An Overview of Security Threats Mitigation through Dynamic Bandwidth Allocation Algorithms in Passive Optical Networks

Abstract

Amid the ceaseless advancements in the telecommunication world, Passive Optical Network (PON) technologies have positioned themselves as paramount solutions in the sphere of global broadband access networks. These technologies offer several distinctive benefits, including remarkable bandwidth capabilities, impressive reach, passive nature, relative ease of maintenance and flexibility for potential future upgrades. Among these technologies, Gigabit PON (GPON) has been extensively adopted across Europe and America, while Ethernet PON (EPON) has found considerable favor in regions such as Korea and Japan. GPON, which notably functions as a potent "last-mile" communication system, leverages the unparalleled high-speed data transfer capabilities that are inherent to fiber-optic technology. However, with technological progress comes the escalating need for robust security measures. The complex landscape of network technologies has seen a rise in cyber threats, making security a paramount concern. Cyber threats can often take several forms, such as masquerading, where unauthorized entities impersonate legitimate users; packet replaying, where network packets are maliciously resubmitted; message modification, where information within network messages is unlawfully altered; and the more notorious Denial of Service (DoS) attacks. This study embarks on an exploratory journey into the impacts of these diverse security strategies on the overall performance of PON technologies. It provides an overview that considers the categorization of PON standards and scrutinizes their respective security schemes, providing insights into their effectiveness. This paper not only reviews existing methods for mitigating security threats in PONs but also introduces a novel Hybrid Security-Aware DBA (HSA-DBA) model. By incorporating machine learning, the HSA-DBA enhances adaptability and resilience in bandwidth allocation while addressing evolving security challenges. This approach ensures optimal bandwidth distribution, adheres to SLAs, and strengthens PONs against future threats, offering a significant advancement in secure and efficient PON architecture. The study examines performance metrics including upstream delay, delay variation, and frame loss across traffic classes T1, T2, T3 and T4.