Adaptive Cooperative Localization Using Relative Position Estimation for Networked Systems with Minimum Number of Communication Links

Abstract

In this paper, a novel solution for cooperative localization problem involving a network with multiple mobile agents accessing to only one beacon agent is presented. The solution can be applied to a network with a minimum number of communication links (subjecting the network topology to be an undirected spanning tree). The objective is to provide each agent in the network with information on its absolute position without using any on-board global positioning sensors such as GPS receivers. The proposed solution uses an adaptive relative position estimation algorithm for each pair of neighboring mobile agents. The proposed estimation algorithm requires each agent to measure the relative inter-agent distance. The local velocity vector is also measured and transmitted to the neighboring agents. The estimation mechanism incorporating a signum function outperforms the recently established relative position estimation algorithms, in terms of positioning and tracking errors. An adaptive cooperative localization (ACL) algorithm is formed by augmenting the relative position estimation in a cooperative observer scheme suitably applicable for accomplishing localization task involving a network of mobile agents. The salient feature of the proposed ACL algorithm is that the communication graph among the agents needs only to have one undirected path between two agents in the network. Such convenience promotes easy practical implementation and lite computation for each agent. The proof of the proposed algorithm is provided using the Lyapunov stability theorem. Three simulation case studies are presented to evaluate the performance of the solution in different scenarios, including the stationary and moving beacon agent as well as the non-cooperatively controlled and cooperatively controlled network of mobile agents. The comparative studies reveal that the ACL algorithm is superior to the recently investigated linear-convex algorithm. The number of communication links required for the localization task to be carried out by the proposed algorithm is minimum, thus promoting a more preferable energy-efficient solution.