arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2306.01159quant-phcs.SYeess.SY
收藏

用于边缘节点部署与工作负载分配的基于量子的分布式算法

Quantum-based Distributed Algorithms for Edge Node Placement and Workload Allocation

Duong The Do, Ni Trieu, Duong Tung Nguyen

AI总结:

针对边缘服务器部署与工作负载分配的NP难MILP问题,提出将其分解为QUBO与LP子问题的量子-经典混合求解框架,并通过实验验证其有效性。

摘要:

边缘计算是一项前景广阔的技术,可提供卓越的用户体验,并支持各类创新的物联网应用。本文提出了一种混合整数线性规划(MILP)模型,用于最优边缘服务器部署和工作负载分配,而该问题已知是NP难的。为此,我们探索利用量子计算来应对这一计算上具有挑战性的问题的可能性。然而,现有量子求解器仅限于求解无约束二进制规划问题。为克服这一障碍,我们提出一种量子-经典混合解决方案,将原始问题分解为一个二次无约束二进制优化(QUBO)问题和一个线性规划(LP)子问题。QUBO问题可由量子求解器求解,而LP子问题可使用传统LP求解器求解。数值实验证明了利用量子优越性解决边缘计算中复杂优化问题的实用性。

英文摘要:

Edge computing is a promising technology that offers a superior user experience and enables various innovative Internet of Things applications. In this paper, we present a mixed-integer linear programming (MILP) model for optimal edge server placement and workload allocation, which is known to be NP-hard. To this end, we explore the possibility of addressing this computationally challenging problem using quantum computing. However, existing quantum solvers are limited to solving unconstrained binary programming problems. To overcome this obstacle, we propose a hybrid quantum-classical solution that decomposes the original problem into a quadratic unconstrained binary optimization (QUBO) problem and a linear program (LP) subproblem. The QUBO problem can be solved by a quantum solver, while the LP subproblem can be solved using traditional LP solvers. Our numerical experiments demonstrate the practicality of leveraging quantum supremacy to solve complex optimization problems in edge computing.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