DRL-MEC-based anti-malicious jamming decision in wireless body area network

Wireless body area networks (WBAN) expose new challenges in security due to resource constraints. In particular, malicious jamming attacks can seriously threaten the security of wireless communications in WBANs. In this paper, deep reinforcement learning (DRL) and mobile edge computing (MEC) techniques are used. Without considering the channel state information, the proposed scheme exploits the selection of edge devices, transmit power, and offload rate to improve the system's utility, including the shared gain of the offloaded signal, computational latency, energy consumption, and SNR. Latency and energy consumption are key to ensuring data transmission and processing. To reduce the exploration time, gaussian noise is added to the DRL. The performance boundary of the suggested approach was theoretically analyzed using three dimensions: system utility, energy consumption, and service time. The results of the simulation experiment demonstrate that the plan performs well overall in terms of reducing energy usage, improving transmission quality, and optimizing computational delay.