Decentralized Optimization Over Tree Graphs

This paper presents a decentralized algorithm for non-convex optimization over tree-structured networks. We assume that each node of this network can solve small-scale problems and communicate approximate value functions with its neighbors based on novel multi-sweep communication protocol. In contrast to existing parallelizable algorithms optimization, the nodes are neither synchronized nor assign any central entity. None needs know whole topology network, but all is tree-structured. discuss conditions under which locally quadratic convergence rates be achieved. The method illustrated by running asynchronous protocol radial AC power case study.