Radar Network Optimization Under Communication Cost Constraint

We consider the network optimization problem of target localization in a distributed Multiple Input Multiple Output (MIMO) radar system under communication cost constraint. Limited by the communication cost, it is impossible to combine all the radar nodes in the network for signal-level data fusion. However, performing full parameter-level data fusion leads to significant localization performance degradation. We group the radar nodes into clusters to perform joint signal-level and parameter-level data fusion. Within a cluster, the radar nodes send received signals to the sub-fusion center (FC)s for signal-level processing. Then the localization results from each cluster are sent to a FC to perform parameter-level data fusion. The aim is to find an optimal network topology, that satisfies the constraints of communication cost while maximizing the localization accuracy of the radar network. A greedy algorithm with shrinkage constraint is proposed to obtain an approximate optimal solution of the network topology. The proposed algorithm offers a considerable reduction in computational complexity when compared with an exhaustive search. Simulation results show that the proposed joint signal-level and parameter-level data fusion scheme provides better localization performance than pure signal-level fusion.