Bulletin of Electrical Engineering and Informatics Vol. 13, No. 4, August 2024, pp. 2418~2426 ISSN: 2302-9285, DOI: 10.11591/eei.v13i4.5773  2418 Journal homepage: http://beei.org Resource allocation for device-to-device communications underlaying uplink cellular networks Ahtirib Haroune 1 , Hettiri Messaoud 1 , Lejdel Brahim 2 1 Department of Electrical Engineering, Faculty of Exact Sciences, University of El-Oued, El-Oued, Algeria 2 Department of Computer Science, Faculty of Exact Sciences, University of El-Oued, El-Oued, Algeria Article Info ABSTRACT Article history: Received Jan 14, 2023 Revised Nov 10, 2023 Accepted Feb 12, 2024 Underlying cellular networks, device-to-device (D2D) communications are a practical network technology that can increase power efficiency and spectrum usage for close-proximity wireless services and applications. However, D2D link interference, when sharing resources with cellular users (CUs), poses a major challenge in such distribution situations. In this research, we primarily utilize wireless channel data that exhibits slowly shifting large-scale fading to conduct spectrum sharing and power allocation. The overall ergodic capacity of all cellular user equipment (CUE) links is initially considered as the optimization target in order to maximize the overall throughput of CUE links while ensuring the reliability of each D2D link. Then, the expansion of the minimum ergodic capacity is measured to ensure a more consistent capacity performance across all CUE links. We utilized algorithms that are resilient to channel fluctuations and produce optimal resource allocation. We use MATLAB, and the computer simulation validates its intended performance. Keywords: Cellular networks Device-to-device communication Ergodic capacity Power allocation Resource allocation This is an open access article under the CC BY-SA license. Corresponding Author: Lejdel Brahim Department of Computer Science, Faculty of Exact Sciences, University of El-Oued BP: 127, El-Oued, Algeria Email: lejdel.brahim@gmail.com 1. INTRODUCTION One of the technological components of the third generation partnership projects (3GPPs) long term evolution advanced (LTE-A) is device-to-device (D2D) communication [1]. When it comes to D2D communication, nearby cellular users (CUs) can communicate with each other directly through a direct link instead of sending and receiving signals via a cellular base station (BS). D2D users remain under BS control while engaging in direct communication. When interacting directly with one another, CUs in close proximity can save energy and resources compared to routing through a BS. Furthermore, because direct contact occurs over short distances, D2D users can benefit from fast data rates and minimal transmission delays [1]. Another benefit of D2D communication is that it reduces stress on the network by bypassing the BS and other network components, allowing cellular traffic to be directly transmitted between users. This helps to increase the network’s effective capacity. Additional advantages and applications are covered in [2]. In this paper, we utilize a random D2D underlaid cellular network model to analyze the performance of an improved power control approach and channel allocation. The study focuses on a diverse scenario where the distance between D2D pairs is randomly modeled. Each D2D transmitter chooses the power level for transmission based on the channel conditions, particularly the distance-dependent path-loss between D2D pairs, with the aim of maximizing its own D2D link rate. We also take into account the fact that one cellular user may share resources with the D2D links.