جايابی چند هدفه و مبتنی بر يادگيری ماشين کنترلکنندهها در شبکههای مبتنی بر نرمافزار
محورهای موضوعی : مهندسی برق و کامپیوترمبین قلی زاده 1 , ناصر مزینی 2 *
1 - دانشكده مهندسی كامپيوتر، دانشگاه علم و صنعت ايران، تهران، ایران
2 - دانشكده مهندسی كامپيوتر، دانشگاه علم و صنعت ايران، تهران، ایران
کلید واژه: جايابي کنترلکنندهها, شبکههاي مبتني بر نرمافزار, يادگيري ماشين.,
چکیده مقاله :
امروزه با رشد روزافزون شبکههاي کامپيوتري و افزايش پيچيدگي ساختارهاي ارتباطي، نياز به مديريت بهينه منابع، کاهش هزينههاي عملياتي و نگهداري به چالشي اساسي تبديل شده است. شبکههاي مبتني بر نرمافزار با جداسازي لايه داده و کنترل، انعطافپذيري، مقياسپذيري و مديريت متمرکز را فراهم آورده است. در لايه کنترل، کنترلکنندهها به عنوان عنصر اصلي مديريت و تصميمگيري در شبکه عمل کرده و چالش اصلي در اين حوزه جايابي بهينه آنها در شبکه است. نحوه توزيع و استقرار کنترلکنندهها تاثير مستقيمي بر عملکرد شبکه همچون کاهش تأخير، برقراري تعادل بار و افزايش پايداري شبکه دارد. در این مقاله، به ارائه يک روش چندهدفه مبتني بر يادگيري ماشين براي جايابي کنترلکنندهها در شبکههاي مبتني بر نرمافزار پرداخته شده است. به این منظور، مسئله را با هدف کمينهکردن تأخير لايه کنترل و تأخير مياندامنهاي، بار لايه کنترل و هزينه شکست گرهها و پيوندهاي توپولوژي شبکه و همچنين بيشينهکردن توان عملياتي مدلسازی شده است. همچنین، با توجه به پیچیدگی محاسباتی مسئله، یک الگوریتم حل مبتنی بر يادگيري تقويتي عميق مورد استفاده قرار گرفته است. نتایج شبیهسازیها در چند توپولوژی مختلف نشان میدهد که الگوریتم پیشنهادی توانسته است تأخير لايه کنترل، تأخير درون دامنهاي و توان عملياتي را به ترتیب تا 20، 30 و 15 درصد نسبت به کارهای پیشین بهبود بخشد. بدین ترتیب، رویکرد پیشنهادی علاوه بر پوشش اهداف متداول همچون کاهش تأخیر و توازن بار، معیارهای مهمی مانند پایداری و تحملپذیری شکست را نیز در فرآیند تصمیمگیری جایابی در نظر میگیرد.
With the rapid growth of computer networks and the increasing complexity of communication infrastructures, the need for efficient resource management and reduced operational and maintenance costs has become a major challenge. Software-Defined Networking (SDN), by decoupling the control and data planes, provides flexibility, scalability, and centralized management. In the control plane, controllers act as the core components responsible for network management and decision-making, and their optimal placement remains a fundamental challenge. The distribution and deployment of controllers directly affect network performance parameters such as delay reduction, load balancing, and stability enhancement. In this paper, a multi-objective machine learning–based approach is proposed for controller placement in SDN. The problem is formulated with the objectives of minimizing control-plane delay, inter-domain delay, controller load, and the failure cost of network nodes and links, while maximizing network throughput. Considering the computational complexity of the problem, a deep reinforcement learning (DRL) algorithm is employed as the solution approach. Simulation results on multiple network topologies demonstrate that the proposed algorithm improves the control-plane delay, inter-domain delay, and network throughput by approximately 20%, 30%, and 15%, respectively, compared with previous works. Therefore, the proposed approach not only addresses common objectives such as delay minimization and load balancing but also incorporates crucial factors such as stability and fault tolerance into the controller placement decision-making process.
[1] C. Urrea and D. Benítez, "Software-defined networking solutions, architecture and controllers for the industrial internet of things: A review," Sensors, vol. 21, no.19, Article ID: 6585, 2021.
[2] Y. Maleh, Y. Qasmaoui, K. El Gholami, Y. Sadqi, and S. Mounir, "A comprehensive survey on sdn security: threats, mitigations, and future directions," Journal of Reliable Intelligent Environments, vol. 9, no. 2, pp.201-239, 2023.
