Distributed Blind Calibration via Output Synchronization in Lossy Sensor Networks

TL;DR

This paper introduces a distributed algorithm for blind sensor calibration in networks, achieving asymptotic agreement on sensor parameters despite noise and communication issues, with proven convergence and simulation validation.

Contribution

It presents a novel gradient-based distributed calibration method using output synchronization, including a modified version robust to noise and outages, with rigorous convergence proofs.

Findings

Algorithm achieves asymptotic consensus on sensor parameters.

Modified algorithm handles measurement noise and communication outages.

Simulation examples demonstrate effectiveness of the proposed methods.

Abstract

In this paper a novel distributed algorithm for blind macro calibration in sensor networks based on output synchronization is proposed. The algorithm is formulated as a set of gradient-type recursions for estimating parameters of sensor calibration functions, starting from local criteria defined as weighted sums of mean square differences between the outputs of neighboring sensors. It is proved, on the basis of an originally developed methodology for treating higher-order consensus (or output synchronization) schemes, that the algorithm achieves asymptotic agreement for sensor gains and offsets, in the mean square sense and with probability one. In the case of additive measurement noise, additive inter-agent communication noise, and communication outages, a modification of the original algorithm based on instrumental variables is proposed. It is proved using stochastic approximation arguments that the modified algorithm achieves asymptotic consensus for sensor gains and offsets, in the mean square sense and with probability one. Special attention is paid to the situation when a subset of sensors in the network remains with fixed characteristics. Illustrative simulation examples are provided.