Adaptive Policies for Perimeter Surveillance Problems

TL;DR

This paper introduces a sequential perimeter surveillance model using a combinatorial multi-armed bandit framework with Poisson rewards, providing a new solution approach that learns and adapts over time to improve intrusion detection.

Contribution

It develops a formal CMAB model with filtered feedback for perimeter surveillance and proposes an upper confidence bound algorithm with proven performance bounds.

Findings

The proposed UCB algorithm outperforms competing methods in simulations.

Theoretical bounds on the algorithm's expected performance are established.

Empirical results demonstrate increased reliability in intrusion detection.

Abstract

Maximising the detection of intrusions is a fundamental and often critical aim of perimeter surveillance. Commonly, this requires a decision-maker to optimally allocate multiple searchers to segments of the perimeter. We consider a scenario where the decision-maker may sequentially update the searchers' allocation, learning from the observed data to improve decisions over time. In this work we propose a formal model and solution methods for this sequential perimeter surveillance problem. Our model is a combinatorial multi-armed bandit (CMAB) with Poisson rewards and a novel filtered feedback mechanism - arising from the failure to detect certain intrusions. Our solution method is an upper confidence bound approach and we derive upper and lower bounds on its expected performance. We prove that the gap between these bounds is of constant order, and demonstrate empirically that our approach is more reliable in simulated problems than competing algorithms.