold - 24 20 20 16 16 ETSI ETSI TS 104 089 V1.1.1 (2024-09) 36 D.2.2.5 Removing the "miniscule" location codes When a location code in the child set overlaps the source data polygon(s) by only a miniscule amount, the inclusion of the location code becomes disproportionate because the vast majority of the area described is outside the source alert area. The overlap between the alert areas defined by the source data polygon(s) and a location code exists if any point of the alert area lies within the location code spherical rectangle. To avoid this situation, location codes in the child set are discarded if they overlap by an area that is smaller than the area threshold. The area threshold is a fraction of the spherical rectangle area with a numerator of 1 and the denominator according to the location code level as defined in table D.4. Table D.4: Area thresholds Level L1 L2 L3 L4 L5 L6 Area threshold (1/N) - 4096 1024 256 64 16 The output set of location codes is the result of discarding the miniscule location codes from the child set. The output set is the encoded alert area and is used for output formation. D.2.3 Output formation The alert area is carried in FIG 0/15 using a formatted set of location codes. In the output set, all the location codes are at the same level, which is the output level. The level below the output level is called the stem level (that is, it has one fewer hexadecimal digit). For efficient coding in the FIG 0/15, location codes can be grouped together using the stem level and sub-coding. The output set of location codes is grouped into sub-sets sorted by the stem level location code: if location codes in the output set share a common stem level code, they fall into the same group. Every group has up to 16 location codes, only different in the last hexadecimal digit. Each group is then encoded to a format that depends on the group size, as defined in table D.5. Table D.5: Location code format in output Group size 1 2 to 15 16 Encoded format Output level Stem level with subcode Stem level If the group has one location code, it is encoded as a single location code at the output level as this is the most efficient coding. If a group has more than one and fewer than 16 location codes, the group is coded as the stem level location code with a sub-code field that has one bit set for every location code in the group determined by its last digit. If a group has 16 location codes, it is coded as a single location code at stem level. NOTE: The translation procedure is defined such that the number of groups will never exceed the parent level threshold. This guarantees that the entire output set can be transmitted within four FIG 0/15 instances. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 37 Annex E (normative): Definition of FIG 0/15 EWS information The EWS information description is encoded in Extension 15 of FIG type 0 (FIG 0/15). Figure E.1 shows the structure of the EWS information field which is part of the Type 0 field (see also ETSI EN 300 401 [1], figure 7). FIG 0/15 signals participation of an ensemble in an EWS, and the status and localization for EWS audio alerts that are carried in the tuned ensemble or in another ensemble. The alert area, when smaller than the complete ensemble coverage area, is defined by the use of location codes (see annex F). In order to allow complex alert areas to be signalled, a set of FIG 0/15s can be signalled: the set comprises between one and four FIG 0/15 instances. One or more alerts may be signalled concurrently: in this case there will be a set of one to four FIG 0/15s for each alert; the complete collection of FIG 0/15s that describe the current alert information is known as an alert group. The FIG type 0 flags (see ETSI EN 300 401 [1], clause 5.2.2.1) are used as follows: C/N flag - SIV; OE flag - OE; P/D flag - special definition. The P/D flag is used to ensure synchronization between the transmission equipment and receiving equipment that operates a sleep mode. It shall be set as follows: • 0: Process • 1: Discard This feature shall use the SIV signalling (see ETSI EN 300 401 [1], clause 5.2.2.1). The database shall be divided by use of a database key. Changes to the database shall not be signalled using the CEI mechanism, because an alert database only exists during the alert. A special form of FIG 0/15 is used as a heartbeat indicator (control function, C/N flag is set to 1) to identify that the ensemble is part of an EWS and will carry alert information when alerts are active. The heartbeat form has an empty type 0 field and is identified by the length field of the FIG type 0 header being equal to 1. The OE flag shall be set to 0. The heartbeat form of FIG 0/15 shall not be transmitted when alerts are signalled. Figure E.1: Structure of EWS information field Location code a EId SubChId 2 bits 6 bits 16 bits Location code l Incident Id .... b15 b0 b7 b5 b0 Type 0 field for extension 15 Tuned ensemble (OE = 0) Other ensemble (OE = 1) Id field 8 or 16 bits Padding Digit 1 0 to 20 bits 0 or 4 bits Num digits 3 bits b6 b4 1 bit b7 Other digits 4 bits 4 bits b3 b0 Stage 3 bits b6 b4 6 bits Zone SCF b5 b0 NFF 2 bits b7 b6 b3 b0 0 or 16 bits Sub-codes Last 1 bit b7 Phase b6 Status field 0 or 8 bits 0 to 48 bits 0 to 48 bits 0 or 2 bits Sec Rfa 0 or 6 bits b5 b0 b7 b6 ETSI ETSI TS 104 089 V1.1.1 (2024-09) 38 The following definitions apply: Id field: this 8- or 16-bit field shall contain the identity of the alert audio service component or the ensemble that carries the alert audio service component. The OE flag in the FIG type 0 header field (see ETSI EN 300 401 [1], clause 5.2.2.1) shall determine if the service component is in the tuned ensemble or another ensemble. The following definitions apply: • Phase: this 2-bit field shall indicate the phase of the tuned ensemble alert, as follows: 0 0: Pre-trigger; 0 1: Trigger; 1 0: Sustain; 1 1: End. • SubChId: this 6-bit field shall identify the sub-channel which contains the audio service component carrying the audio alert. • Rfa: this 2-bit field shall only be present when the Phase field = Pre-trigger. The bits are reserved for future addition and shall be set to 0. • Sec: this 6-bit field, coded as an unsigned binary integer, shall only be present when the Phase field = Pre-trigger and shall contain the seconds count corresponding to the start of the Trigger phase. • EId (Ensemble Identifier): this 16-bit field shall identify the other ensemble that carries the audio alert. Status field: this 8-bit field, when present, shall provide information about the alert. The following definitions apply: • Last: this 1-bit flag shall indicate if this is the last FIG 0/15 to be signalled in this alert group as follows: 0: not the last FIG 0/15 of the alert group 1: the last FIG 0/15 of the alert group • Stage: this 3-bit field shall indicate the stage of the EWS alert, as follows: 0 0 0: Level 1 Start 0 0 1: Level 1 Update 0 1 0: Level 1 Repeat 0 1 1: Level 1 Critical 1 0 0: Level 2 Start 1 0 1: Level 2 Update 1 1 0: Level 2 Repeat 1 1 1: Test • Incident Identifier (IId): this 4-bit field, coded as an unsigned binary integer, shall identify an incident that is composed of a series of audio alerts. The value is assigned by the ensemble that carries the audio alert and shall be maintained through the various stages of an incident. Location code: these n × 8-bit fields define the geographic area that the EWS alert applies to. The areas specified are additive. If no location codes are provided then the alert is relevant to the entire ensemble coverage area. The number of location codes carried is not specified, but location codes fill the rest of the FIG 0/15 specified by the length field of the FIG 0/15 header. The following definitions apply: • NFF: this 2-bit field, expressed as an unsigned binary number in the range 0 to 3, shall indicate the number of FIG 0/15s for this alert set which follow in order to signal the complete alert area. NOTE 1: NFF is not part of the location code, it is provided for determining the integrity of the complete alert area. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 39 NOTE 2: When the C/N flag in the type 0 header is set to 0, (NFF + 1) indicates the total number of FIG 0/15s in the alert set. • Zone: this 6-bit field, expressed as an unsigned binary number, shall identify the global zone as defined in annex E. • SCF: this 1-bit flag shall indicate whether the Sub-codes field is present or not, as follows: 0: Sub-codes field is absent: the location code contains a single spherical rectangle. 1: Sub-codes field is present: the location code contains 2 to 15 spherical rectangles. • Num digits: this 3-bit field, expressed as an unsigned binary number, shall indicate the number of digits, in the range 0 to 5, contained in the subsequent Other digits field. NOTE 3: When SCF=1, the smallest spherical rectangles are sub-coded and Num digits is in the range 0 to 4. • Digit 1: this 4-bit field shall specify the most significant digit of the location code (see annex E). The special value 0 may be used for the polar zones to indicate the entire zone. • Other digits: this 0- to 20-bit field shall only be present when the Other digits field is greater than 0. It shall contain all the remaining digits of the location code when SCF = 0 and all but the least significant digit of the remaining digits when SCF = 1. NOTE 4: When SCF=1, the smallest spherical rectangles are sub-coded and so there are a maximum of 4 other digits. • Padding: this 4-bit field shall only be present when Num digits is an odd number. All bits shall be set to 0. • Sub-codes: this 16-bit flag field shall only be present when the SCF flag is set to 1. It shall specify between 2 and 15 spherical rectangles that are the sub-areas of the area specified by the zone and coded digits. The definition of the sub-areas is given in annex E. The use of sub-codes permits efficient coding of a more precisely defined area to be specified. The flags shall be coded as follows: bi (i = 0 to 15) 0: sub-area i is not part of alert area 1: sub-area i is part of alert area NOTE 5: For 1 spherical rectangle, the coding efficiency is greater by including the least significant digit in the Other digits field; for 16 spherical rectangles, the sub-coding is not required as all sub-areas are part of the alert area. The database key comprises the OE flag (see ETSI EN 300 401 [1], clause 5.2.2.1) and the Id field. There is no Change Event Indication (CEI). When no alert is signalled, the heartbeat form of FIG 0/15 shall be used to indicate that the ensemble is part of the EWS. NOTE 6: If multiple EWS alerts are active concurrently, the capacity of the FIC for all other features could be reduced. Receivers may expect that the nominal repetition rates of FIG 0/15 and of other FIGs may be affected. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 40 Annex F (normative): DAB location coding F.1 Introduction Incidents and activities will rarely affect the same area as the coverage area of a particular DAB ensemble, and in many cases DAB ensembles have overlapping coverage areas. DAB location coding is applicable to any place on earth and uses a hierarchy of spherical rectangles (that is, rectangles defined in terms of polar coordinates). Each location code has a variable length with the first level of the code identifying the largest area, with each subsequent level identifying a sub-division of the previous level. The longer the code, the finer the spatial resolution. By providing a set of location codes, any arbitrary area may be described. A location code consists of a top-level zone using 6-bits and up to six sub-division levels, each using 4-bits, to form location codes with a maximum length of 30 bits and a minimum length of 10 bits (zone plus one sub-division). The maximum spatial resolution of the scheme is 977 m north-south. In the east-west direction, the spatial resolution is 978 m at the equator and 302 m at 72° latitude. This level of precision is appropriate for a broadcast system. Each digit removed from the end of the location code reduces the spatial resolution by factor four, and increases the area described by a factor of 16. F.2 Coordinates In WGS84, the latitude is 0° at the equator, 90° at the north pole and -90° at the south pole, and the longitude is given as the position east of the Greenwich meridian, with 0° at the meridian and increasing positively eastwards to 180° and negatively westwards to -180°. To make calculations simpler for determining the location code, the WGS84 coordinates are translated using a simple mapping to use positive numbers only. The WGS84 latitude is translated to the Southerly Extent (SE) by subtracting the WGS84 latitude from 90°; thus the north pole becomes 0°, the equator 90° and the south pole 180°. The WGS84 longitude is translated to the Easterly Extent (EE) by adding 360° to negative values only; thus, for example, -10° becomes 350°. F.3 Division into zones The top level of the hierarchy is a zone. There are 42 zones, as shown in figure F.1: • 40 banded zones, which are spherical rectangles 36° × 36°; and • 2 polar zones, also 36° × 36°, which extend 18° from the poles. The vast majority of the populated area of the globe is contained within the banded zones, that is between 72°N and 72°S from the equator. