Resource allocation modelling using methods of feasible directions in phased array radar systems

In this paper, we report on recent progress results on optimal real-time resource allocation in phased array radar systems. A recently proposed resource allocation approach, Q-RAM, is considered and is observed to generate non-optimal results. We identify the shortcomings of this method and firstly extend it using Karesh-Kuhn-Tucker (KKT) optimality conditions for the single-resource-type case to obtain a globally optimal algorithm. We later generalize this further for the multiple-resource-type case. However, this particular approach has its origins in quality-of-service domain and is fundamentally limited to sampled cost functions. The availability of empirically obtained sampled convex tracking performance curves for phased array radar, and the feasibility of continuous approximations to these curves lead our study to the consideration of well formulated alternative methods from optimization literature belonging to the Methods of Feasible Directions. We apply in particular the Gradient-Projection Algorithm for the more general multiple resource type case. Our experimental studies using simulated radar performance curves show that superior performance can be obtained in closeness to optimal and execution speed. Closeness to optimality improvement becomes significant in particular for dense target scenarios with large number of targets.