Generic potential field based distributed node coordination in flying adhoc network (FANET)

Abstract

Modern and contemporary technological developments have been the need of the hour in the present time. Extensive research showed that there has been significant enhancements in technologies that work on improving network paradigms like FANET and VANET creating evolutionary ideas. Latest studies have shown that, the research is budding in multi-drone communications with the swarm of UAVs called flying adhoc network (FANET). Not only are FANETs more reliable than single UAV but are also likely to be more efficacious and advantageous in finishing operational tasks. Interestingly, due to the high mobility characteristic, FANETs don’t have a fixed topology and expeditiously changes its topological structure which makes coordination between UAVs in the FANET arduous. The UAVs deployed for real time applications possess myriad of features such as on-board video streaming, streaming and more. The on-board systems on the UAVs utilise the protocol with high bandwidth, mobility, link stability and high energy consumption. However, the extant solutions used in the conventional adhoc network for node coordination cannot be applied in the FANET. Hence, this paper addresses the above mentioned issues by proposing a novel distributed node coordination algorithm for FANETs. The crux in this proposed algorithm is to make use of the General Potential Field based node coordination (GPFnc). The next stage involves validating the experimental results by developing two simulated environments—dynamic and static respectively in which , the performance metrics—feasibility and effectiveness of FANET were measured and analysed. The results of the experiment clearly demonstrated that the proposed GPFnc achieves enhanced scalability, reliability and fast network formation than the existing contemporary algorithms of MANET. Finally, the proposed GPFnc achieves a maximum of 63% reliable throughput with extremely low jitter and the latency is computed to be is 1.5 × times better than the present state-of-art algorithms.