A Comparative Study of Resource Allocation Algorithms Used in Network Virtualization Function

Abstract

A new networking technology in the telecom sector called network function virtualization lowers operating and capital costs while enabling network service deployment. In NFV, seeks to tackle challenges by utilizing standardized IT virtualization technology to combine various types of network equipment onto industry-standard, high volume servers, switches and storage systems, which may be situated in data centers, network nodes, or at the end user’s location. The Virtual Network Functions (VNFs) worked as software oriented approach, it creates a high flexible and dynamic network to meet more several demand along with a series of research challenges, such as, VNF management and orchestration, service chaining, VNF scheduling for low latency and efficient virtual network resource allocation with Network Function Virtualization Infrastructure (NFVI), among others. However, because network conditions and workloads are dynamic, effective resource management is still a major difficulty in NFV systems. This survey paper presents an overview of resource allocation algorithms in Network Functions Virtualization (NFV). We examine state-of-the-art approaches as Deep Reinforcement Learning (DRL), Parallel VNF Placement Framework (PVFP), RL Based Framework and Online Coordinated Resource Allocation (OCRA), analyzing speed, limitations, and suitability for dynamic environments. This paper has been prepared as an effort to reassess the research studies on the relevance of machine learning techniques in the domain of Network Function Virtualization.

Introduction

Traditionally, telecom networks relied on dedicated, physical hardware devices for each network function, leading to rigid service chains, slow product cycles, and dependency on specialized equipment. Network Function Virtualization (NFV) introduces a transformative approach by virtualizing these functions as software instances running on standard hardware, improving flexibility, agility, and cost-efficiency.

NFV’s architecture decouples network functions from hardware, enabling deployment on commercial servers, switches, and storage across various locations. However, realizing NFV’s potential faces challenges such as efficient resource management, autonomous scaling, and maintaining service reliability, especially critical for 5G networks and applications requiring high dependability.

Related Research Highlights:

• Several studies address NFV resource allocation, VNF scheduling, and service function chain (SFC) placement using heuristics, genetic algorithms, machine learning, and reinforcement learning to optimize performance, cost, and latency.

• Algorithms have been developed for dynamic VNF placement and migration, delay-aware scheduling, and coordinated resource allocation, often using mixed-integer programming, tabu search, or genetic approaches.

• Machine learning and deep reinforcement learning techniques have been applied to predict resource needs, optimize scaling, and reduce latency and resource consumption.

• Advanced models like multi-agent DRL, federated learning, and graph neural networks improve adaptability and efficiency in managing VNFs and service chains.

• Research also focuses on addressing real-time demands, online backup, and fault tolerance in edge and cloud environments.

Overall, NFV research is rapidly evolving with innovative algorithms and intelligent frameworks aimed at optimizing virtual network resource usage, improving quality of service, and enabling scalable, flexible telecom infrastructure for future technologies.

Conclusion

In this paper, a comprehensive review was conducted to examine resource allocation problems in NFV with various approaches. The implementation is carried out between RL, DRL, OCRA and PVFP algorithms for Virtual network functions across Multiple servers, to pivot on Memory, CPU and execution Time synthetic dataset featuring 5 servers and 10 VNFs is utilized, creating a controlled environment for testing. The dataset encompasses a range of resource needs for the VNFs and differing capacities for the servers, replicating real-world situations with diverse resources. The variety in the dataset ensures that each algorithm\'s strengths and weaknesses are highlighted.RL Coordination offers a compromise between efficiency and resource use, making it ideal for typical tasks. Optimal for Maximizing Resource Use: The DRL-Based Allocation ensures peak CPU utilization, making it suitable for tasks that require significant resources. Optimal for Simplicity: OCRA is efficient for basic allocation, but it may not fully utilize resources. Ideal for Parallelism: PVFP performs exceptionally well in parallel execution situations, although it tends to have longer execution durations. Choosing an algorithm must take into account the nature of the workload, the availability of resources, and the particular needs of the application setting.

