Machine Learning-Based Strategy to Detect Sybil Attacks on Mobile Ad Hoc Networks

Resumen

Mobile ad hoc networks pose significant security challenges due to their decentralized nature and limited resources. These networks allow nodes to join or leave freely, without any central authority controlling their entry or exit. Dynamic multi-hop networks can be categorized as either stationary or mobile, and provide rapid and effortless access to information. However, the unpredictable and constantly changing topology of MANETs, along with the dispersion and self-organization of nodes, can make it difficult to predict how the network will evolve. Mobile ad hoc networks are inherently less secure than wired networks due to vulnerabilities in security and constraints in energy resources. In contrast to fixed networks, mobile ad hoc networks use wireless transmission, which exposes them to higher loss rates, delays, and jitter. Moreover, the nodes in these networks rely on finite energy sources, such as batteries. To gain unauthorized access to a large portion of the system, a Sybil attack refers to the ability of a small group of actors to mimic several peers. This study proposes a machine learning-based strategy to identify a Sybil attacks in MANETs by collecting network metrics such as traffic characteristics, communication patterns, and node behaviours.