A DNN-orchestrated cognitive radio-based power control scheme for D2D communication

Abstract

Device-to-device (D2D) communication demonstrates its efficacy across diverse domains, including military, surveillance, public safety, disaster management, content sharing, and local advertising, through its provision of proximity-based connections. Moreover, it elevates data rates, spectrum efficiency, and network throughput with efficient resource allocation while simultaneously reducing latency. However, it encounters challenges related to unregulated high-power consumption, leading to interference in D2D communication. Hence, in this paper, we proposed a deep neural network (DNN) and cognitive radio-driven efficient power control scheme for D2D communication. The proposed scheme operates in two distinct phases: the cognitive radio and the DNN-based phase. In this context, cognitive radio aids in selecting secondary users or cellular mobile equipment (CUE) through energy detection spectrum sensing. Subsequently, the DNN model is employed to optimize the transmission power of CUE in D2D communication. The effectiveness of the proposed scheme is assessed through various metrics, including loss, optimizers, learning rates, and sum rates.