The Limits of Motion Prediction Support for Ad hoc Wireless Network Performance

A fundamental understanding of gain provided by motion prediction in wireless ad hoc routing is currently lacking. This paper examines benefits in routing obtainable via prediction. A theoretical best-case non-predictive routing model is quantified in terms of both message overhead and update time for non-predictive routing. This best- case model of existing routing performance is compared with predictive routing. Several specific instances of predictive improvements in routing are examined. The primary contribution of this paper is quantification of predictive gain for wireless ad hoc routing.