Modeling Search Costs in Wireless Sensor Networks

We develop approximate closed-form expressions of expected minimum search energy costs for data-centric wireless sensor networks showing the search performance with respect to the network size N and the number of randomly placed copies of the target event r. We consider both unstructured sensor networks, which use blind sequential search for querying, and structured sensor networks, which use efficient hash-based querying. We also consider two kinds of deployments: a fixed transmit power (FTP) model and the geometric random graph (GRG) model. We find that the search cost of unstructured networks under the FTP deployment is proportional to N and inversely proportional to (r + 1) regardless of the spatial dimension d in which nodes are deployed, while that of the GRG is proportional to N(log N)^eta/d/r + 1 where eta is the path-loss exponent. The search cost of structured networks under the FTP deployment is found to be proportional to ^dradicN/^dradicr, while that of the GRG deployment is proportional to ^dradicN(log N)^eta-1/^dradicr. In all cases, we also provide bounds on the coefficient of proportionality.