Abstract
Protecting a large number of wireless communication links against the interception of enemy in Warfighter Information Network-Tactical (WIN-T) systems is challenging. Due to the high mobility of ground combat vehicles (GCVs), the Low Probability of Intercept (LPI) capacity can easily be violated. Prior work focuses mainly on a single interception technique, which exposes vulnerabilities when multiple interception techniques are deployed simultaneously. We propose a strategy against both energy-based and correlation-based interception techniques by jointly optimizing power allocation (PA) and spreading factor assignment (SA) of the WIN-T. This non-convex problem is then solved by advanced optimization techniques such as decomposition and difference of convex functions (DC). We also propose a communication mode selection strategy to improve the throughput performance in the context of LPI conservation. To obtain the optimized solution in near real-time, we design a Multi-Agent Deep Reinforcement Learning (MADRL) algorithm. Our numerical results show the performance of the proposed MADRL algorithm is close to the optimal solution, making it applicable for practical systems.
Funding Information
  • Ultra TCS and REMI

This publication has 36 references indexed in Scilit: