Wanniarachchi, Shashini ThamarasieShashini ThamarasieWanniarachchiTurau, VolkerVolkerTurau2024-09-122024-09-122024-08-2512th International Conference on Networked Systems, NETYS 20249783031673207https://hdl.handle.net/11420/49045Device-to-device (D2D) communication is a key technology introduced with the fifth generation (5G) communication standards. D2D aids massively in catering the continuously increasing demand on the connectivity needs. Despite that, there are several challenges involved with D2D communication. Among other issues, interference management is of high importance when coexisting with the traditional cellular network. Improper interference handling in the network leads to increased system delay, low sum rate and throughput along with reduced spectral efficiency. The strategy of allocating resources such as channel, spectrum and transmission power is a fundamental component leading to interference management. In our work, we address this challenge of interference management in 5G D2D communication. We first investigate the resource allocation methods for D2D communication suggested in the literature. Then we propose a dynamic resource allocation technique for 5G D2D users based on Expected SARSA reinforcement learning algorithm. With simulation results, we show that the system throughput performance can be uplifted to a greater extent while maintaining a high reliability by applying the proposed dynamic resource allocation algorithm.en5G device-to-device communicationExpected SARSAResource allocationMLE@TUHHTechnology::621: Applied Physics::621.3: Electrical Engineering, Electronic Engineering::621.38: Electronics, Communications EngineeringConference Paper10.1007/978-3-031-67321-4_16Conference Paper