Towards Accelerated Rates for Distributed Optimization over Time-Varying Networks

We study the problem of decentralized optimization with strongly convex smooth cost functions. This paper investigates accelerated algorithms under time-varying network constraints. In our approach, nodes run a multi-step gossip procedure after taking each gradient update, thus ensuring approximate consensus at iteration. The outer cycle is based on Nesterov scheme. Both computation and communication complexities method have an optimal dependence global function condition number \(\kappa _g\). particular, algorithm reaches complexity \(O(\sqrt{\kappa _g}\log (1/\varepsilon ))\).