Path selection and resource allocation for 5G multi-hop D2D networks

Multi-hop Device-to-Device (D2D) communication has emerged as one of the promising technologies to improve the network capacity of fifth generation (5G) networks. Existing research works focus on either increasing the throughput by streamlining radio-resource and power allocation approach or finding the best suitable path based on next-hop distance. However, there is very limited research work that covers the selection of the best path and resource allocation in multi-hop D2D communication. Therefore, in this paper, we proposed a path selection and resource allocation method. The proposed path selection method is based on the Q learning model in which the cumulative reward is calculated from source to destination. The proposed resource allocation method is divided into two-stage. In the first stage, a Max Mean Far (MMF) method is used to allocate the resource block in forming a multi-hop D2D. In the second stage, a hybrid combination of particle swarm optimization and gray wolf optimization (HPSOGWO) is applied over the first stage results. The HPSOGWO maximizes the system throughput by optimizing the power distribution of cellular and D2D users. In comparison to previously proposed approaches, this research process and findings ensure that system consistency and implementation are improved in all aspects. © 2022 Elsevier B.V.