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 41 Figure F.1: Representation of the zones The north polar zone is zone 0; the south polar zone is zone 41. The banded zones are numbered 1 to 40, in four rows of 10 zones, moving southwards and eastwards. The identity of the zone may be determined from the SE and EE values as follows: case (SE < 18) zone = 0; case (18 ≤ SE < 162) zone = 10 × int((SE - 18)/36) + int(EE/36) + 1; case (SE ≥ 162) zone = 41; where int indicates that only the integer part of the result is used. Zone 1 therefore comprises the area contained within WGS84 coordinates from (36, 0) to (72, 36); zone 2 from (36, 36) to (72, 72); … ; zone 10 from (36, -36) to (72, 0); zone 11 from (0, 0) to (36, 36); and so on to zone 40 from (-72, -36) to (-36, 0). This is shown in figure F.2. Polar zone Banded zone Zone 0 Zone 1 Zone 10 Zone 11 Zone 40 Zone 41 Equator ETSI ETSI TS 104 089 V1.1.1 (2024-09) 42 Figure F.2: Position of the zones (variant of the Mercator projection) The polar zones have a radius of 18°, approximately 2 000 km. The zones bordering the equator are approximately 4 000 km north-south; 4 000 km at the equator, and 3 200 km at 36° from the equator, east-west. The zones adjoining the polar zones are also approximately 4 000 km north-south; 3 200 km at 36° from the equator, and 1 250 km at 72° from the equator, east-west. F.4 Division of the banded zones The first division of the banded zones, and all subsequent divisions, are performed as follows. The area is divided into 16 sub-areas of equal polar co-ordinate dimensions in a 4 × 4 pattern, as shown in figure F.3. Each sub-area is numbered to identify its position, beginning at the north-west corner and moving eastwards and southwards. NOTE: Using a 4-bit integer to identify a sub-area, the left-most (most significant) 2 bits count north-south and the right-most (least significant) 2 bits count west-east. ETSI ETSI TS 104 089 V1.1.1 (2024-09) 43 Figure F.3: Division of spherical rectangles For any point on the earth's surface within the banded zones, the digits for the location code are calculated as follows: Southerly Code (SC) = int(frac((SE - 18)/36) × 212) expressed as a 12-bit integer Easterly Code (EC) = int(frac(EE/36) × 212) expressed as a 12-bit integer where int indicates that only the integer part of the result is used and frac indicates that only the fractional part of the result is used. The Combined Code (CC) is then formed by interleaving the SC and EC, 2-bits at a time, to produce a 24-bit integer, the most significant 2-bits are from the SC, the next 2-bits from the EC, and so on. That is, the digits representing a sub-area at each level of the 24-bit integer are formed by combining piecewise 2-bits from the SC and 2-bits from the EC. EXAMPLE: BBC Broadcasting House in London is located at WGS84 (51,5187412, -0,1434571) First, the coordinates are translated: SE = 90 - 51,5187412 = 38,4812588 EE = -0,1434571 + 360 = 359,8565429 Second, the zone number is calculated: Zone = 10 × int((38,4812588 - 18)/36) + int(359,8565429/36) + 1 = 10 Third, the digits are calculated: SC = int(frac((38,4812588 - 18)/36) × 4 096) = 2330 = 91A16 = 10 01 00 01 10 102 EC = int(frac( 359,8565429/36) × 4 096) = 4079 = FEF16 = 11 11 11 10 11 112 CC = 1011 0111 0011 0110 1011 10112 = B736BB16 The location code for BBC Broadcasting House is thus Z10:B736BB. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 South East ETSI ETSI TS 104 089 V1.1.1 (2024-09) 44 F.5 Division of the polar zones The polar zones have a special coding for their first digit, because the area of a polar zone is circular. The first division of the polar zones is shown in figure F.4. Each sub-area is numbered to identify its position. Figure F.4: First division of polar zones (WGS84 co-ordinates) Subsequent digits are divided in the same way as for the banded zones (see figure F.3). For any point on the earth's surface within the north polar zone, the digits for the location code are calculated as follows: if (SE < 9) then { Digit 1 = int(EE/72) + 11 Southerly Code (SC) = int(SE/9 × 210) expressed as a 10-bit integer Easterly Code (EC) = int(frac(EE/72) × 210) expressed as a 10-bit integer} else { Digit 1 = int(EE/36) + 1 Southerly Code (SC) = int(frac((SE - 9)/9) × 210) expressed as a 10-bit integer Easterly Code (EC) = int(frac(EE/36) × 210) expressed as a 10-bit integer}; the Combined Code (CC) is then formed by concatenating Digit 1 (most significant) with the result of interleaving the SC and EC, 2-bits at a time, to produce a 20-bit integer, the most significant 2-bits are from the SC, the next 2- bits from the EC, and so on. For any point on the earth's surface within the south polar zone, the digits for the location code are calculated as follows: if (SE < 171) then { Digit 1 = int(EE/36) + 1 Southerly Code (SC) = int(frac((SE - 162)/9) × 210) expressed as a 10-bit integer Easterly Code (EC) = int(frac(EE/36) × 210) expressed as a 10-bit integer} else { Digit 1 = int(EE/72) + 11 Southerly Code (SC) = int(frac((SE - 171)/9) × 210) expressed as a 10-bit integer Easterly Code (EC) = int(frac(EE/72) × 210) expressed as a 10-bit integer}; (72, 0) (72, 36) (72, 72) (72, -36) (81, 0) 7 6 5 4 3 2 1 10 9 8 11 15 14 13 12 North pole - Zone 0 (-72, 0) (-72, 36) (-72, 72) (-72, -36) (-81, 0) 7 6 5 4 3 2 1 10 9 8 11 15 14 13 12 South pole - Zone 41 ETSI ETSI TS 104 089 V1.1.1 (2024-09) 45 the Combined Code (CC) is then formed by concatenating Digit 1 (most significant) with the result of interleaving the SC and EC, 2-bits at a time, to produce a 20-bit integer, the most significant 2-bits are from the SC, the next 2- bits from the EC, and so on. NOTE: In the polar zones, the linear dimension of the edge of spherical rectangles that touch the pole is 0. EXAMPLE: Svalbard Museum in Longyearbyen is located at WGS84 (78,222609, 15,651605) First, the coordinates are translated: SE = 90 - 78,222609 = 11,777391 EE = 15,651605 Second, the zone number is calculated: Zone = 0 Third, the digits are calculated: Digit 1 = int(15,651605/36) + 1 = 1 SC = int(frac((11,777391 - 9)/9) × 1 024) = 316 = 13C16 = 01 00 11 11 002 EC = int(frac(15,651605/36) × 1 024) = 445 = 1BD16 = 01 10 11 11 012 CC = 0001 0101 0010 1111 1111 00012 = 152FF116 The location code for Svalbard Museum is thus Z0:152FF1 ETSI ETSI TS 104 089 V1.1.1 (2024-09) 46 History Document history V1.1.1 September 2024 Publication |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 1 Scope | The present document specifies methods to efficiently build and instantiate Key Encapsulation Mechanisms (KEMs) with hidden access policies, while having the privacy of encapsulated keys relying on the best security of two hybridized schemes, namely with an instantiation where the privacy relies on the Computational Diffie-Hellman (CDH) classical assumption and the Learning With Errors (LWE) post-quantum assumption. Both problems have to be broken to endanger the privacy of the encapsulated key. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 2 References | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 2.1 Normative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. Referenced documents which are not found to be publicly available in the expected location might be found in the ETSI docbox. NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee their long term validity. The following referenced documents are necessary for the application of the present document. [1] NIST SP 800-186: "Recommendations for Discrete Logarithm-Based Cryptography: Elliptic Curve Domain Parameters". [2] IETF RFC 7748: "Elliptic Curves for Security". [3] NIST SP800-185: "SHA-3 Derived Functions: cSHAKE, KMAC, TupleHash and ParallelHash". [4] FIPS PUB 180-4: "Secure Hash Standard (SHS)". [5] FIPS PUB 202: "SHA-3 Standard: Permutation-Based Hash and Extendable-Output Functions". [6] FIPS PUB 203: "Module-Lattice-Based Key-Encapsulation Mechanism Standard". |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 2.2 Informative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee their long term validity. The following referenced documents are not necessary for the application of the present document but they assist the user with regard to a particular subject area. [i.1] Théophile Brézot, Paola de Perthuis, and David Pointcheval: "Covercrypt: an Efficient Early-Abort KEM for Hidden Access Policies with Traceability from the DDH and LWE". ESORICS 2023. [i.2] Vipul Goyal, Omkant Pandey, Amit Sahai, and Brent Waters: "Attribute-Based Encryption for Fine-Grained Access Control of Encrypted Data". ACM CCS 2006. [i.3] Amit Sahai and Brent Waters: "Fuzzy identity-based encryption". EUROCRYPT 2005. [i.4] ISO/IEC 18033-2: "Information technology — Security techniques — Encryption algorithms — Part 2: Asymmetric ciphers". ETSI ETSI TS 104 015 V1.1.1 (2025-02) 7 |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 3 Definition of terms, symbols and abbreviations | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 3.1 Terms | For the purposes of the present document, the following terms apply: adversary's advantage: probability for an adversary to distinguish two distributions NOTE: Formally, for an adversary A, given two distributions D0 and D1, the advantage is defined as: Adv A = Pr A x = 1 −Pr A x = 1 = 2 . Pr , A x = b −1. negligible probability in κ: probability that is smaller than the inverse of any polynomial in κ, for κ large enough oracle access: efficient evaluation of a function for inputs of their choice overwhelming probability in κ: probability p such that 1-p is negligible in κ polynomial time: running time can be expressed as a polynomial in the security parameter security parameter: number of bits in the security level NOTE: If the security parameter is equal to κ, then the security should hold except with probability less than 2-κ. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 3.2 Symbols | For the purposes of the present document, the following symbols apply: 1κ The security parameter κ taken as input to an algorithm || The logical (non-exclusive) OR && The logical AND ⊕ The logical XOR x ← f(y) x is the output of the algorithm f applied to the input y. Unless stated otherwise, f is a randomized algorithm, implicitly also using an input of random coins x $← S x is drawn from a uniform distribution on the finite set S ¬A For an event A, the event in which A does not happen D = { A : B } The distribution of B given A (where A will specify the distribution from which B is taken). For instance, D = { a $← : a } denotes the distribution of a knowing that a is drawn from a uniform distribution on the finite set f : X → Y The function f takes input values in the space X and outputs values in Y ⊥ Output to an algorithm that indicates that it has failed and returns nothing, except for the indication that it did not terminate correctly |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 3.3 Abbreviations | For the purposes of the present document, the following abbreviations apply: ABE Attribute-Based Encryption CCA Chosen-Ciphertext Attacks CDH Computational Diffie-Hellman CPA Chosen-Plaintext Attacks DEM Data Encryption Mechanism DNF Disjunctive Normal Form IND INDistinguishability KEM Key Encapsulation Mechanism KEMAC KEM with Access Control LWE Learning With Errors NIKE Non-Interactive Key Exchange ETSI ETSI TS 104 015 V1.1.1 (2025-02) 8 PKE Public-Key Encryption PK-IND Public-Key privacy INDistinguishability PPT Probabilistic Polynomial Time SK-IND Session-Key privacy INDistinguishability |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 4 Cryptographic primitives | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 4.1 Hash functions | Hash functions are used to produce a fixed length random output y from an arbitrary length input x: • H(x) → y Approved hash functions for the purpose of the present document are: • SHA-256, SHA-384, SHA-512, SHA-512/256 as defined in FIPS PUB 180-4 [4]. • SHA3-256, SHA3-384, SHA3-512 as defined in FIPS PUB 202 [5]. