Efficient Network Localization in the Presence of Likelihood Ambiguity

Fisher information matrices (FIMs) have been employed to determine performance limits on network localization. However, it is unclear whether the performance limit indicated by an FIM in network localization can be achieved asymptotically in cases where likelihood ambiguity is present. Indeed, these cases are important as likelihood ambiguity often occur when the network is designed under the guideline of maximizing the network’s power efficiency. We prove in this letter that, even in the presence of ambiguity, an FIM can be achieved asymptotically via joint design of a power allocation strategy and a position estimator. Results in this letter sharpen our understanding on FIMs and provide design guidelines for network localization.