Abstract

Incorporating error correcting coding techniques into target localization provides an immediate advantage that existing decoding algorithms can be used to determine which area the target is most likely located in. The important knowledge of exact sensor positions is, however, ignored in these decoding algorithms. This paper revisits the problem and shows that based on the weighted average of sensor positions with binary weightings from local decisions, a newly proposed decoding criterion can achieve a much better accuracy in target localization than the soft-and hard-decision rules proposed in [1], particularly when a certain number of sensors are under Byzantine attacks.