Distributed target tracking in sensor networks using multi-step consensus

The authors propose a new algorithm for distributed target tracking by sensor networks with sensors having limited sensing range (radars, sonars, cameras), using an adaptive multi-step consensus scheme. The proposed distributed adaptation strategy represents the core element of the algorithm, providing: (i) asymptotic consensus gains giving the desired importance to the nodes observing the target, and (ii) fast convergence rate of the consensus scheme, enabling efficient implementation. Another contribution of this study is a theoretical study of the algorithm properties, including asymptotic stability and reduction of noise influence. The given simulation results show that the proposed algorithm outperforms the existing algorithms, preserving, at the same time, much lower communication and computation requirements.