[3] A. N. Alhaj and N. Dutta, "Analysis of security attacks in sdn network: A comprehensive survey," Contemporary Issues in Communication, Cloud and Big Data Analytics, pp.27-37, 2022.
[4] K. Nisar, et al., "A survey on the architecture, application, and security of software defined networking: Challenges and open issues," Internet of Things, vol. 12, Article ID: 100289, 2020.
[5] S. Ahmad and A. H. Mir, "Scalability, consistency, reliability and security in sdn controllers: a survey of diverse sdn controllers, " Journal of Network and Systems Management, vol. 29, pp.1-59, 2021.
[6] B. Isong, R. R. S. Molose, A. M. Abu-Mahfouz, and N. Dladlu, "Comprehensive review of sdn controller placement strategies," IEEE Access, vol. 8, pp.170070-170092, 2020.
[7] M. Rahouti et al., "SDN security review: Threat taxonomy, implications, and open challenges," IEEE Access, vol. 10, pp. 45820-45854, 2022.
[8] E. H. Bouzidi, A. Outtagarts, R. Langar, and R. Boutaba, "Dynamic clustering of software defined network switches and controller placement using deep reinforcement learning," Computer Networks, vol.207, Article ID: 108852, 2022.
[9] K. Saeed and M. O. Ullah, "Toward reliable controller placements in software-defined network using constrained multi-objective optimization technique," IEEE Access, vol. 10, pp.129865-129883, 2022.
[10] C. Liao, et al., "Modecp: A multi objective based approach for solving distributed controller placement problem in software defined network," Sensors, vol.22, no.15, Article ID: 5475, 2022.
[11] M. M. Kazemian and M. Mirabi, "Controller placement in software defined networks using multi-objective antlion algorithm, " The Journal of Supercomputing, vol. 78, no. 4, pp.5626-5649, 2022.
[12] Y. Wu, S. Zhou, Y. Wei, and S. Leng, "Deep reinforcement learning for controller placement in software defined network," in Proc. IEEE Conf. on Computer Communications Workshops, pp.1254-1259, Toronto, Canada, 6-9 Jul. 2020.
[13] H. Mostafaei, M. Menth, and M. S. Obaidat, "A learning automaton-based controller placement algorithm for software-defined networks," in Proc. IEEE Global Communications Conf., 6 pp., Abu Dhabi, United Arab Emirates, 9-13 Dec. 2018.
[14] S. S. Nande, O. Lhamo, S. Biswas, R. Bassoli, and F. H. Fitzek, "Quantum machine learning for controller placement in software defined networks," in Proc. 28th European Wireless Conf., pp.382-387, Rome, Italy, 2-4 Oct. 2023.
[15] S. Rahman, et al., "Virtualized controller placement for multi-domain optical transport networks using machine learning," Photonic Network Communications, vol. 40, pp.126-136, 2020.
[16] V. Srivastava and R. S. Pandey, "Machine intelligence approach: To solve load balancing problem with high quality of service performance for multi controller based software defined network," Sustainable Computing: Informatics and Systems, vol. 30, Article ID: 100511, 2021.
[17] C. Kurra, V. P. Dadi, A. Awasthi, V. Janyani and R. B. Battula, "CAPFUL - Coherent dynamic controller placement method using deep reinforcement learning for SDN," in Proc. Int. Conf. on Communication, Computing, Networking, and Control in Cyber-Physical Systems, Dubai, United Arab Emirates, pp. 331-336, 10-12 Jun. 2025.
[18] B. Li, et al., "DDPG-based load-aware QoS guaranteed SDN controller placement for Internet of vehicles," IEEE Internet of Things Journal, vol. 12, no. 24, 15 Dec. 2025.
[19] A. Kumari, A. Roy and A. Singh Sairam, "Optimizing SDN controller load balancing using online reinforcement learning," IEEE Access, vol. 12, pp. 131591-131604, 2024.
[20] M. Farhan, et al., "ReCAP: Reliability–capacity aware joint controller placement and routing using a hybrid AI approach," IEEE Trans. on Reliability, vol. 74, no. 4, pp. 5686- 5700, 2025.
[21] J. Hyun and J. W. Hong, "Knowledge-defined networking using in-band network telemetry," in Proc. 19th Asia-Pacific Network Operations and Management Symp., pp. 54-57, Seoul, South Korea, 27-29 Sept. 2017.
[22] P.T.A. Quang, Y. Hadjadj-Aoul, and A. Outtagarts, "A deep reinforcement learning approach for VNF forwarding graph embedding," IEEE Trans. on Network and Service Management, vol. 16, no 4, pp. 1318-1331, Dec. 2019.