ACPM: An associative connectivity prediction model for AANET

Abstract

Aeronautical ad hoc networks (AANETs) can have a hybrid topology of intermittently connected clusters and mesh network. The hybrid network topology compounded with highly dynamic nature of AANET leads to variable connectivity in the network. The connectivity in the network for direct air-to-air communication between aircrafts is primarily a function of velocity of air vehicles, position of air vehicles, direction of flight, range of communication and congestion. In this paper, we present a connectivity prediction model for the AANET. The proposed model does a space time analysis of connectivity in the regions of AANET. The model can identify areas of low connectivity in a region by grading connectivity in the AANET. The connectivity prediction complements the change in state of an aircraft as it joins the AANET in a cluster or mesh. The model directs idle or migrating members of the network to the higher regions of connectivity. Connectivity gradation reduces network setup time in the events of network disruptions, strengthening the network's self-healing capability. FCM clustering is used for translating 3D topology of the network in conjunction with network traffic to multiple domains, that represent an aircraft.