CUDA-supported 5G multi-access edge computing modifications on 5G-air-simulator

(2025) 2025:29 Erdogan and Ozsoy J Wireless Com Network https://doi.org/10.1186/s13638-025-02438-z EURASIP Journal on Wireless Communications and Networking Open Access RESEARCH CUDA‑supported 5G multi‑access edge computing modifications on 5G‑air‑simulator Hasan Tugrul Erdogan1*   and Adnan Ozsoy1 *Correspondence: 1 Department of Computer Engineering, Hacettepe University, Beytepe Kampüsü, No:21, 06800 Ankara, Çankaya, Turkey Abstract In the recent years, 5G radio and core networking researches have become very popular because of its potential to be one of the promising technologies in the future with its unique use cases. Because of the specific nature of the 5G radio and core infrastructure, researchers in this area need special simulators and development sets to evaluate their hypothesis. However, these tools are not enough to cover the needs of the research areas of 5G architecture, and there are open areas. With this work, we are trying to improve one of the commonly used 5G simulator’s, 5G-air-simulator [1], simulating abilities by adding GPGPU capable MEC (Multi-access Edge Computing) support to it, which is a first in the literature. Keywords: 5G simulator, MEC (multi-access edge computing), System-level simulation 1 Introduction 5G systems are waiting to become parts of our lives in the near future in many use cases. These valuable use cases are commonly constructed on new opportunities of 5G systems such as very low latency and high bandwidths of large user groups. An advantageous design of 5G systems to achieve low latency and high bandwidth is placing the server application near user-equipments (UEs). This server application placement strategy points to Multi-access Edge Computing (MEC) concept. MEC incorporates general-purpose computing servers at network edge instead of centralized server locations. Alongside the high bandwidth potential of MEC used scenarios, they have also some extra advantages like bringing an x86-based General-Purpose Graphic Processing Unit (GPGPU) programming capable infrastructure to operate on telecommunication network. As shown in Fig. 1, MEC framework places at edge side of regular 5G infrastructure. The traffic between core network and base station can be directed over MEC host based on intelligent network routing managers. As the result of this intelligent network routers, MEC host become capable to serve different types of services for 5G users, operators and telecommunication infrastructure share holders. In the literature, there is a gap between theoretically defining the MEC architecture and physically presenting a system-level simulation. So with this work, we are proposing a system-level simulator having MEC architecture with GPGPU programming capability. © The Author(s) 2025. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/. Erdogan and Ozsoy J Wireless Com Network (2025) 2025:29 Fig. 1 MEC framework places at edge side of regular 5G infrastructure We have used 5G-air-simulator [1] as the base system-level 5G simulator because of its modular design and researcher-friendly licensing policies. We propose a simulator that simulates a GPGPU powered MEC platform from the perspective of Nvidia and also GPGPU-supported European Telecommunications Standards Institute (ETSI) MEC simulation ability. With both MEC implementation, our proposed system can respond to a broader range of MEC researchers’ needs. Our proposition has contributions to the literature as follows: • Implementation of a simulator for GPGPU-supported MEC environment – Both with Nvidia perspective MEC architecture implementation – Both with ETSI MEC reference design perspective MEC architecture implementation • Definition of some specialized 5G use cases for a GPGPU-supported MEC infrastructure • To our best knowledge, all of these contributions are first in the literature. In the following section, we describe related works in the area. In the third section, we briefly share the design of 5G architecture, and with the fourth section, we explain ETSI MEC with its design and interface details. The fifth section gives evaluation in some key works which add value to future 5G use cases. In the sixth section, we explain our proposed design. In the seventh section, we demonstrate the execution results of our modification. We also share the performance results of selected algorithms by comparing Central Processing Unit (CPU) and Graphics Processing Unit (GPU) results, respectively. Finally, we conclude the paper with last remarks and future work. 2 Related works In the simulator world, a grouping is made around two different concepts, System-level simulators and Link-level simulators [2]. While System-level simulators focus on data and the transmission of that data as implementation, it is not concerned with the fact that the transmission paths of simulator are identical to the real world. Thus, the provision of the service and the relationship of the service provided with the upper-level variables (number of users, interactions between the service provider and the service receiver, etc.) can be modeled. Link-level simulators, on the other hand, aim to ensure that all interfaces and layers which transport the data are the same with the real world. While simulators in this class can model the interactions of the interfaces, but they are not able to present the interactions of service and service quality parameters (mainly since they are limited the number of users because of complicated nature of link-level simulating). Page 2 of 19 Erdogan and Ozsoy J Wireless Com Network (2025) 2025:29 Following a brief introduction to the general classification criteria, System-level simulators and Link-level simulators, a group of 5G simulators work is described below. The 5G K-SimNet [3] simulator is not only a System-level simulator but can also be classified as a Link-level simulator. 5G K-SimNet is a project developed based on ns-3 [4] simulator so it extends ns-3 simulator by adding 5G technologies an (...truncated)