Resource Allocation for Distributed Multitarget Tracking in Radar Networks With Missing Data

In this paper, an effective joint measurement selection and power allocation (JMSPA) scheme is proposed for the distributed multi-target tracking task in radar networks with missing data. Missing data may occur during data exchange between radars and a fusion center (FC) due to unreliability of communication channels. First, we derive the predicted conditional Cramér-Rao lower bound (PC-CRLB) in the presence of missing data and measurement origin uncertainty, which are scaled by an improved information reduction factor (IIRF). Second, an overall cost function is formulated based on the derived PC-CRLB to quantify the global multi-target tracking performance. Combined with the practical resource constraints of the multi-radar architectures, the formulated JMSPA problem is shown to be an NP-hard optimization problem containing coupled continuous and binary variables. Third, we propose an efficient solver integrating convex relaxation with accelerated penalized sequential convex programming (AP-SCP) algorithm for solving this problem. Finally, numerical simulation results demonstrate that the proposed JMSPA has better performance than the traditional resource allocation strategy that ignores missing data, and is robust to resource allocation in different missing data scenarios.