Communication Control in Multiple UAV applications

In this paper, communication control algorithms are proposed for Unmanned air Vehicles (UAVs) operating in absence of central communication systems. Such situations arise in many scenarios. Operating over enem y territory or uneven terrain are a few examples. The communication network formed by these UAVs belongs to a larger class called ad-hoc networks. These networks are dynami c in nature and hence conventional algorithms of routing information cannot be applied. Most o f the algorithms used are heuristics based and in such algorithms, stability a nd convergence cannot be established. A Grossberg neural network (GNN) based approach is proposed to obtain routing of data packets. When there are many connections in a n etwork or all nodes are not equally capable, clustering of several nodes into one group is done. By clustering, number of connections and data loss can be reduced. GNN based algorithm s are implemented in a decentralized manner and have very low computation and storage re quirements. GNNs are dynamic in nature and hence are well suited for applications in ad-hoc networks. Any change in the ad-hoc network is absorbed by the GNN. The proposed algorithms are tested on a simple network simulator developed for this purpose. Satisfy ing results show the potential of the approach.