Alleviating the Siphon Effect on Age of Information in D2D-Enabled Social IoT Networks

The increasing demand for real-time information freshness and efficient communication in Social Internet of Things (SIoT) environments for consumer electronics application necessitates the development of optimized resource allocation strategies. Device-to-Device (D2D) communication especially fits, where numerous nearby devices require timely updates. In this paper, we propose the actor-critic reinforcement learning-based age of information optimization (PACAO) algorithm, which leverages the Siphon Effect to dynamically regulate resource allocation, mitigating network congestion and optimizing data dissemination. The proposed algorithm utilizes an actor-critic reinforcement learning framework to minimize Age of Information (AoI) and energy consumption to make communication more efficient. Simulation results demonstrate that proposed algorithm minimizes AoI by 57.67% and energy consumption is minimized by 25.58% compared to benchmark algorithms. Furthermore, the proposed approach effectively enhances network throughput, ensuring seamless and low-latency D2D communication in SIoT environments. Simulation result confirms that proposed algorithm outperforms benchmark techniques by adapting to varying network conditions and minimizing unnecessary energy consumption, making it highly suitable for consumer electronics applications in SIoT networks.