MANET routing with chutes and ladders

The general problem of communications routing in mobile ad hoc networks (MANETs) can be stated simply as the search for the ldquobestrdquo path from point A to point B by actively seeking out the ladders and avoiding the chutes in the topological landscape. We propose a model of this landscape as a Markov random field in order to provide X-awareness, where X is interference or coverage, or some other spatially-based property of the network to be observed and controlled. A distributed and computationally efficient means has been developed to obtain the relevant indicators of the underlying potential field. This paper builds on prior work that used power-spatial density to identify and avoid interference hot spots (chutes) using routing. Alternatively, for battery-limited sensor networks, we route through regions with surplus monitoring lifetime (ladders). The path cost function for routing is modified to include the indicator for either of these potentials. It is shown that the two indicators are identical in form but have different signs.