Probability-aware Qubit-to-Processor Mapping in Distributed Quantum Computing
Yingling Mao, Yu Liu, Yuanyuan Yang
Abstract
Quantum computing has exciting potential but a current technological hurdle lies in the limited number of qubits in a single processor. One way to address this challenge is to assemble small, specialized quantum processors into a larger computing system, named distributed quantum computing. In this work, we focus on a key problem in distributed quantum computing: how to map logical qubits of a specific quantum circuit to different processors in a heterogeneous quantum network with the goal of minimizing the overall communication overhead. To solve this problem, we formulate a probability-aware qubit-to-processor mapping model, where the communication overhead between each pair of processors is determined through probabilistic analyses based on the link entanglement generation rates. We also introduce a multi-flow routing protocol in our model to improve the overall entanglement rates. Afterward, we employ a multistage hybrid simulated annealing algorithm to minimize the total communication overhead. Lastly, we perform extensive simulations to show the superiority of our solutions under a wide range of system settings.