A Novel Quality-aware Cross-layer Framework for Adaptive Routing in MANETs
Mobile Ad hoc Networks (MANETs) have emerged as a prominent solution for providing communication, especially in scenarios where infrastructure-based networks are absent or impractical. With advances in mobile and wireless communication technology and their potential applications, infrastructure-less wireless networks and ad-hoc networks have attracted much attention from the research community and industry. Mobile ad hoc networks are formed by an autonomous group of mobile devices such as smartphones, laptops, and tablets that communicate with each other through wireless links (such as Wi-Fi or Bluetooth) without the use of any pre-installed infrastructure. That is, this infrastructure-less network type is designed to be both self-organized and self-configured, and without any central administration. This means that they do not impose any initial cost for setting up base stations or maintenance costs compared to standard networks that have both. However, the dynamic nature of MANETs, characterized by node mobility and varying network topologies, poses significant challenges to efficient and reliable routing that can negatively affect the quality of service (QoS). This thesis presents a novel quality-aware cross-layer framework that addresses the broadcasting problem and path selection challenge in reactive routing protocols in MANETs by employing fuzzy logic and Q-learning. To overcome the routing challenges in a high-density network environment and improve the quality of service, the proposed framework leverages fuzzy logic to select a subset of nodes by considering their conditions (i.e. available bandwidth, signal strength, queue congestion, MAC layer collision, node’s remaining energy, and local network density) to forward broadcast packets. The proposed framework also incorporates a Q-learning-based technique to evaluate path quality, which is used to dynamically adjust routing decisions based on the quality of intermediate nodes. Simulations and performance evaluations were conducted to validate the effectiveness of the proposed approaches. The results demonstrate that the proposed methods outperform traditional approaches in terms of network throughput, end-to-end delay, and packet loss.
Safari, F., Savi�?, I., Kunze, H., Ernst, J., & Gillis, D. (2023, June). A Review of AI-based MANET Routing Protocols. In 2023 19th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob) (pp. 43-50). doi: 10.1109/WiMob58348.2023.10187830
Safari, F., Savi�?, I., Kunze, H., & Gillis, D. (2023). �??The diverse technology of MANETs: A survey of applications and challenges. Int. J. Future Comput. Commun, 12(2) : 37-48. DOI: 10.18178/ijfcc.2023.12.2.601