Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RL-Aided Cognitive ISAC: Robust Detection and Sensing-Communication Trade-offs

Adam Umra, Aya M. Ahmed, Aydin Sezgin

TL;DR

This work tackles robust radar sensing and spectrum-efficient communications in a cognitive ISAC system employing a 2D UPAs-based massive MIMO BS. It integrates a Wald-type detector to handle non-Gaussian clutter and a SARSA-based RL agent to learn target positions without environmental priors, guiding a joint waveform design that balances sensing accuracy and downlink throughput. The authors derive a closed-form, computation-efficient solution to the trade-off problem and demonstrate through Monte Carlo simulations that the RL-aided approach significantly improves detection probability while maintaining competitive sum-rate across stationary and dynamic scenarios. The results highlight the potential of RL-driven sensing to enable robust, spectrum-efficient ISAC in future wireless networks.

Abstract

This paper proposes a reinforcement learning (RL)-aided cognitive framework for massive MIMO-based integrated sensing and communication (ISAC) systems employing a uniform planar array (UPA). The focus is on enhancing radar sensing performance in environments with unknown and dynamic disturbance characteristics. A Wald-type detector is employed for robust target detection under non-Gaussian clutter, while a SARSA-based RL algorithm enables adaptive estimation of target positions without prior environmental knowledge. Based on the RL-derived sensing information, a joint waveform optimization strategy is formulated to balance radar sensing accuracy and downlink communication throughput. The resulting design provides an adaptive trade-off between detection performance and achievable sum rate through an analytically derived closed-form solution. Monte Carlo simulations demonstrate that the proposed cognitive ISAC framework achieves significantly improved detection probability compared to orthogonal and non-learning adaptive baselines, while maintaining competitive communication performance. These results underline the potential of RL-assisted sensing for robust and spectrum-efficient ISAC in next-generation wireless networks.

RL-Aided Cognitive ISAC: Robust Detection and Sensing-Communication Trade-offs

TL;DR

This work tackles robust radar sensing and spectrum-efficient communications in a cognitive ISAC system employing a 2D UPAs-based massive MIMO BS. It integrates a Wald-type detector to handle non-Gaussian clutter and a SARSA-based RL agent to learn target positions without environmental priors, guiding a joint waveform design that balances sensing accuracy and downlink throughput. The authors derive a closed-form, computation-efficient solution to the trade-off problem and demonstrate through Monte Carlo simulations that the RL-aided approach significantly improves detection probability while maintaining competitive sum-rate across stationary and dynamic scenarios. The results highlight the potential of RL-driven sensing to enable robust, spectrum-efficient ISAC in future wireless networks.

Abstract

This paper proposes a reinforcement learning (RL)-aided cognitive framework for massive MIMO-based integrated sensing and communication (ISAC) systems employing a uniform planar array (UPA). The focus is on enhancing radar sensing performance in environments with unknown and dynamic disturbance characteristics. A Wald-type detector is employed for robust target detection under non-Gaussian clutter, while a SARSA-based RL algorithm enables adaptive estimation of target positions without prior environmental knowledge. Based on the RL-derived sensing information, a joint waveform optimization strategy is formulated to balance radar sensing accuracy and downlink communication throughput. The resulting design provides an adaptive trade-off between detection performance and achievable sum rate through an analytically derived closed-form solution. Monte Carlo simulations demonstrate that the proposed cognitive ISAC framework achieves significantly improved detection probability compared to orthogonal and non-learning adaptive baselines, while maintaining competitive communication performance. These results underline the potential of RL-assisted sensing for robust and spectrum-efficient ISAC in next-generation wireless networks.

Paper Structure

This paper contains 25 sections, 55 equations, 9 figures, 3 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Methods
  3. System Model
  4. Radar Signal Model
  5. Communication Signal Model
  6. Cognitive ISAC for Sensing–Communication Trade-offs
  7. Reinforcement Learning Framework
  8. States and Actions
  9. Reward
  10. SARSA-based Algorithm
  11. Results and Discussion
  12. Parameters
  13. Comparison of RL-Based and Benchmark Schemes in a Stationary Scene
  14. Impact of Sensing–Communication Trade-Offs on System Performance
  15. Effect of the Number of Communication Users on Detection and Sum Rate
  16. ...and 10 more sections

Figures (9)

  • Figure 1: ISAC BS system serving $K$ single-antenna users while operating as a monostatic cognitive radar.
  • Figure 2: RL framework for the SARSA-based cognitive ISAC.
  • Figure 3: Probability of detection over pulses for Target 1 and 2 and with $\rho = 0.2$ and $K = 48$
  • Figure 4: Sum rate over time with $\rho = 0.2$, $K = 48$ and transmit SNR of $12$ dB
  • Figure 5: Probability of detection over time and sum rate over varying transmit SNRs for different trade-off $\rho$
  • ...and 4 more figures