I. Introduction
Cooperative localization is gaining more and more attention throughout the research community. The striking advantage over non-cooperative localization is that infrastructure requirements can be greatly relaxed [1]–[3], i.e. agents require fewer anchors to obtain unambiguous position estimates. Consider the example in Fig. 1, where two agents want to localize themselves. Both agents have obtained distance estimates with respect to two anchors. With non-cooperative approaches, the agents cannot determine their positions unambiguously. On the other hand, cooperation among the agents could resolve the ambiguity because each agent becomes a surrogate anchor (an anchor with some uncertainty) for the other agent. The primary challenge of cooperative localization breaks down to expressing the uncertainties of agents’ positions accurately.