Likelihood function for target tracking in single frequency network

Abstract

The single frequency network (SFN) refers to a special kind of sensor networks whose illuminators broadcast the same frequency band of signals. The remarkable feature of SFN is that it generates multiple simultaneous measurements from the same target with illuminators uncertain, which brings heavy computational loads in the context of target tracking. To solve this problem, decomposition of the likelihood function (LF) of SFN into the sum of sub-LFs corresponding to each illuminator is proposed. Then, the sub-LFs are approximately modeled through the random finite set framework. The advantage is that it avoids enumerating all the associations between the illuminators and the measurements. Simulation results demonstrate the effectiveness of the proposed LF.