2018/10/04 by Sulaiman A. Alghunaim, Kun Yuan, Alghunaim, Sulaiman A. +3 · 1 citation
Computer Science · #Cooperative Communication and Network Coding #Distributed Control Multi-Agent Systems #FOS: Mathematics #Neural Networks Stability and Synchronization #Optimization and Control (math.OC)
paper · pdf · doi:10.48550/arxiv.1810.02124
openalex publication_date 2018/10/04 · openalex created_date 2022/09/29 · openalex updated_date 2026/07/28
This work develops a proximal primal-dual decentralized strategy for\nmulti-agent optimization problems that involve multiple coupled affine\nconstraints, where each constraint may involve only a subset of the agents. The\nconstraints are generally sparse, meaning that only a small subset of the\nagents are involved in them. This scenario arises in many applications\nincluding decentralized control formulations, resource allocation problems, and\nsmart grids. Traditional decentralized solutions tend to ignore the structure\nof the constraints and lead to degraded performance. We instead develop a\ndecentralized solution that exploits the sparsity structure. Under constant\nstep-size learning, the asymptotic convergence of the proposed algorithm is\nestablished in the presence of non-smooth terms, and it occurs at a linear rate\nin the smooth case. We also examine how the performance of the algorithm is\ninfluenced by the sparsity of the constraints. Simulations illustrate the\nsuperior performance of the proposed strategy.\n