Cooperative multi-AUV localization using distributed extended information filter

Cooperative multi-AUV localization has the potential to outperform single-AUV localization, by taking advantage of data sharing among the team members. Unlike terrestrial communication links, underwater communication links have many issues pertaining to the channel stability and bandwidth, and therefore a decentralized localization is preferred. However this opens up new challenges to team members working in cooperation. This paper uses analytical examples to illustrate the cooperation problems in the decentralized architecture. To solve these problems, we propose a new cooperative multi-AUV localization algorithm using distributed extended information filter (DEIF). The proposed method only requires small transmission packets and is designed for use under constrained underwater communication. It is robust to packet loss, providing consistent position estimates when fusing correlated data. We demonstrate the effectiveness and advantages of the proposed method with comparative results using data from simulation and field experiments.