Convergence and Applications of a Gossip-based Gauss-Newton Algorithm

Abstract

The Gauss-Newton algorithm is a popular and efficient centralized method for solving non-linear least squares problems. In this paper, we propose a multi-agent distributed version of this algorithm, named Gossip-based Gauss-Newton (GGN) algorithm, which can be applied in general problems with non-convex objectives. Furthermore, we analyze and present sufficient conditions for its convergence and show numerically that the GGN algorithm achieves performance comparable to the centralized algorithm, with graceful degradation in case of network failures. More importantly, the GGN algorithm provides significant performance gains compared to other distributed first order methods.

accepted by IEEE Transactions on Signal Processing

References in corpus (4)

• Gossip Algorithms for Distributed Signal Processing
• Diffusion Adaptation Strategies for Distributed Optimization and Learning over Networks
• Distributed Robust Power System State Estimation
• Convergence analysis of a proximal Gauss-Newton method

Cited by in corpus (8)

• Decentralized Frank-Wolfe Algorithm for Convex and Non-convex Problems
• Acoustic Local Positioning With Encoded Emission Beacons
• Stability and Performance Limits of Adaptive Primal-Dual Networks
• A Framework for Phasor Measurement Placement in Hybrid State Estimation via Gauss-Newton
• Evaluation of Zadoff-Chu, Kasami and Chirp based encoding schemes for Acoustic Local Positioning Systems
• Simultaneous Calibration and Navigation (SCAN) of Multiple Ultrasonic Local Positioning Systems
• NEXT: In-Network Nonconvex Optimization
• Robust Decentralized State Estimation and Tracking for Power Systems via Network Gossiping