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 4.2 Key Encapsulation Mechanisms (KEMs) | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 4.2.1 KEMs description | A Key Encapsulation Mechanism KEM is a public-key scheme defined by three algorithms: • KEM.KeyGen(1κ) → (pk, sk): on input of a security parameter κ, returns a public key pk and a secret key sk; • KEM.Enc(pk) → (C, K): on input of the public key pk, generates a session key K, and its encapsulation C, and returns (C, K); • KEM.Dec(sk, C) → K: on input of the encapsulation C and the secret key sk, returns the session key K encapsulated in C. Correctness: A KEM is said to achieve correctness if the probability that KEM.Dec(sk,C) is not equal to K is negligible in κ, on the distribution of {(pk,sk) ←KEM.KeyGen(1κ), (C,K)←KEM.Enc(pk)}. Security: The main goal of a KEM is to encapsulate a session key K that can only be recovered from the encapsulation C with knowledge of the secret key. This is called Session Key Indistinguishability (SK-IND). One may also wish to protect the privacy of the recipient, meaning that an adversary cannot identify whether a particular encapsulation has been prepared for a specific public key. This is called Public Key Indistinguishability (PK-IND). The adversary can be modelled as having access to an encapsulation oracle (equivalently, the KEM's public key), in which case the scheme should be resistant to a Chosen Plaintext Attack (CPA security), or having additional access to a decapsulation oracle, in which case the scheme should be resistant to a Chosen Ciphertext Attack (CCA security). The adversary is not allowed to submit any challenge values to the decapsulation oracle. For a more detailed description of these properties, including the precise security games, see clause A.1. Approved KEMs for the purpose of the present document are: • ML-KEM [6]. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 9 |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 4.2.2 KEMs with Access Control (KEMAC) | When several users are in a KEM system, a KEM with Access Control (KEMAC) can issue users keys according to a key-policy Y, and encapsulate session keys with respect to an encapsulation-policy X, so that a user with key-policy Y can decapsulate if and only if R(X,Y) evaluates to 1, for a fixed Boolean rule R. Said differently, the access control is defined with respect to the rule R on policies X and Y. For any user-policy Y and encapsulation-policy X, R(X,Y) evaluates to 1 if the user with keys corresponding to the policy Y is allowed to decapsulate an encapsulation made with the policy X; else, R(X,Y) evaluates to 0. A KEMAC KEMAC is defined with the following algorithms: • KEMAC.Setup(R,1κ) → (MPK, MSK): on input of the rule R and the security parameter κ, outputs the global public parameters MPK and the master secret key MSK; • KEMAC.KeyGen(MSK, Y) → USK: on input of the master secret key MSK and the user-policy Y, outputs a user secret key USK; • KEMAC.Enc(MPK, X) → (C, K): on input of the global public parameters MPK and the encapsulation -policy X, outputs the session key K and an encapsulation C of K; • KEMAC.Dec(USK, C) → K: on input of the user secret key USK and the encapsulation C, outputs the key K encapsulated in C. Correctness. KEMAC is said to achieve correctness with respect to the rule R if for each user-policy Y and encapsulation-policy X such that R(X,Y)=1, given the security parameter κ, the distribution of user keys built with respect to Y, and of encapsulations C of keys K with respect to the policy X is such that except with probability negligible in κ, the decapsulation of C using these user keys is equal to K. Security. The challenge setup consists of chosen policies X and Y according to R, a random key pair (MPK,MSK) ← KEM.Setup (R,1κ), a random encapsulation (C,K0)←KEM.Enc(MPK, X), a random bit b, and a random key K1. For SK-IND-CPA security, given (C, Kb), no adversary, that can only ask keys for user-policies Y' such that R(X,Y')=0, can guess b with non-negligible advantage. Note that allowing key queries for a user-policy Y' such that R(X,Y')=1 would allow decapsulating C, and trivially guess b. For PK-IND-CCA security, the adversary has additional access to a decapsulation oracle, which provides the encapsulated key K for any encapsulation C' under a key USK, according to any user-policy Y', except for the challenge encapsulation C. Traceability. An optional feature of a KEMAC is offering traceability in the case of a pirate decoder in which a particular user's key has been embedded. This is a recommended but not required feature, which gives each user distinct keys even if they have common attributes. Several levels of traceability exist. The simplest one is called white-box tracing, where from the key extracted in the pirate decoder one can trace back the traitor. In this case, the KeyGen algorithm takes an additional input U, the identity of the user. Then no adversary should be able to design a decapsulating algorithm that uses a key that does not correspond to a user U. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 4.2.3 NIKE-based KEM | A Non-Interactive Key Exchange (NIKE) is defined by two algorithms: • NIKE.KeyGen(1κ) → (pk, sk): on input of a security parameter κ, returns a public key pk and a secret key sk; • NIKE.SessionKey (sk, pk') → K: on input of a secret key sk and a public key pk', generates a session key K. With the two properties: • Correctness: for any (pk0, sk0), (pk1, sk1) ← NIKE.KeyGen(1κ), NIKE.SessionKey(sk1, pk0) = NIKE.SessionKey(sk0, pk1); • Security: for (pk0, sk0), (pk1, sk1) ← NIKE.KeyGen(1κ), K ← NIKE.SessionKey(sk1, pk0), given (pk0, pk1) only, recovering K is hard. Then one can derive a KEM: • KEM.KeyGen(1κ) → (pk, sk): for (pk, sk) ← NIKE.KeyGen(1κ); ETSI ETSI TS 104 015 V1.1.1 (2025-02) 10 • KEM.Enc(pk) → (C, K): for (pk', sk') ← NIKE.KeyGen(1κ) and K ← NIKE.SessionKey(sk', pk), then C ← pk'; • KEM.Dec(sk, C) → K': for K' = H (K), with K ← NIKE.SessionKey(sk, pk'), where pk' ← C. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 4.2.4 Key-Homomorphic NIKE (KH-NIKE) | A NIKE is called key-homomorphic, if there are two internal group-laws ⊗,⊙ on the secret and the public keys that make them correspond to each other: from (pk0, sk0), (pk1, sk1) ←NIKE.KeyGen(1κ), the secret key sk ← sk0 ⊗ sk1 corresponds to the public key pk ← pk0 ⊙ pk1. So, for any scalar x, the secret key sk' ← x . sk = sk ⊗… ⊗ sk corresponds to the public key pk' ← x . pk = pk ⊙… ⊙ pk. Approved KEMs for the purpose of the present document are based on the Diffie-Hellman NIKE in a group (G, P, p), where P is a generator of G, of prime order p. The DH algorithms are: • DH.KeyGen(1κ) → (pk, sk): for sk $← ⟦1; −1⟧ and pk ← sk.P; • DH.SessionKey (sk1, pk2) → Q: where Q ← sk1.pk2. The security relies on the Computational Diffie-Hellman (CDH) problem, and it provides key homomorphism. Approved NIKEs for the purpose of the present document are the above DH scheme on elliptic curves where G is instantiated with the P-256, P-384 and P-521 [1] or the Curve25519 and Curve448 [2] elliptic curves. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 5 Hybrid Traceable KEMAC (HTKEMAC) | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 5.1 Description | After the definitions given in previous clauses, this clause specifies the KEM instantiation recommended in the present document, combining hybridization, access control and traceability, from a set Ω of rights (which are combinations of attributes, as shown below) that defines the rule: for any pair , of subsets of Ω, R , = 1 if and only and have a non-empty intersection. As already explained, and will be the encapsulation-policy and the user-policy, respectively, as lists of rights or equivalently lists of the indices of the rights in the set Ω. It makes use of a key-homomorphic NIKE (with secret keys in ⟦1; −1⟧) and a KEM with output session keys in K = 0,1 . It will use the following notations: • Ω = {, . . . , } is the set of rights, as described in clause 6; • G is a group of prime order p, in which the CDH is assumed to be hard. It will be instantiated with the P-256, P-384 and P-521 [1] or the Curve25519 and Curve448 [2] elliptic curves; • KEM is a KEM scheme achieving SK-IND-CCA and PK-IND-CCA security. It will be implemented with ML-KEM [6]; • G, H, J are hash functions, mapping elements to ⟦0; −1⟧, 256-bit strings and 384-bit string respectively where p is the order of group G, an elliptic curve field defined by the curve. They will be implemented with SHAKE [3], [5]. The algorithms are: • HTKEMAC.Setup(Ω, 1κ) → (MPK, MSK): for G a group of prime order p corresponding to the security parameter κ, and P a generator of G: - the algorithm samples (, ), (, ), (, ) ←NIKE. KeyGen(1 ); - the set of user identities ID, is initialized as an empty set, the tracing secret key is then set to: tsk ←( , , ,ID) and the tracing public key to: tpk ←, , , ; ETSI ETSI TS 104 015 V1.1.1 (2025-02) 11 - the set of users' secret keys showing their permissions is initialized as an empty set with UP ←∅; - for each right Si of index i in Ω, the algorithm samples (pk, sk) ←KEM. KeyGen(1 ), (, ) ← NIKE. KeyGen(1 ), computes ←NIKE. SessionKey( , ), and sets pki' ←(, pki) and ski' ←(, ski); - finally, the global public key is set to MPK← (tpk, {pki'}i), and the master secret key to MSK ← (tsk, {ski'}i, UP). The algorithm returns (MPK, MSK). • HTKEMAC.KeyGen(MSK, U, Y) → (USK, MSK', tsk'): on input a username U, along with Y a set of indices corresponding to U's rights in Ω, parsing the master secret key MSK as an output of the Setup algorithm: - draws ⟦1; −1⟧, sets to be the representative of ( − / ) that is in ⟦0; −1⟧, so that = (. ) ⊗( . ) and thus = (. ) ⊙( . ), and sets U's secret identifier to uid ←, ; - updates the tracing secret key by setting tsk' to be equal to tsk in which (U, uid) is added to ID; - U's secret key is defined as USK ← (uid, {sk'j}∈), and the master secret key is updated to MSK' equal to MSK in which USK was added to UP. Finally, the algorithm outputs (USK, MSK', tsk'). • HTKEMAC.Enc(MPK, X) → (C, K): parsing the public key MPK as an output of the Setup algorithm, and X as a set of indices of rights in Ω: - denoting K the key space of KEM, the encryption algorithm draws S $← K, r ← G(S), sets ←. ←. and ←, , and for each index i in X, sets ← NIKE. SessionKey(, ), (, ) ← KEM.Enc(pki), and ← ⊕(, , , { } ∈ ) - the algorithm then computes (, ) ← J, ,{, }∈ , sets the encapsulation as ←, {, }∈ , and the encapsulated key to be K. The algorithm outputs (K, C). • HTKEMAC.Dec(USK, C) → K: parsing USK as an output of the KeyGen algorithm, and = = (, ), {, }∈ , as an output of the Enc algorithm, for each index ∈ with a pair , in C, and for each index ∈ with an element skj' in USK, the decryption algorithm: - runs ′, ← KEM.Dec (skj, Ei); - computes ← NIKE. SessionKey(, (α. c) ⊙( . )); - computes , ←⨁(, ,, , { } ∈ ); it computes both r' ← G , and ,, , ← J ,, ,{, }∈ , and checks whether = . , . and ′, = , returns ← , and stops. Otherwise, it continues with the next pair (i,j). If for all indices i and j, no key was returned, the algorithm returns ⊥. When instantiated as above, the HTKEMAC hybridizes a pre-quantum NIKE scheme and a post-quantum KEM scheme to obtain the best of both their secret-key privacy: it provides CCA security for secret-key privacy under the CDH or the MLWE problems, and CCA security for access-control privacy under both the CDH and the MLWE problems. