Distributed Data Fusion for On-Scene Signal Sensing With a Multi-UAV System

Abstract

This article presents an algorithmic framework for distributed data fusion using a network of multiple unmanned aerial vehicles (UAVs). Each UAV in the network is capable of sensing signals, performing computational tasks, and exchanging information with its immediate neighbors. Based on the limited local measurement and exchanged information only from neighbors, we aim to propose a fully distributed algorithm to integrate information from various measurements of individual UAVs. As an important building block of the distributed data fusion algorithm, a novel sum consensus scheme is also developed with only assuming row-stochastic weight matrices. Consequently, through the proposed distributed algorithm, each UAV is able to obtain a more reliable fused sensing result with higher accuracy, compared to the local measurement collected by the UAV itself. The effectiveness of the proposed algorithmic framework is validated by both theoretical analysis and numerical simulations.