Performance Evaluation of Routing Protocols and ML-based Enhancements for UAV-Assisted Post-Disaster Communication Networks

During natural disasters, the existing communication systems collapse, and the disaster-affected areas become disconnected without any means of exchanging information. The collapse of existing communication networks poses challenges for First-Responders (FRs) in locating survivors during Search and Rescue (SAR) operations and for survivors to communicate for emergency aid. To alleviate post-disaster consequences and save lives, Uncrewed Air Vehicles (UAVs), commonly known as drones, can be employed to establish adaptable and reliable emergency communication networks. UAVs offer portability and rapid deployment, making them effective in crises. This thesis presents two main contributions: 1) Evaluation and Comparison of Routing Protocol Performance: The research evaluates and compares the performance of four network routing protocols – Ad Hoc On Demand Distance Vector (AODV), Optimized Link State Routing (OLSR), Destination Sequenced Distance Vector (DSDV), and Dynamic Source Routing (DSR) - that are suitable for UAV-assisted communication networks. These protocols are assessed through extensive NS-3 simulation experiments, focusing on key functional requirements such as delay, throughput, packet delivery ratio, and jitter. 2) Integration of Machine Learning (ML) specifically K-means for CC-AODV enhancement, we hypothesize that K-means will enhance CC-AODV by refining the route discovery process and minimizing redundant packet transmissions under high-demand conditions. Additionally, non-functional requirements such as scalability, reliability, and adaptability are examined in the work.