Retaining connectivity in multi-task communications network with multiple agents: Connectability theory approach

Abstract

Practical retention of mobile ad hoc network communications via connectability theory is presented and compared to predictive modeling techniques. Network communication disruptions is prevented by driving relay agents to computed waypoints using sliding mode and LQ control, or using predictive modeling to optimally control relay agents. The connectability matrix is used to determine where future node isolation will occur. This paper expands the connectability matrix concept into connectability theory to not only predict node isolation, but to directly compute the waypoints for relay agents. The existing methods of computing waypoints, of controlling robotic routers to form so called network bridges, and the outcome of predictive modeling are shown to be special cases of the proposed connectability theory. Also, case studies and simulations are presented to show this connectability theory's utility in various network configurations. © 2013 AACC American Automatic Control Council.