Interference Aware Routing for Tactical Networks using Power Spatial Density

Performance of wireless communications and sensor networks is subject to limitations imposed by available bandwidth, power, or interference levels. Nodes may be able to report their consumption of these resources as congestion indicators so that appropriate routing decisions can be made to reduce their load and guarantee certain levels of service. To the extent that the congestion involves only one node or is due to logical flows between nodes, this approach is adequate. However mutual interference is a function of the power spatial density of nodes and therefore some link qualities are correlated. This suggests that routing with regional congestion control may be more effective than with nodal congestion control - provided the appropriate metrics can be obtained. This paper introduces a new approach to interference aware routing that results in more effective load balancing by using power spatial density as a predictor of interference. The analogy of temperature is used in a novel and computationally efficient algorithm to detect "hot spots" that may emerge anywhere in the domain of the network. The detection algorithm provides a regional congestion indicator to a routing algorithm as an additional penalty, so that an entire region may be avoided when making routing decisions. Operation requires the periodic exchange of indicators, as well as position and transmitted power reports between nodes, which may be accommodated by many tactical networking protocols. Preliminary simulation results are presented demonstrating that the algorithm exhibits good dynamic behavior, and is robust with respect to stale reports.