Cloud Computing and Machine Learning-Driven Security Optimization and Threat Detection Mechanisms for Telecom Operator Networks

Abstract

Telecom operator networks are increasingly migrating toward cloud-native architectures enabled by network function virtualization (NFV) and software-defined networking (SDN). This transformation brings flexibility but also exposes new security challenges such as virtualization vulnerabilities, multi-tenant isolation, and dynamic threat propagation. This study proposes a machine learning-driven security optimization framework that integrates adaptive threat detection with reinforcement learning-based policy control. The framework formulates network security management as a multi-objective optimization problem balancing detection accuracy, response latency, and resource efficiency. A layered architecture enables dynamic coordination among detection, orchestration, and policy modules, supporting intelligent and self-adaptive defense in telecom environments. Simulation-based validation verifies the framework's logical feasibility and adaptability, providing a theoretical foundation for intelligent and automated network protection.