Non-myopic Beam Scheduling for Multiple Smart Target Tracking in Phased Array Radar Network

TL;DR

This paper introduces a scalable non-myopic beam scheduling policy for tracking multiple smart targets in phased array radar networks, improving performance over previous methods by using Whittle indices within a Markov decision process framework.

Contribution

It formulates the target tracking as a restless multi-armed bandit problem and proposes a Whittle index-based policy that is computationally efficient and effective for large radar networks.

Findings

The proposed policy achieves near-optimal tracking performance.

The method scales linearly with the number of targets.

Numerical simulations validate the policy's effectiveness across scenarios.

Abstract

A smart target, also referred to as a reactive target, can take maneuvering motions to hinder radar tracking. We address beam scheduling for tracking multiple smart targets in phased array radar networks. We aim to mitigate the performance degradation in previous myopic tracking methods and enhance the system performance, which is measured by a discounted cost objective related to the tracking error covariance (TEC) of the targets. The scheduling problem is formulated as a restless multi-armed bandit problem (RMABP) with state variables, following the Markov decision process. In particular, the problem consists of parallel bandit processes. Each bandit process is associated with a target and evolves with different transition rules for different actions, i.e., either the target is tracked or not. We propose a non-myopic, scalable policy based on Whittle indices for selecting the targets to be tracked at each time. The proposed policy has a linear computational complexity in the number of targets and the truncated time horizon in the index computation, and is hence applicable to large networks with a realistic number of targets. We present numerical evidence that the model satisfies sufficient conditions for indexability (existence of the Whittle index) based upon partial conservation laws, and, through extensive simulations, we validate the effectiveness of the proposed policy in different scenarios.