Resource-Aware Exact Decentralized Optimization Using Event-Triggered Broadcasting

This article addresses the decentralized optimization problem where a group of agents with coupled private objective functions work together to exactly optimize summation local interests. Upon modeling as an equality-constrained centralized one, we leverage linearized augmented Lagrangian method design event-triggered algorithm that only requires light computation at generic time instants and peer-to-peer communication sporadic triggering instants. The for each agent are locally determined by comparing deviation between true broadcast primal variables certain thresholds. Provided threshold is summable over time, establish new upper bound effect behavior on primal-dual residual. Based this, same convergence rate O(1/k) periodic algorithms secured nonsmooth convex problems. Stronger results obtained strongly smooth problems, is, iterates linearly converge exponentially decaying Finally, developed strategy examined two common problems; comparison illustrate its performance superiority in exploiting resources.