Variable rate adaptive modulation (VRAM) for introducing small-world model into WSNs

Abstract

Data communication has a strong impact on the design of a Wireless Sensor Network (WSN), since the data transmission energy cost is typically higher than the data processing cost. In order to reduce the data transmission cost, small world phenomenon is introduced into WSNs. Networks that do not have the small world structure can be converted to achieve a small world property by the addition of few extra links. The problem is that most large scale WSNs are inherently unstructured and a node has no precise information of the overall model of the network and thus has to rely on the knowledge of its neighbor. For this reason, in most unstructured networks, information is propagated using gossiping. In this paper, we exploit this information propagation mechanism and use Neighbor Avoiding Walk (NAW), where the information is propagated to node that has not been visited previously and which is not the neighbor of a previously visited node. Using this, a novel approach is presented, in which nodes with highest betweenness centrality form a long distance relay path by using a lower order modulation scheme and therefore resulting in a relatively reduced data rate, but maintaining the same bit error rate. Our empirical and analytical evaluations demonstrate that this leads to a significant reduction in average path length and an increase in node degree.