text
stringlengths
0
8.13M
rui001@e.ntu.edu.sg;han.yu@ntu.edu.sg);dniyato@ntu.edu.sg).
J. Kang is with the School of Automation, Guangdong University of quantum tasks can be fully computed. Thirdly, distributed
Technology,China(e-mail:e-mail:kavinkang@gdut.edu.cn. quantumcomputingmaysufferfromfidelitydegradation.This
X. (Sherman) Shen is with the Department of Electrical and Computer
is unavoidable at the moment, and reduces the efficiency of
Engineering, University of Waterloo, Waterloo, ON, Canada, N2L 3G1 (e-
mail:sshen@uwaterloo.ca). quantum teleportation in quantum networks. Thus, deploying
Provision of the deployed
Applications require the
quantum computers Entanglement
use of quantum computing
Superposition
Military
Measurement/Interference
Properties of quantum mechanics
Minimize costs of using the
deployed quantum computers
under quantum resources
Quantum computer
Internet of Things Smart Grid
operator
Provision of on-demand Perform quantum computing on
quantum computer deployment distributed quantum computing
Fig. 1: The adaptive resource allocation approach with two-stage stochastic programming for distributed quantum computing.
quantum resources in distributed quantum computing in its 2n possible outcomes and have the same chance of being
current form may result in highly inefficient utilization of measured for each. Therefore, quantum computers can store
quantum computing resources due to the inherent uncertainty and manipulate more information than classical computers,
in real-world circumstances. providing far more diverse possibilities and opportunities.
To address the above challenges, in this paper, we propose 2) Interference: Qubits must be subjected to some kind
an adaptive resource allocation approach towards efficient of measurement in order to represent and store their values
and scalable distributed collaborative quantum computing. It and results. If one intervenes in the process, it is possible to
consists of deterministic and stochastic programming models measure and see the results of the paths. When the process is
for quantum resource allocation with uncertainty in collabo- interrupted, the states of the qubits collapse to classical bits,
rative settings to help quantum computer operators minimize and the computation results appear. For example, the result of
total deployment costs. It jointly considers the uncertainty of a coin toss is known definitively (e.g., heads or tails) when
future quantum computing demands, the computing power of the coin reaches the bottom.
quantum computers, and the fidelity in distributed quantum 3) Entanglement: Two qubits can be entangled with each
computing in order to optimally deploy and utilize quantum otherasanentanglementpair[5],whichmeansthatwhenone
computers. We conduct extensive experiments to reveal the qubit is measured, the other qubit can also be known because
importance of the optimal deployment of quantum computers of their entangled nature. In addition, a pair of entanglement
in distributed quantum computing. In comparison to other qubitsisentangledmaximallyalsoknownasaBellstatewhen
resource allocation models, the proposed approach can reach the results of measuring one of them will certainly affect
the lowest total deployment cost. Finally, we highlight oppor- the outcome of measuring the other one later. The fidelity
tunities and challenges in distributed quantum computing for of entanglement pairs is the metric of attenuation for the
various military applications in future networks. entangledqubitsbetweentworemotequantumcomputers.The
fidelity scale runs from 0 to 1, where 1 indicates the best
performance that the entanglement can achieve.
II. FUNDAMENTALSOFQUANTUMCOMPUTING The main analogies between classical computing and the
technology used to realize a true quantum computer are the
A. Quantum Computing
following. In classical computing, a circuit is a computational
Three properties of quantum mechanics define quantum
model that enables the processing of input values through
computing: superposition, interference, and entanglement, as
gates and operations. Similarly, the quantum circuit model
shown in Fig. 1.
proceeds with implementations on qubits and involves an
1) Superposition: In classical computers, the binary bits
ordered sequence of quantum gates that permit logical in-
0 and 1 are used to encode information for computations.
teraction between qubits. In particular, the measurement of
A superposition in quantum computing allows the encoding
qubits must occur near the end of the quantum circuit. A
of qubits in a combination of two classical binary states. A
quantum processor is a small quantum computer that can
well-known example of quantum mechanics is the tossing of
executequantumgatesonasmallnumberofqubitsandallows
a coin. When a coin is tossed in the air, the exact outcome is
the entanglement of qubits inside.
unknown, and its probability values reflect qubits. Each qubit
B. Distributed Quantum Computing
can be expressed as a linear combination of binary states,
where the coefficients correspond to the probabilities of the Theconceptofdistributedquantumcomputingreliesonthe
qubit amplitudes. Due to the superposition, n qubits can store following principles:
Ref. Scenarios Problems Performance Metrics Quantum Algorithms
[6] Quantum-enhancedsmartgrid Paradigm and elements of Grid efficiency, security, and Quantum Fourier transform,
quantum computing in power stability quantumsearch,andquantum
systems neural networks
[2] Realizing IoT environments IoT-sensor space problem Data accuracy and data tem- Quantum-basedalgorithmand
porary efficiency quantum Fourier transform
[7] UAV-mounted wireless net- Trajectory and uplink trans- Convergenceperformanceand Quantum-based action selec-
works mission rate optimization learningqualityoftrajectories tion strategy
[8] Quantum machine learning Improvement computation of Computation complexity Quantum mechanics, Shor’s,
machine learning algorithms and Grover’s algorithm
[9] Entanglement flow in quan- Maximization of fidelity over Entanglement rate and router Quantum router architecture
tum networks long-distance links infidelity design
[10] Adaptive measurements in Estimation state of qubits Successprobabilityofestima- Grover’s algorithm
quantum tion
[11] Resource allocation in quan- Quantum network flow opti- The number of served apps, (Weighted round robin algo-
tum networks mization the Bell pair capacity, and rithm for resource allocation)
Jain’s fairness
TABLE I: Summary of representative related works demonstrated through scenarios, problems, performance metrics, and
quantum algorithms
1) Quantum Networks through Entanglement: To enable QPUs, which are subsequently utilized to read output after
distributed quantum computing, a quantum network must be a measurement. The QPUs should have specialized space to
developedtoconnectquantumcomputers.However,duetothe store and operate matter qubits that improve the ability of
non-cloning theorem in quantum mechanics, qubits cannot be distributed quantum computing to work quickly and consis-
duplicated or cloned across quantum computers. Thus, two tently. In addition, network communication is also needed for
quantum computers must exchange or transfer qubits through QPUs to send signals across the network while measuring
the concept of quantum teleportation. In quantum telepor- and reading qubits. The effort required to input data and the