References

[1] B.Han,V. Gopalakrishnan, L. Ji, and S. Lee, “Network function virtualization: Challenges and opportunities for innovations,” Communications Magazine, IEEE, vol. 53, no. 2, pp. 90–97, Feb 2015 [2] Mijumbi, R., Serrat, J., Gorricho, J. L., Bouten, N., De Turck, F., and Davy, S, ”Design and Evaluation of Algorithms for Mapping and Scheduling of Virtual Network Functions” , Network Softwarization (NetSoft), 2015 1st IEEE Conference on, pp.1-9, 2015 [3] Long Qu, Chadi Assi, Senior Member, IEEE and Khaled Shaban, Senior Member,”Delay-Aware Scheduling and Resource Optimization with Network Function Virtualization”, IEEE ,: DOI [4] A. Suzuki et al., “A method of coordinating multiple control algorithms for NFV,” in IEICE Tech. Rep., ser. IN2016-103, vol. 116, no. 485,2017, pp. 37–42, (in Japanese). [5] Nahida Kiran, Xuanlin Liu, Sihua Wang, Changchuan Yin,”Optimising resource allocation for virtual network functions in SDN/NFV-enabled MEC networks”,© 2021 IET Communications published by John Wiley &Sons Ltd on behalf of The Institution of Engineering and Technology [6] J. Zhang, Z. Wang, C. Peng, L. Zhang, T. Huang, and Y. Liu, “RABA:Resource-aware backup allocation for a chain of virtual network functions,”in Proc. IEEE Conf. Comput. Commun., 2019, pp. 1918–1926. [7] Zhiyuan Li , Lijun Wu, Xiangyun Zeng, Xiaofeng Yue, Yulin Jing, Wei Wu , and Kaile Su, Member, IEEE,”Online Coordinated NFV Resource Allocation via Novel Machine Learning Techniques”,IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT, VOL. 20, NO. 1, MARCH 2023. [8] L. Wang, Z. Lu, X. Wen, R. Knopp, and R. Gupta, “Joint optimization of service function chaining and resource allocation in network function Virtualization,” IEEE Access, vol. 4, pp. 8084–8094, 2016. [9] R.Mijumbi,S.Hasija,S. Davy, A. Davy, B. Jennings, and R. Boutaba, “Topology-aware prediction of virtual network function resource requirements,” IEEE Transactions on Network and Service Management, vol. 14, no. 1, pp. 106–120, 2017. [10] Stefan Schneider?,Narayanan Puthenpurayil Satheeschandran,Manuel Peuster , and Holger Karl,”Machine Learning for Dynamic Resource Allocation in Network Function Virtualization “, IEEE Conference on Network Softwarization (NetSoft 2020)2020. [11] M OHAMMAD BANY TAHA,, YOUSEF SANJALAWE , AHMAD AL-DARAISEH , SALAM FRAIHAT, AND SALAM R. AL-E’MARI ,”Proactive Auto-Scaling for Service Function Chains in Cloud Computing Based on Deep Learning “ IEEE VOLUME 12, 2024 Identifier 10.1109/ACCESS.2024.3375772. [12] Hurmat Ali Shah and Lian Zhao,” Multi-Agent Deep Reinforcement Learning Based Virtual Resource Allocation Through Network Function Virtualization in Internet of Things”,DOI 10.1109/JIOT.2020.3022572, IEEE Internet of Things Journal. [13] Hang Li, Luhan Wang, Xiangming Wen?, Zhaoming Lu, Jinyan Li,”MSV: An Algorithm for Coordinated Resource Allocation in Network Function Virtualization,IEEE Access ( Volume: 6)ISSN: 2169-3536. [14] Zhiwei Liu, Zhaogang Shu * , Shuwu Chen, Yiwen Zhong, Jiaxiang Lin,”Low-latency Virtual Network function Scheduling Algorithm Based on Deep Reinforcement Learning*”, 1389-1286/© 2024 Elsevier B.V.https://doi.org/10.1016/j.comnet.2024.110418. [15] Haojun Huang , Jialin Tian, Geyong Min , Member, IEEE, Hao Yin,”Parallel Placement of Virtualized Network Functions via Federated Deep Reinforcement Learning “,IEEE/ACM TRANSACTIONS ON NETWORKING, VOL. 32, NO. 4, AUGUST 2024. [16] M.A. Abdelaal, G.A. Ebrahim, W.R. Anis, Efficient placement of service function chains in cloud computing environments, Electronics 10 (3) (2021). [17] J. Chen, J. Chen, R. Hu, MOClusVNF: Leveraging multi-objective for scalable NFV orchestration, 2021 [18] Mahsa Moradi a , Mahmood Ahmadi a, , Latif PourKarimi b,”Virtualized network functions resource allocation in network functions virtualization using mathematical programming”,0140-3664/© 2024 Elsevier B.V.https://doi.org/10.1016/j.comcom.2024.107963. [19] R.Cohen,L.Lewin-Eytan,J-S.Naor,and D.Raz,”Near optimal placement of Virtual network functions “,in IEEE INFOCOM,2015-PP.1346-1354.

Copyright

Copyright © 2025 A. Kavitha, Dr. M. Gobi. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.