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 5.2 Parameter sets | Parameter sets consist of tuples of specific choices for the hash function SHAKE [3], [5], the ML-KEM [6], and the elliptic curves [1], [2]: • SHAKE128_P256_ML-KEM-512 ETSI ETSI TS 104 015 V1.1.1 (2025-02) 12 • SHAKE128_Curve25519_ML-KEM-512 • SHAKE256_P384_ML-KEM-768 • SHAKE256_Curve448_ML-KEM-768 • SHAKE256_P521_ML-KEM-1024 |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6 Access structure | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6.1 High-level description | The previous clause describes an access control from two subsets and of rights in associated to the encapsulations and the user's keys respectively, so that the latter can decrypt the former if and only if ∩ ≠∅. This clause explains how to transform an access structure and a Boolean policy into the set of all the possible rights and the desired subsets and . The access structure is specified by rights, which are combinations of attributes along different dimensions. For the sake of clarity, one can consider a concrete case, where there are three dimensions of attributes: • The countries CTR={EN,FR} • The departments DPT={DEV,MKG} • The security levels SEC=(LOW,MED,HIG) This defines the following qualified attributes along the 3 dimensions: CTR::FR, CTR::EN, DPT::DEV, DPT::MKG, and SEC::LOW, SEC::MED, SEC::HIG. The two first dimensions CTR and DPT are defined by unordered sets, whereas the last security level SEC is defined by an ordered set, meaning that a user with the SEC::HIG attribute also possesses the SEC::LOW and SEC::MED qualified attributes, as SEC::HIG ⇒ SEC: : MED ⇒ SEC::LOW, or equivalently SEC::LOW ≤ SEC: : MED ≤ SEC::HIG, whereas attributes within the dimensions CTR and DPT are incomparable. A right is a combination of attributes. Such a right is valid when represented as a list of attributes if it involves (some or none) attributes of different dimensions only. One can then define as the set of valid rights, that are enough to define any monotonous access structure. One can note this includes fully defined rights, such as {CTR::FR, DPT::MKG, SEC::MED}, but also partially defined rights, such as {CTR::FR}, {SEC::HIG}, or even the empty one {}, which designates a broadcast. This notation is used to optimize the encapsulation size, with several kinds of broadcasts along the different dimensions: an encapsulation with right {CTR::FR} can be decrypted by users with rights {CTR::FR, DPT::MKG, SEC::MED} or {CTR::FR, DPT::DEV}, and even more. In order for decryption to be possible, a user-right has to contain all of the attributes required by an encapsulation-right. Hence, the sets X and Y will have to be carefully derived, as explained below, to allow decryption of an encapsulation under by a user's key under if and only if ∩ ≠∅. The full ordering along some dimensions can be extended to a partial ordering between all the rights, in a trivial way: for the right to be smaller than the right , denoted ≤, one requires that along each dimension, the attribute in has to be absent, equal to or smaller (if comparable) than the attribute in . For example, {CTR::FR, SEC::MED} ≤ {CTR::FR, DPT::MKG, SEC::HIG}, and {CTR::FR} ≤ {CTR::FR, DPT::MKG}, but {CTR::FR, SEC::HIG} and {CTR::FR, DPT::MKG, SEC::MED} are incomparable. Hence, when one expresses the access control for a given encapsulation (in the Ciphertext-Policy ABE spirit), by a specific monotonous Boolean formula , it can be converted into its Disjunctive Normal Form (DNF), that is a disjunction of clauses. Clauses are exactly the above rights. The encapsulation will be associated to all the rights/clauses in the DNF, whereas the user's key will be associated to all the rights/clauses owned by the user. The following explains how to find the smallest encapsulation within the above framework. This allows the minimization of encapsulation sizes. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 13 |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6.1.1 Attributes, dimensions and hierarchies | Let a dimension be a set of attributes. Each dimension is constrained to be either: • a hierarchy , which defines a set of ordered attributes: ∀!, " ∈, ! ≤" or " ≤! • an anarchy #, which defines a set of incomparable attributes: ∀!, " ∈#, ! = " or (¬ ! ≤") and ¬" ≤! A Boolean policy can then be described as a disjunction and/or conjunction of attributes. Each dimension is further constrained to be uniquely named, and each attribute has to have a unique name across its dimension. Then, the following context-free grammar can be used to parse an access policy: <access-policy> = <qualified-attribute> | <access-policy> <op> <access-policy> <qualified-attribute> = <dimension>::<attribute> <dimension> = <name> <attribute> = <name> <op> = `&&' | `||' <name> = [^ `&&' `||' `::'] where [^ ‘&&’ ‘||’ ‘::’] stands for any non-empty combination of characters (name) in which neither ‘&&’, ‘||’ nor ‘::’ occur. 6.1.2 Spatial representation Without loss of generality, a right can be defined as a conjunction of attributes belonging to different dimensions. The universe can then be defined as the set of all those conjunctions. Moreover, there is a one-to-one correspondence between an access policy on an access structure with dimensions, and a set of points in a -dimensional space. Each right is associated to a point in the -dimensional space: • the associated point of the right is the only point that belongs to this access policy, but in no further restriction of it. EXAMPLE: One can associate rights to points, by associating attributes to a specific integer along the dimension. The absence of an attribute in a dimension is then represented by 0: - CTR::EN && DPT::DEV to 1,1; - CTR::EN && DPT::MKG to 1,2; - CTR::FR && DPT::DEV to 2,1; - CTR::FR && DPT::MKG to 2,2; - CTR::EN to 1,0; - CTR::FR to 2,0; - DPT::DEV to 0,1; - DPT::MKG to 0,2. The total number of rights in a 2-dimensional universe is therefore 8 and becomes 9 if one adds a right that maps to the origin 0 , corresponding to the global broadcast, denoted right BROADCAST. These points will be used for an encapsulation-policy . On the other hand, for a user-policy, they will be expanded for efficiency reasons, in order to represent the Boolean policy on rights in the key. This allows the use of the smaller of the two sets for encapsulation generation, which minimizes the size. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 14 One can furthermore define a partial ordering on the rights and/or points. This was explained above for rights and can be translated as follows for points: • For the rights: ≤, if and only if for each dimension, the attribute in shall be absent, equal to or smaller, for a hierarchical dimension, than the attribute in ; • For the points: = $ ≤ = % if and only if for each dimension : $ = 0 or $ ≤% for a hierarchical dimension. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6.2 Efficiency considerations | As explained above, a Boolean formula expresses the access control for a given encapsulation. After the DNF conversion, this leads to the set with all the clauses (or points associated to the rights) that appear in the DNF. However, building the set associated to the attributes for the user's key is more complex. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6.2.1 Encapsulation rights | Once converted into its DNF, the Boolean formula associated to an encapsulation is a list of rights from . This is the set . |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6.2.2 User's key rights | Two additional spaces have to be defined to build the set : • the semantic space &'( , of a right is the subspace of points in which the attributes involved in this right can be expressed. EXAMPLE: In the 3-dimensional space, the semantic space of CTR::EN, associated to 1,0,0, is the 1-dimensional subspace of generated by the CTR dimension: < 1,0,0 >. The semantic space of the right CTR::FR && DPT::DEV && SEC::MED, associated to 2,1,2, is the whole space = < 1,0,0, 0,1,0, 0,0,1 >. • the complementary space )*(+ , of a right is defined as )*(+ , = { , ≤} + (Ω ∖ &'( , ). EXAMPLE: The complementary space of CTR::EN is the 2-dimensional space < 0,1,0, 0,0,1 > generated by DPT and SEC which origin is 1,0,0: {1,0,0, 1,1,0, 1,1,2, 1,1,3, 1,2,0, 1,2,1, 1,2,2, 1,2,3}. In case of a hierarchical dimensions, one has to consider all the origins that are implied by the right : the complementary space of CTR::FR && DPT::DEV && SEC::MED is the combination of the 0-dimensional spaces which origins are 2,1,1, 2,1,2: {2,1,1, 2,1,2}. As a final example, the complementary space of CTR::FR && SEC::MED is the combination of the 1-dimensional spaces < 0,1,0 > generated by DPT which origins are 2,0,1, 2,0,2: {2,0,1, 2,1,1, 2,1,2, 2,0,2, 2,1,2, 2,2,2}. In the user-policy set , all the points in all the subsets associated to the rights are concatenated, plus the BROADCAST, associated to the origin point 0 , that is 0,0,0 in the 3-dimensional case. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6.2.3 Cardinality of an encapsulation | Following the previous description, the number of rights used in an encapsulation should be equal to the number of clauses in the DNF of its associated access policy, since each clause is either a broadcast to a subspace of , or to a singleton. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 6.2.4 Cardinality of a user secret key | The number of rights associated to a user secret key is the number of points in the complementary spaces generated by the rights associated to the user. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 15 |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7 Specification | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.1 Introduction | This clause specifies all the objects and functions needed to implement Covercrypt, with their input/output types, where the brackets (such as [* Name]) stand for optional components. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.2 Access Structure | |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.2.1 Type | An AccessStructure is a set of dimensions, where a Dimension is an object that holds its name and attributes, which are themselves composed of their name and an id that is unique across this access structure: AccessStructure = Set Dimension Dimension = Hierarchy (Name * OrderedSet (Id * Name)) | Anarchy (Name * Set (Id * Name)) Then a right is uniquely associated to each set of IDs from an access structure: Right = Set Id |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.2.2 API | An object of type AccessStructure should expose the following API: • ap_to_usk_rights (AccessStructure * AccessPolicy) -> Set Right Generates the set of USK rights described by the given access policy. • ap_to_enc_rights (AccessStructure * AccessPolicy) -> Set Right Generates the set of ciphertext rights described by the given access policy. • add_anarchy (AccessStructure * Name) -> AccessStructure Adds an anarchic dimension with the given name to the access structure. Requires USK refresh: On addition of a dimension, users will need to refresh keys in order to decrypt new encapsulations with access policies including the new dimension. • add_hierarchy (AccessStructure * Name) -> AccessStructure Adds a hierarchic dimension with the given name to the access structure. Requires USK refresh: On addition of a dimension, users will need to refresh keys in order to decrypt new encapsulations with access policies including the new dimension. • del_dimension (AccessStructure * Name) -> AccessStructure Removes the dimension with the given name from the access structure. Requires USK refresh: On deletion of a dimension, users that refresh their keys will lose the ability to decrypt old encapsulations with access policies including the old dimension. • add_attribute (AccessStructure * QualifiedAttribute [* Name]) -> AccessStructure Adds the given qualified attribute to the access structure. If the dimension is hierarchical, specifying the name of an existing attribute of the same dimension sets the rank of the new attribute to be in-between this existing attribute and the next attribute if any. Gives the new attribute the lowest rank in case no such attribute name is specified. If this name does not match any valid attribute, an error is returned. Specifying an attribute name when adding an attribute to an anarchy has no effect. Requires USK refresh: On addition of an attribute, users will need to refresh keys in order to decrypt new encapsulations with access policies including the new attribute. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 16 • del_attribute (AccessStructure * QualifiedAttribute) -> AccessStructure Removes the given qualified attribute from the access structure or returns an error if this attribute does not belong to this access structure. Requires USK refresh: On deletion of an attribute, users that refresh their keys will lose the ability to decrypt old encapsulations with access policies including the old attribute. