Energy-Efficient and Intelligent ISAC in V2X Networks with Spiking Neural Networks-Driven DRL

Abstract

Integrated sensing and communication (ISAC) is emerging as a key enabler for vehicle-to-everything (V2X) systems. However, designing efficient beamforming schemes for ISAC signals to achieve accurate sensing and enhance communication performance in the dynamic and uncertain environments of V2X networks presents significant challenges. While artificial intelligence technologies offer promising solutions, the energy-intensive nature of neural networks imposes substantial burdens on communication infrastructures. To address these challenges, this work proposes an energy-efficient and intelligent ISAC system for V2X networks. Specifically, we first leverage a Markov Decision Process framework to model the dynamic and uncertain nature of V2X networks. This framework allows the roadside unit to develop beamforming schemes relying solely on its current sensing information, eliminating the need for numerous pilot signals and extensive CSI acquisition. We then introduce an advanced deep reinforcement learning (DRL) algorithm, enabling the joint optimization of beamforming and power allocation to guarantee both communication rate and sensing accuracy in dynamic and uncertain V2X scenario. To alleviate the energy demands of neural networks, we integrate spiking neural networks (SNNs) into the DRL algorithm. The event-driven, sparse spike-based processing of SNNs significantly improves energy efficiency while maintaining strong performance. Extensive simulation results validate the effectiveness of the proposed scheme with lower energy consumption, superior communication performance, and improved sensing accuracy.