IRSML: An intelligent routing algorithm based on machine learning in software defined wireless networking

Abstract

In software-defined wireless networking (SDWN), the optimal routing technique is one of the effective solutions to improve its performance. This routing technique is done by many different methods, with the most common using integer linear programming problem (ILP), building optimal routing metrics.These methods often only focus on one routing objective, such as minimizing the packet blocking probability, minimizing end-to-end delay (EED), and maximizing network throughput. It is difficult to consider multiple objectives concurrently in a routing algorithm. In this paper, we investigate the application of machine learning to control routing in the SDWN. An intelligent routing algorithm is then proposed based on the machine learning to improve the network performance. The proposed algorithm can optimize multiple routing objectives. Our idea is to combine supervised learning (SL) and reinforcement learning (RL) methods to discover new routes. The SL is used to predict the performance metrics of the links, including EED quality of transmission (QoT), and packet blocking probability (PBP). The routing is done by the RL method. We use the Q-value in the fundamental equation of the RL to store the PBP, which is used for the aim of route selection. Concurrently, the learning rate coefficient is flexibly changed to determine the constraints of routing during learning. These constraints include QoT and EED. Our performance evaluations based on OMNeT++ have shown that the proposed algorithm has significantly improved the network performance in terms of the QoT, EED, packet delivery ratio, and network throughput compared with other wellknown routing algorithms.

Cognitive routing in software defined networks using learning models with latency and throughput constraints

“…The environment's reinforcement signal is an appraisal of the outcome, good or poor, rather than instruction on how to generate the desired behaviour. According to our findings, RL is typically employed to promote robustness and scalability [18], [19], and it enables choosing routes or route optimisation in SDNs [20], [21]. When delay minimise and throughput enhancement are used as the operation and maintenance approach for deterministic policy gradient (DDPG) routing optimization mechanism (DROM) [22], the network's performance is enhanced with reliable and superior routing services, and convergence and effectiveness are boosted.…”

A novel and effective method for solving the router nodes placement in wireless mesh networks using reinforcement learning

“…Based on the rewards acquired in the prior learning, the agent selects the action that provides the best reward in the following learnings. The total reward for taking the a t action in s t state is Q ( s t , a t ), which is typically determined by the Q-learning algorithm as follows [ 36 ]: where α and γ ∈ [0, 1] are the learning rate and the discount factors, respectively.…”

“…The RL has recently been successfully employed to solve technical challenges in wireless communication such as routing [ 27 , 36 ], topology management [ 37 ], and resource allocation. In this study, we use RL to solve the RNP problem.…”

Một thuật toán định tuyến cân bằng năng lượng trong mạng cảm biến không dây dựa trên SDN

“…Các giao thức điều khiển hoạt động của hệ thống mạng như định tuyến, báo hiệu, điều khiển tô-pô, v.v được thực hiện tập trung tại bộ điều khiển SDN. Mô hình WSN sử dụng SDN đã được triển khai để cải tiến hầu hết các giao thức điều khiển trong mạng WSN, điển hình như các giao thức định tuyến [4,5,[10][11][12], điều khiển tô-pô [6,7,13], điều khiển phân cụm [8,9]. Trong đó, định tuyến sử dụng SDN là chủ đề nhận được sự quan tâm của nhiều nhóm nghiên cứu trong nước và trên thế giới thời gian gần đây.…”