Analysis Of Connectivity In Opportunistic Sensor Netowrks

The rapid development of Wireless Sensor Network (WSN) technology makes muchof its applications to industrial production, agricultural control, environmentalmonitoring, health care, military defense and many other domains, the demand ofreal-time movement of sensor nodes speeded up the coming out of opportunistic sensornetworks. Nodes in WSNs are static once they were deployed, while in OpportunisticSensor Network (OSN), nodes can move randomly or according to certain rules. OSNcan meet the needs of node movement in surveillance scenes, which make up for thedeficiency of WSN in this area. As a basic nature of network, connectivity is theguarantee of normal communication. The connectivity of network, if can be obtained,could be useful for the design of data disseminations mechanisms, and be helpful for thenodes’ selection of routing protocols to forward data. But in OSNs, nodes’ movementresult to the dynamic changes in connectivity, which leads to the researches of theconnectivity in static networks are unsuitable for the connectivity analysis inopportunistic sensor network. So, it is meaningful to describe how well the network isconnected accurately.Firstly, this article introduces the concept and features of the OSN comprehensively,and generalizes Multi-region Opportunistic Sensor Network (MOSN) model. On thebasis of the model, we put forward the study of connectivity in OSN, and propose theconcept of Regional Connectedness (RC) timely. This concept is used to describe theconnectivity of each region. Through a large number of experiments and analysis, theinfluence factors for RC are discussed in detail, and the major factors are selected in themean of grey relational analysis for the various factors and RC. And mathematicaldefinition of RC is obtained through stepwise regression analysis method, whichprovides a reference for improving the regional connectivity.Then, from the view of the forwarding message for each region by ferry node, wedefine the Ferry Connectedness (FC), which is used to describe the ferry node’s abilityof carring and forwarding message. To accurately master the changing rule of the FC,the similar method is adopted to study the RC comprehensively, and the regression expression of RC is obtained, which provides basis for improving the Ferryconnectivity.Finally, on the basis of Regional Connectedness and Ferry Connectedness, thenetwork connectivity model is constructed by analyzing the pathes of ferry nodes, andthis provides the basis for describing the network connectivity.Experimental results show that the proposed Regional Connectedness is helpful fordepicting the connectivity of region, and Ferry Connectedness can describe carrying andreflect the connectivity of the network preferably.