PROMISING DIRECTIONS FOR DEVELOPING, IMPROVING AND APPLICATION SDN NETWORKS BASED ON ARTIFICIAL INTELLIGENCE METHODS

Abstract

This article explores promising directions for developing and enhancing Software-Defined Networking (SDN) using artificial intelligence (AI) methods. Software-Defined Networking (SDN) represents a modern networking approach that decouples control functions from data forwarding mechanisms, thereby enabling centralized oversight and flexible reconfiguration of network components. However, in conditions of limited resources, such as computing power, memory, or bandwidth, traditional SDN approaches face significant challenges in maintaining performance, security, and adaptability.

The study emphasizes the integration of machine learning (ML) algorithms for solving tasks related to access control, intelligent traffic distribution, detection of anomalies, and efficient allocation of cloud and edge resources. Particular attention is given to lightweight AI models that can operate on conventional hardware without requiring specialized equipment. Several practical cases are considered, including the use of clustering and classification algorithms to support decision-making in resource-constrained SDN nodes.

The paper also analyses the benefits of AI for improving the resilience and scalability of SDN controllers, including support for distributed intelligence and partial decentralization of control logic. This is especially relevant for hybrid infrastructures where centralized cloud services coexist with edge computing devices. Additionally, the article discusses the potential of AI in reducing energy consumption, enhancing quality of service (QoS), and enabling predictive analytics in dynamic network environments.

The results of this research highlight the relevance of AI-driven SDN for future communication systems, particularly in scenarios where access to computing and network resources is limited. The proposed approaches contribute to the vision of building smarter, more adaptive, and resource-efficient software-defined infrastructures.