%0 Journal Article
%A HE Shuo
%A NI Wan-li
%A WANG Wen
%A ZHENG Jing-heng
%A TIAN Hui
%T Data-Importance-Aware Resource Allocation in IRS-Aided Edge Intelligent System
%D 2020
%R 10.13190/j.jbupt.2020-162
%J Journal of Beijing University of Posts and Telecommunications
%P 51-58
%V 43
%N 6
%X In order to solve the problem of model aggregation in intelligent reflecting surface (IRS) aided edge intelligent system, a data-importance-aware resource allocation algorithm is proposed by using convex optimization and branch-and-bound methods to alternately design the user's uplink power, transmission time, and the phase shifts of IRS. Simulation results show that the proposed algorithm can effectively aggregate the model parameters of the distributed agents based on the importance difference of local data, and can maximize the uplink weighted sum rate.
%U https://journal.bupt.edu.cn/EN/10.13190/j.jbupt.2020-162