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.3 Master Secret Key | The master secret key has the following type: MSK = AccessStructure * Keypairs [* SigningKey] [* Tracers] where: • the AccessStructure is described above and contains all necessary information to validate and translate an access structure into a set of rights; • the Keypairs structure contains the history of the keypairs associated to each right. This includes old keys that are no longer in use for encryption, because encapsulations are not necessarily re-encrypted after a key refresh; • the SigningKey is used to produce publicly verifiable signatures of the user secret keys and master public keys; • the Tracers is used to allow tracing user keys. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.4 Master Public Key | A master public key has the following type: MPK = Version * AccessStructure * (Right -> PublicKey) [* Signature] [* TracingPoints] where: • Version holds a number that is increased at each generation of a new master public key; • AccessStructure holds all necessary information for validating and translating an access structure into a set of rights; • Right -> PublicKey maps each right to the latest version of its associated public key; • Signature holds a publicly verifiable signature generated by the master secret key; • TracingPoints holds public points required in the encapsulation process if the master secret key defines tracers. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.5 User Secret Key | A user secret key has the following type: USK = SecretKeys [* Signature] [* UserId] where: • SecretKeys contains the history of the secret keys associated to each right that have been given to the user; • Signature holds a publicly verifiable signature generated by the master secret key; • UserID contains a unique identifier required for opening encapsulations if the master secret key defines tracers. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 17 |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.6 Encapsulation | An encapsulation has the following type: XEnc = List Encapsulation * Tag [* Traps] [* Version] where: • List Encapsulation contains the encapsulation associated to each right, in a cryptographically-random order. Using a List allows serialization of this order, which is needed since it matters when recomputing the Tag; • Tag is the early-abort tag that allows determination of whether opening an encapsulation was successful; • Traps contains public points required in the opening process if the master secret key defines tracers; • Version holds the version number of the master public key used to generate this encapsulation. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 7.7 Covercrypt | A Covercrypt implementation exposes the following API: • mk_gen (AccessStructure) -> MSK * MPK It takes as input the access structure and generates new master secret and public keys: omega ← ap_to_usk_rights(AccessStructure, "*") (msk, mpk) ← HTKEMAC.setup(omega) • usk_gen (MSK * U * AccessPolicy) -> USK It takes as input the master secret key, a user and an access policy and generates a new user secret key that holds secret keys as follows, or signals an error if this access policy is invalid w.r.t. this master secret key: y ← ap_to_usk_rights(msk.AccessStructure, ap) usk ← HTKEMAC.keygen(msk, u, y) • enc (MPK * AccessPolicy) -> (K * XEnc) It takes as input a master public key and an access policy and generates an encapsulation of a random session key for the rights returned by ap->enc-rights or signals an error if this access policy is invalid w.r.t. this master public key: x ← ap_to_enc_rights(mpk.AccessStructure, ap) (k, c) ← HTKEMAC.enc(mpk, x) • dec (USK * XEnc) -> K It takes as input a user secret key associated to an access policy ! and an encapsulation associated to an access policy !, and returns a session key if ! ∩! ≠∅ and this user secret key holds the correct version of at least one of secret key associated to a right in this intersection: k ← HTKEMAC.dec(usk, c) • rotate (MSK * MPK * AccessPolicy) -> MSK * MPK Generates new keypairs for all rights given by ap_to_usk_rights. All user secret keys need to be refreshed, and older ciphertexts may be re-encrypted. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 18 • refresh (MSK * USK) -> USK It takes as argument the master secret key and a user secret key and return an updated version of this user secret key that holds each newer version of the secret keys it previously held. |
08c7c12d92f08acc1ca41679e7fb8a4f | 104 015 | 8 Conclusion | The present document shows the construction of a KEM which achieves privacy and correctness properties, while allowing the encapsulation of keys with respect to hidden access policies. As shown in [i.1], this scheme allows an order of magnitude speedup with respect to Attribute-Based Encryption implementations in practical use-cases in which the access-policies can be expressed with less than ten logical gates. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 19 Annex A (informative): Security Definitions A.1 KEM Security Definitions Session-Key Privacy. KEM is said to achieve session-key privacy (denoted SK-IND or SK-IND-CPA) if session keys generated by the encapsulation algorithm are indistinguishable from elements taken uniformly at random in the session-key space K. In other words, KEM is SK-IND secure if and only if for any PPT adversary A, A has a negligible advantage in the security parameter κ in distinguishing between distributions D0 and D1 defined for a bit b with: D = , -., /. ←KEM. 1 , , ←KEM. , $← $ K : , , In the SK-IND-CCA security game, the adversary is given access to a KEM. Dec oracle, except for the challenge ciphertext. Public-Key Privacy. Defined analogously to the anonymity of a PKE scheme, the public-key privacy (denoted PK-IND or PK-IND-CPA) of an encapsulation scheme similarly states the outputs of encapsulations using one or the other of two output public keys of the KeyGen algorithm will be indistinguishable except with probability negligible in the security parameter κ. More formally, for any PPT adversary A, defining for a bit b: D = For = 0,1: , ←KEM. 1 , , ←KEM. : , , , The advantage of A in distinguishing between D0 and D1 is negligible in κ. In the PK-IND-CCA security game, the adversary is given access to a KEM. Dec oracle, except for the challenge ciphertext. A.2 KEMAC Security Definitions Correctness. Formally, if defining the distribution D and event Ev as: D = ⎩ ⎨ ⎧∀, such that ℛ, = 1, MP , ←KEMAC. 1 , ←KEMAC. , , , ←KEMAC. , : , , ⎭ ⎬ ⎫ = !KEMAC. " , = #. The probability that Ev happens on the distribution D is overwhelming in κ then the KEMAC is said to be correct. Session-Key Privacy. KEMAC is said to be session-key private (denoted SK-IND or SK-IND-CPA) if any PPT adversary A provided with the public key MPK, and with an oracle access to the KeyGen algorithm for a set of user policies Y, choosing an encapsulation-policy X such that for each user-policy Y in Y, R(X,Y)=0, and given an encapsulation C of MPK under the policy X, can distinguish between the session-key encapsulated in C and a random element of the key-space K only with probability negligible in κ. In the SK-IND-CCA security game, the adversary is given access to a KEMAC. Dec oracle, except for the challenge ciphertext. In addition to avoiding leaking information about encapsulated session-keys to non-recipients, a scheme with this property will also not reveal any information about the policies used, which will be granted by the access-control privacy property defined hereafter. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 20 Access-Control Privacy. KEMAC is said to be access-control private (denoted AC-IND-CPA) if in the same setting as in the session-key privacy security game, any PPT adversary A choosing two encapsulation policies X0 and X1 for which she cannot have user keys enabling decapsulation, cannot distinguish between encapsulations using one or the other policy. In the AC-IND-CCA security game, the adversary is given access to a KEMAC. Dec oracle, except for the challenge ciphertext. Traceability. Let A be a PPT adversary that can: • ask for the generation of user keys through an oracle OKeyGen(·) taking usernames as inputs, and, on input U, running USKU ← KEMAC.KeyGen(MSK, U) and adding U and the corresponding key USKU to the system; • corrupt registered users with access to an oracle OCorrupt(·), taking as input a username U, adding it to the system if it is not in it yet, as well as in the list of traitors T, and returning U's secret key USKU to A. A KEMAC is called white-box traceable if from the key extracted in a pirate decoder one can get the identity of the traitor. ETSI ETSI TS 104 015 V1.1.1 (2025-02) 21 History Document history V1.1.1 February 2025 Publication |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 1 Scope | The present document defines the PEMEA File Exchange (PFE) capability, and the need for this functionality. The required entities and actors are identified along with the protocol, specifying message exchanges between entities. The message formats are specified and procedural descriptions of expected behaviours under different conditions are detailed. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 2 References | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 2.1 Normative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. Referenced documents which are not found to be publicly available in the expected location might be found at https://docbox.etsi.org/Reference/. NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee their long term validity. The following referenced documents are necessary for the application of the present document. [1] ETSI TS 103 478: "Emergency Communications (EMTEL); Pan-European Mobile Emergency Application". [2] IETF RFC 2617 (June 1999): "HTTP Authentication: Basic and Digest Access Authentication". [3] IETF RFC 6750 (October 2012): "The OAuth 2.0 Authorization Framework: Bearer Token Usage". [4] IETF RFC 6838 (January 2013): "Media Type Specifications and Registration Procedures". [5] IETF RFC 7578 (July 2015): "Returning Values from Forms: multipart/form-data". [6] IETF RFC 8089 (February 2017): "The "file" URI Scheme". [7] IETF RFC 9110 (June 2022): "HTTP Semantics". [8] IETF RFC 9112 (June 2022): "HTTP/1.1". [9] ISO 8601-1:2019: "Date and time - Representations for information interchange - Part 1: Basic rules". [10] WHATWG: "HTML Living Standard". |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 2.2 Informative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee their long term validity. The following referenced documents are not necessary for the application of the present document but they assist the user with regard to a particular subject area. [i.1] ETSI TS 103 756: "Emergency Communications (EMTEL); PEMEA Instant Message Extension". ETSI ETSI TS 104 014 V1.1.1 (2024-07) 8 [i.2] ETSI TS 103 871: "Emergency Communications (EMTEL); PEMEA Real-Time Text Extension". [i.3] ETSI TS 103 945: "Emergency Communications (EMTEL); PEMEA Audio Video Extension". [i.4] IETF RFC 7519 (May 2015): "JSON Web Token (JWT)". |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 3 Definition of terms, symbols and abbreviations | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 3.1 Terms | For the purposes of the present document, the following terms apply: security: techniques and methods used to ensure: • authentication of entities accessing resources or data; • authorization of authenticated entities prior to accessing or obtaining resources and/or data; • privacy of user data ensuring access only to authenticated and authorized entities; • secrecy of information transferred between two authenticated and authorized entities; • trusted is used as defined in ETSI TS 103 478 [1]. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 3.2 Symbols | Void. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 3.3 Abbreviations | For the purposes of the present document, the following abbreviations apply: AEAD Authenticated Encryption with Associated Data AES Advanced Encryption Standard AESGCM Advanced Encryption Standard key used with GCM AP Application Provider App Application CPE Customer Premises Equipment DHE Diffie-Hellman key Exchange ECDHE Elliptic-Curve Diffie-Hellman key Exchange EDS Emergency Data Send (message) GCM Galois/Counter Mode ISO International Organization for Standardization JSON JavaScript Object Notation JWT JSON Web Token MAC Message Authentication Code MIME Multipurpose Internet Mail Extensions OAuth Open Authorization Pa PEMEA Application to AP interface PEMEA Pan-European Mobile Emergency Application PFE PEMEA File Exchange PIM PSAP Interface Module PSAP Public Safety Answering Point PSP PSAP Service Provider RFC Request For Comments RSA Rivest Shamir Adleman public key encryption algorithm SSE Server-Sent Events TLS Transport Layer Security ETSI ETSI TS 104 014 V1.1.1 (2024-07) 9 TS Technical Specification tPSP terminating PSP URI Uniform Resource Identifier URL Uniform Resource Locator UTC Universal Time Coordinated UTF-8 8-bit Unicode Transformation Format WHATWG Web Hypertext Application Technology Working Group |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 4 PEMEA capability extensions | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 4.1 Overview of extension in PEMEA | PEMEA extension capabilities are defined in ETSI TS 103 478 [1] and are implemented through the use of "reach- back" URIs. The Application Provider (AP) node advertises capabilities as part of the initial forward message through the network, the Emergency Data Send (EDS) message, and the terminating PSAP Service Provider (PSP) or PSAP responds with the subset of capabilities that it supports, thus binding the emergency session between the AP and the terminating PSP or PSAP node. Specifically, the capabilities are sent as information elements in the apMoreInformation element of the EDS message. The information element and apMoreInformation structures are defined in clauses 10.3.11 and 10.3.12 of ETSI TS 103 478 [1]. An information element in a PEMEA EDS message identifies a capability and each capability is made up of three distinct parts: • typeOfInfo: what function does the information element serve; • protocol: the specific semantics for using the function; • value: the URI through which the service is invoked. Table 10 in ETSI TS 103 478 [1] identifies an initial set of "typeOfInfo" values used to specify a range of capability extensions for PEMEA. However, beyond the Location_Update and SIP_Request values described in Table 11 of ETSI TS 103 478 [1], protocols are left for further study and definition in subsequent specifications such as the present document. ETSI TS 103 756 [i.1] defines the concrete specification for PEMEA Instant Message protocol, ETSI TS 103 871 [i.2] defines the concrete specification for PEMEA Real-Time Text and ETSI TS 103 945 [i.3] defines the concrete specification for PEMEA Audio Video. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 4.2 Service support indication and response | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 4.2.1 Service definition | The present document provides a concrete definition of the "File_Exchange" typeOfInfo in PEMEA through the present document of a protocol value. The definition in Table 1 shall be considered as an extension to Table 11 in ETSI TS 103 478 [1]. Table 1: Extended AP Information Type Protocol Registry Info type Value Protocol Token Description File_Exchange PEMEA File exchange functionality is supported using the PFE protocol |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 4.2.2 Service support indication | An AP needing to indicate that the Application it is serving can support file exchange using the PEMEA protocol would include the following information element in the apMoreInformation element of the EDS associated with the emergency session: <information typeOfInfo="File_Exchange" protocol="PEMEA"> https://ap.example.pemea.help/37agq1cyusbo </information> ETSI ETSI TS 104 014 V1.1.1 (2024-07) 10 |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 4.2.3 Service support response | A terminating node that can support the "File_Exchange" "PEMEA" capability includes this capability in the apMoreInformation element returned to the AP in the onCapSupportPost, as defined in clause 11.1.4 of ETSI TS 103 478 [1], with the value for "File_Exchange" "PEMEA" provided in the example below. <apMoreInformation xmlns="urn:pemea:apps:xml:ns:pemea:base"> <information typeOfInfo="File_Exchange" protocol="PEMEA"/> </apMoreInformation> |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 5 Architecture | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 5.1 Overview | The PEMEA File Exchange (PFE) capability defines a protocol to exchange files between Apps and PSAPs with a PEMEA emergency request. In order to exchange the files, a file server is needed, and the present document defines all the interfaces and procedures that the file server shall provide. This also helps to understand explicitly what is normatively specified in the present document, what is semantic, and what is normatively referred to from the present document but normatively specified in other documents. The PFE capability was realized with disability usage in mind and the usage model for this necessitates multi-party communications where the Caller and the PSAP Call-Taker represent two parties, but other third-party user may also participate in the emergency session. Where the procedures described in clause 5 refer to the PSAP Interface Module (PIM) PEMEA node, they may be performed by a terminating PSP or by a PIM depending on architectural decisions in PSAP infrastructure according to ETSI TS 103 478 [1]. References in clause 5 to the App may be made by Apps directly or by an App server acting as a proxy. Whether there are Apps directly connected to the AP or there is an App server connected to the AP is beyond the scope of the present document and depends on the implementation details. The File Exchange Server shall accept requests from any source as long as they are made with the authentication token that was delivered to the AP node as described in clause 0 and following all security procedures described in clause 6. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 5.2 Architecture and high-level flows | PEMEA is structured around the AP being the gateway between the App and the PSAP. This model requires communications to occur, first between the App and the AP over the proprietary Pa interface and then between the AP and the associated PSAP service using the protocol mechanisms defined in the specific extension capability document, such as the present document or other PEMEA extension documents, e.g. ETSI TS 103 756 [i.1], ETSI TS 103 871 [i.2], ETSI TS 103 945 [i.3] or any other related one. For most services, this approach is fine, however, exchanging files requires high bandwidth, thus this approach may need to be relaxed somewhat to ensure that the capability delivers the required functionality. Therefore, the present document does not require that all requests to the File Exchange Server are done from the AP node, Apps can make requests to the File Exchange Server directly once they have received the URI and the token from the AP. The present document does not explicitly define the security measures that shall be taken at the File Exchange Server to detect malicious files, but it is highly recommended that all files uploaded to the File Exchange Server are analysed with malware detection software. Figure 1 depicts the PFE service architecture. It includes a malware detection software that is not mandatory but highly recommended. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 11 Figure 1: File Exchange architecture for PEMEA A. Pa interface, the application makes a call to the AP indicating the PFE capability. Invocation information is returned to the App over this interface too. B. The AP packages the information from the App into an EDS message and sends it into the PEMEA network via the Ps interface. The EDS arrives at the PIM over the Pp interface. The PIM sends an onCapSupportPost message to the AP binding the connection between the AP and the PIM. C. The PIM notifies the PSAP-CPE which in turn notifies the PSAP Call-Taker. The call is answered and controlled by the PSAP Call-Taker over this interface also. This includes requesting the creation of a PFE session. The PSAP Call-Taker is also able to request connection credentials for additional participants over this interface. D. On direction from the PSAP Call-Taker the PIM creates a new PFE session and requests Bearer tokens from the Security Token Management system. It shall generate at least a token for the PSAP Call-Taker and a token for the Caller. E. The PIM invokes the PFE capability in the AP passing the URI for the PFE session as well as the Bearer token required to access it. Similar information is provided to the PSAP Call-Taker's application. The AP then passes the received information down to the App. F. The App and the PSAP Call-Taker can make requests to the File Exchange Server using the PFE session URI and passing in their respective security tokens. Upload and download of files occurs over this interface. It is highly recommended that files received through this interface are scanned with anti-malware systems to prevent malicious files from being uploaded. G. The File Exchange Server verifies the Bearer tokens with the Security token management system before accepting requests from the participants. Whilst the division of the File Exchange Server in these sub-components does not need to be strictly followed, the notion of a signalling component and a media or streaming component are important for traffic path differentiation. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 12 |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 6 Security | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 6.1 Transport security | The PEMEA File Exchange (PFE) session is identified as an HTTPS URI. The requests should be made using TLS 1.3 but may be made using 1.2 and this shall not support fallback below TLS 1.2. All requests shall authenticate to the File Exchange Server using a Bearer token in the HTTP Authentication header field as defined in IETF RFC 6750 [3]. The token should be provided to the recipient when the service is invoked. The PFE session is expected to remain open while the entity is "online", that is while the EDS session remains active in the PSAP. The lists for the TLS 1.3 and TLS 1.2 acceptable cipher suites are included in Annex B. These lists are informative and are based on best information at the time of writing. Older cipher suites not included in either of these lists shall not be used. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 6.2 Security token usage | The HTTP Authorization header field is defined in IETF RFC 2617 [2] and it specifies that the usage is a scheme followed by a value, where the value may have a structure, as is the case for the digest authentication scheme. Security token usage in the HTTP Authorization header field was originally specified for use with OAuth and is defined in IETF RFC 6750 [3]. Here the use of the OAuth "Bearer token" is specified so the scheme of the Authorization header field is Bearer, following the scheme a token is placed. The token shall not contain readable information unless it is encoded in base64. Token usage in the PFE specification follows the Bearer scheme defined in IETF RFC 6750 [3]. Tokens issued by entities in the PFE architecture are expected also to be the validating entities, or to have ties to the validating entities, consequently, whether the tokens are opaque or follow a convention such as JSON Web Token (JWT) IETF RFC 7519 [i.4] is not considered relevant to usage and is not specified further. IETF RFC 6750 [3] mandates the usage of TLS for use with Bearer tokens, this usage is further defined in clause 6.1 of the present document. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7 Procedures and signalling | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.1 Overview | In order to be able to exchange files using PEMEA File Exchange (PFE) capability the App, AP, PIM and PSAP have to follow the procedures and signalling specified in clause 7 of the present document. Where the procedures defined in clause 7 refer to the PSAP Interface Module (PIM) PEMEA node, they may be performed by a terminating PSP or by a PIM depending on architectural decisions in PSAP infrastructure according to ETSI TS 103 478 [1]. References in clause 7 to the App may be made by Apps directly or by an App server acting as a proxy. The File Exchange Server shall accept requests from any source as long as they are made with the authentication token that was delivered to the AP node as described in clause 0 and following all security procedures described in clause 6. When clause 7 refers to participants, it refers not only to the App or the PSAP Call-Taker, but also to other participants that are added to the session as indicated in clause 8. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 13 |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.2 Service invocation | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.2.1 Service invocation procedures | Once the PIM has responded to the AP that it can support the PEMEA File Exchange (PFE) service with the procedures defined in clause 4.2, then the AP shall be capable of accepting a service invocation on the provided URI at any time. The AP shall only accept a PFE service invocation from the PIM that sent the onCapSupportPost message. Figure 2 provides the sequence diagram illustrating the flow of events between entities for the invocation procedure. Figure 2: PFE Invocation sequence diagram 1) The App initiates an emergency session with the AP over the Pa interface indicating that it can support the PFE capability. 2) The AP creates an EDS message from the data provided by the App and includes the PFE capability. The AP sends the EDS into the PEMEA network. 3) The EDS arrives at the PIM. The PIM supports the PFE capability and includes this option in the onCapSupportPost back to the AP. 4) The AP binds the emergency session to the PIM that sent the onCapSupportPost message and then signals to the App over the Pa interface the PSAP can support the PFE functionality. 5) The PIM notifies the PSAP Call-Taker that a new EDS has arrived. 6) The PSAP Call-Taker requests the PIM to initiate the creation of a File Exchange session. 7) The PIM requests that the File Exchange Server to create a File Exchange session. The session and 2 tokens are created. 8) The File Exchange Server returns the File Exchange session URI to the PIM. 9) The PIM returns the File Exchange session URI, one token and its expiry time to the PSAP Call-Taker. 10) The PIM invokes the PFE capability in the AP using the provided reach-back URI from the EDS. The PIM includes the File Exchange session URI, token and expiry time in the body of the HTTP POST to the reach- back URI. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 14 11) The AP signals to the App over the Pa interface that the PSAP has invoked the PFE capability and provides the File Exchange session URI, token and expiry. Once the PFE service is invoked, the operations specified in clause 7.3 can be performed using the PFE session URI and the security token. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.2.2 Service invocation object | The PIM invokes the PEMEA File Exchange (PFE) service in the AP by posting to the URI provided in the File_Exchange information element included in the apMoreInformation contained in the EDS. The POST message includes a body containing a JSON object. The JSON object provides the File_Exchange session URI as well as a security token and corresponding expiry time. The JSON schema for the PFE service invocation message is provided in Annex A. Table 2: PFE service invocation message Property Type Description url String The URI of the File Exchange session. token String A security token used to authenticate the App to the File Exchange session. The App shall include the token in the HTTP Authorization header using the Bearer token scheme. The App shall use the token in all requests to the File_Exchange session for the duration of the App emergency session. expiry Integer Specifies the expiry time of the security token. It is an integer specifying the number of second since UTC epoch, 00:00:00 1st of January 1970. Invocation example: { "url":"https://file_exchange_server.example.com/session/534wafds21s21fdf", "token":"PPtzs5zzG5Pkf61KPz51", "expiry":1574092280231 } |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.2.3 File Exchange session creation and deletion | The File Exchange session is created by the File Exchange Server under direction of the PSAP Call-Taker via the PIM as defined in clause 7.2. Once the File Exchange session is created it remains active as long as the PIM maintains a context for the EDS. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3 File Exchange operations | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.1 Overview | The File Exchange session provides a RESTful interface to perform the operations defined in clause 7.3. All the operations shall be authenticated as defined in clause 6. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.2 List files | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.2.1 Description | Participants can list the available files in the File Exchange session. This allows to check what files had been uploaded to the File Exchange session. To list the available files the participants shall make an HTTP GET request to the URI of the File Exchange session. The HTTP request shall have the following headers: • Accept: The value shall be "application/json". ETSI ETSI TS 104 014 V1.1.1 (2024-07) 15 • Authorization: The token to access the File Exchange session as defined in clause 6.2. This is obtained as defined in clause 7.2. The HTTP request body shall be empty. The HTTP response shall have the following header: • Content-Type: The value shall be "application/json; charset=utf-8". The HTTP response shall have a 200 OK status code. The HTTP response body shall contain the list of files represented as a JSON array of FileMetadata elements which structure is defined in clause 10.2.1. Figure 3 provides the sequence diagram illustrating the flow of events between entities for the list files operation. Figure 3: PFE List files sequence diagram 1) To obtain the list of files of the File Exchange session the App makes an HTTP GET request to the File Exchange session URI. The request contains the Accept header with value "application/json" and the Authorization header with the "Bearer token" as defined in clause 6.2. 2) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The file Exchange Server answers the HTTP GET request from the App with a 200 OK status code, with the Content-Type header with value "application/json; charset=utf-8" and with the body containing the list of files in the File Exchange session represented as a JSON array of FileMetadata elements. 3) To obtain the list of files of the File Exchange session the PSAP Call-Taker makes an HTTP GET request to the File Exchange session URI. The request contains the Accept header with value "application/json" and the Authorization header with the "Bearer token" as defined in clause 6.2. 4) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The file Exchange Server answers the HTTP GET request from the PSAP Call-Taker with a 200 OK status code, with the Content-Type header with value "application/json; charset=utf-8" and with the body containing the list of files in the File Exchange session represented as a JSON array of FileMetadata elements. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 16 |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.2.2 Example | HTTP request: GET /session/534wafds21s21fdf HTTPS/1.1 Host: file_exchange_server.example.com Accept: application/json Authorization: Bearer PPtzs5zzG5Pkf61KPz51 HTTP response: HTTPS/1.1 200 OK Content-Type: application/json; charset=utf-8 [ { "name":"example.png", "size":328049, "timestamp":1574092280231, "type":"image/png", "url":"https://file_exchange_server.example.com/session/534wafds21s21fdf/example.png" } ] |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.3 Upload file | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.3.1 Description | Participants can upload a file to the File Exchange session. This operation adds a new file to the files of the File Exchange session. To upload a file the participants shall make an HTTP PUT request to the URI of the File Exchange session. The HTTP request shall have the following headers: • Accept: The value shall be "application/json". • Authorization: The token to access the File Exchange session as defined in clause 6.2. This is obtained as defined in clause 7.2. • Content-Type: The value shall be multipart/form-data with the boundary that delimits the end of the file as defined in clause 4 of IETF RFC 7578 [5]. The HTTP request body shall be a multipart/form-data body as defined in IETF RFC 7578 [5] with the following considerations: • It shall have only 1 part that starts and ends with a boundary according to IETF RFC 7578 [5]. • The part shall have the Content-Disposition header field. The value shall have the following format: "form- data; name="attachment"; filename="example.png"". The name of the field shall be "attachment" and the file name shall be placed in the filename field like "example.png" is in the example. • The part shall have the Content-Type header field. The value shall be a valid MIME type as defined in clause 4.5 of IETF RFC 6838 [4]. If the file is too big to fit in one single HTTP request, the file shall be split in multiple HTTP requests with the same structure and the limit of the file shall be delimited by the boundary property as defined in clause 4.1 of IETF RFC 7578 [5]. If the File Exchange session receives multiple files with the same filename, the files shall be named by the File Exchange session according to the following rule: • If there is a file in the File Exchange session with the same filename provided in the upload file operation, the filename for the new file shall have "(1)" appended to it. • Each time that a file is uploaded with the same filename, the number shall be increased by 1. That is, the second file shall have "(2)" appended to it, the third file shall have "(3)" appended to it, etc. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 17 The HTTP response shall have the following headers: • Content-Type: The value shall be "application/json; charset=utf-8". • Location: The value shall be the URI to download the uploaded file as defined in clause 10.2.2 of IETF RFC 9110 [7]. The HTTP response shall have a 201 Created status code. The HTTP response body shall contain the file represented as a FileMetadata element which structure is defined in clause 10.2.1. Figure 4 provides the sequence diagram illustrating the flow of events between entities for the upload file operation. Figure 4: PFE Upload file sequence diagram 1) To upload a file to the File Exchange session the App makes an HTTP PUT request to the File Exchange session URI. The request contains the Accept header with value "application/json", the Authorization header with the "Bearer token" as defined in clause 6.2 and the Content-Type header with value "multipart/form-data" with the boundary that delimits the end of the file as defined in IETF RFC 7578 [5]. 2) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The file Exchange Server answers the HTTP PUT request from the App with a 201 Created status code, with the Content-Type header with value "application/json; charset=utf-8" and with the body containing the uploaded file represented as a FileMetadata data structure. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 18 3) To upload a file to the File Exchange session the PSAP Call-Taker makes an HTTP PUT request to the File Exchange session URI. The request contains the Accept header with value "application/json", the Authorization header with the "Bearer token" as defined in clause 6.2 and the Content-Type header with value "multipart/form-data" with the boundary that delimits the end of the file as defined in IETF RFC 7578 [5]. 4) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The file Exchange Server answers the HTTP PUT request from the PSAP Call-Taker with a 201 Created status code, with the Content-Type header with value "application/json; charset=utf-8" and with the body containing the uploaded file represented as a FileMetadata data structure. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.3.2 Example | HTTP request: PUT /session/534wafds21s21fdf HTTPS/1.1 Host: file_exchange_server.example.com Accept: application/json Authorization: Bearer PPtzs5zzG5Pkf61KPz51 Content-Type: multipart/form-data; boundary=----WebKitFormBoundaryDRKRUlAriyYJ28VG ------WebKitFormBoundaryDRKRUlAriyYJ28VG Content-Disposition: form-data; name="attachment"; filename="example.png" Content-Type: image/png <binary> ------WebKitFormBoundaryDRKRUlAriyYJ28VG-- HTTP response: HTTPS/1.1 201 Created Content-Type: application/json; charset=utf-8 Location: https://file_exchange_server.example.com/session/534wafds21s21fdf/example%20(1).png { "name":"example (1).png", "size":328049, "type":"image/png", "timestamp":1715858118715, "url":"https://file_exchange_server.example.com/session/534wafds21s21fdf/example%20(1).png" } |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.4 Download file | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.4.1 Description | Participants can download a file from the File Exchange session. To download a file the participants shall make an HTTP GET request to the URL of the file they want to download. The URL is obtained using the list files or the upload file operations defined in clauses 7.3.2 and 7.3.3. The HTTP request shall have the following header: • Authorization: The token to access the File Exchange session as defined in clause 6.2. This is obtained as defined in clause 7.2. The HTTP request body shall be empty. The HTTP response shall have the following headers: • Content-Disposition: The value shall have the following format: "attachment; filename="example (1).png"" where example (1).png is the name of the file that is being downloaded. • Content-Type: The value shall be a valid MIME type as defined in clause 4.5 of IETF RFC 6838 [4]. The HTTP response shall have a 200 OK status code. The HTTP response body shall contain the downloaded file. If the file is too big to fit in one single HTTP response, the file shall be chunked as defined in IETF RFC 9110 [7] and IETF RFC 9112 [8]. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 19 Figure 5 provides the sequence diagram illustrating the flow of events between entities for the download file operation. Figure 5: PFE Download file sequence diagram 1) To download a file from the File Exchange session the App makes an HTTP GET request to the URL of the file. The request contains the Authorization header with the "Bearer token" as defined in clause 6.2. 2) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The file Exchange Server answers the HTTP GET request from the App with a 200 OK status code, with the Content-disposition header with value "attachment; filename="<filename>"" and the header Content- Type with the MIME type of the file. 3) To download a file from the File Exchange session the PSAP Call-Taker makes an HTTP GET request to the URL of the file. The request contains the Authorization header with the "Bearer token" as defined in clause 6.2. 4) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The file Exchange Server answers the HTTP GET request from the PSAP Call-Taker with a 200 OK status code, with the Content-disposition header with value "attachment; filename="<filename>"" and the header Content-Type with the MIME type of the file. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.3.4.2 Example | HTTP request: GET /session/534wafds21s21fdf/example%20(1).png HTTPS/1.1 Host: file_exchange_server.example.com Authorization: Bearer PPtzs5zzG5Pkf61KPz51 HTTP response: HTTPS/1.1 200 OK Content-Disposition: attachment; filename="captura-de-pantalla-2024-03-11-143604 (1).png" Content-Type: image/png <binary> ETSI ETSI TS 104 014 V1.1.1 (2024-07) 20 |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4 Notifications channel | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.1 Overview | The operations defined in clause 7.3 allow users to upload files, download files and get the list of uploaded files. With these operations a participant can only know when another participant has uploaded a new file by polling with the list files operation, i.e. by making periodic HTTP requests at a certain time interval. This polling implies many extra requests during the emergency session. This is not efficient and could saturate the File Exchange Server if the requests are done with a high frequency or if there are many parallel sessions. To solve this issue a notifications channel that allows participants to be notified of new uploaded files is required. There are many subscription mechanisms available, but as PEMEA can be used by many different types of applications, each with its own architecture and needs, Server-Sent Events (SSE) is the selected mechanism to be used in the present document. As specified in clause 9.2 of HTML Living Standard [10], the Server-Sent Events mechanism allows for efficient real-time communication between servers and clients. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.2 Subscribe to the File Exchange session | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.2.1 Description | Participants can subscribe to events produced in a File Exchange session by opening a Server-Sent Event (SSE) connection to the File Exchange session URI as defined in HTML Living Standard [10]. To open the SSE connection a participant shall make an HTTP GET request to the URI of the File Exchange session. The HTTP request shall have the following headers: • Accept: The value shall be "text/event-stream". • Authorization: The token to access the File Exchange session as defined in clause 6.2. This is obtained as defined in clause 7.2. The HTTP request body shall be empty. The HTTP response shall have the following header: • Content-Type: The value shall be "text/event-stream". The HTTP response shall have a 200 OK status code. The HTTP connection shall remain opened as described in HTML Living Standard 10. In order to avoid the connection to be closed by proxies, the ping pong mechanism of SSE shall be used, this is done by sending a line starting with the ":" character. This message shall be sent periodically from the file sharing session to the connected participants every 60 seconds. Figure 6 provides the sequence diagram illustrating the flow of events between entities for the download file operation. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 21 Figure 6: PFE Subscribe sequence diagram 1) To subscribe to the File Exchange session events the App makes an HTTP GET request to the File Exchange session URI. The request contains the Accept header with value "text/event-stream" and the Authorization header with the "Bearer token" as defined in clause 6.2. 2) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The File Exchange Server answers the HTTP GET request from the App with a 200 OK status code, with the Content-Type header with value "text/event-stream" which indicates that the SSE connection has been successfully established. The File Exchange Server shall maintain the connection opened while the File Exchange session is open, and shall send events through the SSE as defined in clause 7.4.3. 3) To subscribe to the File Exchange session events the PSAP Call-Taker makes an HTTP GET request to the File Exchange session URI. The request contains the Accept header with value "text/event-stream" and the Authorization header with the "Bearer token" as defined in clause 6.2. 4) The File Exchange Server validates that the token is not expired and is valid for the requested File Exchange session. The File Exchange Server answers the HTTP GET request from the PSAP Call-Taker with a 200 OK status code, with the Content-Type header with value "text/event-stream" which indicates that the SSE connection has been successfully established. The File Exchange Server shall maintain the connection opened while the File Exchange session is open, and shall send events through the SSE as defined in clause 7.4.3. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.2.2 Example | HTTP request: GET /session/534wafds21s21fdf HTTPS/1.1 Host: file_exchange_server.example.com Accept: text/event-stream Authorization: Bearer PPtzs5zzG5Pkf61KPz51 HTTP response: HTTPS/1.1 200 OK Content-Type: text/event-stream ETSI ETSI TS 104 014 V1.1.1 (2024-07) 22 |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3 Receive notification events | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3.1 Overview | Once the SSE connection is established as defined in clause 7.4.2 the File Exchange Server can send messages over the opened HTTP connection as it is defined in clause 9.2 of HTML Living Standard [10]. The data transmitted shall be encoded as UTF-8. The data shall be separated by lines, and in each line an event may be sent. The events shall be transmitted with the following format: data:{event1} data:{event2} Where {event1} and {event2} are 2 consecutive events. The present document defines an event to notify when new files are successfully uploaded in clause 7.4.3.2 and another event to notify when the File Exchange session has been closed in clause 7.4.3.3. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3.2 File uploaded to the File Exchange session | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3.2.1 Description | Whenever a new file is successfully uploaded to a File Exchange session the File Exchange Server shall send a notification to all participants subscribed to the File Exchange session. It is done by sending a FILE_UPLOADED event which is a JSON object which structure is defined in clause 10.3. Figure 7 provides the sequence diagram illustrating the flow of events between entities to receive notifications from the File Exchange session. Figure 7: PFE Download file sequence diagram 1) After establishing the SSE channel subscription as described in clause 7.4.2 the File Exchange Server will send to the App the events produced in the File Exchange session. When a new file is successfully uploaded, the File Exchange Server will send a "FILE_UPLOADED" event to the App through the SSE. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 23 2) After establishing the SSE channel subscription as described in clause 7.4.2 the File Exchange Server will send to the PSAP Call-Taker the events produced in the File Exchange session. When a new file is successfully uploaded, the File Exchange Server will send a "FILE_UPLOADED" event to the PSAP Call-Taker through the SSE. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3.2.2 Example | Subscription is done as defined in clause 7.4.2, so the HTTP request would be opened and only the SSE messages would be sent, but for clarification the HTTP request is written again in this example. HTTP request: GET /session/534wafds21s21fdf HTTPS/1.1 Host: file_exchange_server.example.com Accept: text/event-stream Authorization: Bearer PPtzs5zzG5Pkf61KPz51 HTTP response: HTTPS/1.1 200 OK Content-Type: text/event-stream data:{"type":"FILE_UPLOADED","file":{"name":"example%20(1).png","size":328049,"type":"image/png","ti mestamp":1715858118715,"url":"https://file_exchange_server.example.com/session/534wafds21s21fdf/exam ple%20(1).png"}} data:{"type":"FILE_UPLOADED","file":{"name":"example.png","size":328049,"type":"image/png","timestam p":1574092280231,"url":"https://file_exchange_server.example.com/session/534wafds21s21fdf/example.pn g"}} |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3.3 File Exchange session closed by the File Exchange Server | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3.3.1 Description | Whenever a File Exchange session is closed as described in clause 9 the File Exchange Server shall send a notification to all participants subscribed to the File Exchange session before closing all the subscriptions. It is done by sending a FOLDER_CLOSED event which is a JSON object which structure is defined in clause 10.3. |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.3.3.2 Example | Subscription is done as defined in clause 7.4.2, so the HTTP request would be opened and only the SSE messages would be sent, but for clarification the HTTP request is written again in this example. HTTP request: GET /session/534wafds21s21fdf HTTPS/1.1 Host: file_exchange_server.example.com Accept: text/event-stream Authorization: Bearer PPtzs5zzG5Pkf61KPz51 HTTP response: HTTPS/1.1 200 OK Content-Type: text/event-stream data:{"type":"FOLDER_CLOSED","reason":"Folder was closed"} |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.4.4 Unsubscribe from the File Exchange session | To end the subscription the HTTP connection that established the SSE channel shall be closed. This can be done either by the participants or by the File Exchange Server. The File Exchange Server shall only close the SSE connections when a File Exchange session is closed as described in clause 9 and the File Exchange Server shall send a File Exchange session closed notification before closing the connections as defined in clause 7.4.3.3. ETSI ETSI TS 104 014 V1.1.1 (2024-07) 24 |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.5 Errors | |
a4c69013c9592b3b49c0122983240b32 | 104 014 | 7.5.1 Description | The File Exchange Server may encounter various situations where a request results in an error, these errors may occur because the schemas and operations defined in the present document are not followed or by circumstances that are out of control of the File Exchange Server. When the File Exchange Server shall return an error to a received request, the response shall be an HTTP answer with JSON formatted body with the structure defined in clause 7.5. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.