An adaptive resource allocation strategy for multiple target tracking with different performance requirements

The resource aware design has a considerable impact on the improvement of multiple target tracking (MTT) performance for colocated MIMO radar. A common approach is to enhance the tracking performance of worst case among multiple targets through running out of system resources, namely Minmax method. The resulting tracking performance of all targets would reach the same level as the worst one. In practice, it is not reasonable to require all these targets to reach a same performance since the targets with different distinct range, angle and Doppler have different threatening degree. In this paper, we design a novel resource aware framework to ensure more suitable MTT performance rather than Minmax method. In particular, we introduce an adaptive cost function (ACF), involving the tracking accuracy requirements, the cost and resource constraints. Then we formulate a novel but non-convex optimization problem by taking into account the ACF. Meanwhile, we use the posterior Cramér-Rao lower bound (PCRLB) as a metric of tracking performance, which is capable of robustly controlling the different tracking accuracy to each target. Then, we turn the considered problem into a tractable problem sharing theoretical convexity feature. Finally, numerical results are performed showing that the proposed algorithm outperforms in the tracking error under the track accuracy requirements and the limited resource than the available counterparts.