Joint Optimization Strategy of Computation Offloading and Resource Allocation in Multi-Access Edge Computing Environment
Joint Optimization Strategy of Computation Offloading and Resource Allocation in Multi-Access Edge Computing Environment
复制标题
DOI：
10.1109/tvt.2020.3003898
复制
发表时间：
2020-09-01
影响因子：
6.8
通讯作者：
Han, Zhu
中科院分区：
文献类型：
--
作者：
Li, Huilin;Xu, Haitao;Han, Zhu
为了帮助用户终端设备（UTDS）有效处理计算密集型和时间延迟敏感的计算任务，已经提出了多访问边缘计算（MEC）。但是，由于UTD的性能之间的差异以及MEC服务器的资源限制，因此UTDS的卸载决策与网络中资源分配之间的联合优化仍然是研究的重点。本文研究了多用户和多服务器方案中的联合计算卸载和资源分配策略。首先，我们通过约束卸载决策，渠道选择，权力分配和资源分配，将计算卸载和资源分配的联合优化问题作为混合整数非线性编程（MINP）问题。其次，我们提出了一种基于遗传算法的两阶段启发式优化算法，该算法将计算卸载和资源分配的关节优化问题分为两个阶段。基于卸载决策与资源分配方案之间的耦合关系，我们迭代地更新了问题的解决方案，并最终获得了优化问题的稳定收敛解决方案。最后，将提出的算法与其他经典方法进行了比较，以证明有效性。
In order to help user terminal devices (UTDs) efficiently handle computation-intensive and time-delay sensitive computing task, multi-access edge computing (MEC) has been proposed. However, due to the differences among the performance of UTDs, and the resource limitation of MEC servers, the joint optimization between the offloading decisions of UTDs and the allocation of resources in network is still a focus of the research. This paper studies the joint computation offloading and resource allocation strategy in multi-user and multi-server scenarios. Firstly, we formulate the joint optimization problem of computation offloading and resource allocation as a mixed integer nonlinear programming (MINP) problem to minimize the energy consumption of UTDs, by constraining the offloading decision, channel selection, power allocation and resource allocation. Secondly, we propose a two-stage heuristic optimization algorithm based on genetic algorithms, which divides the joint optimization problem of computation offloading and resource allocation in two stages. Based on the coupling relationship between the offloading decision and the resource allocation scheme, we iteratively update the solution of the problem, and finally obtain the stable convergence solution of the optimization problem. Finally, the proposed algorithm is compared with other classical methods to prove the effectiveness.