Hierarchical aggregation of distributed data for sensor networks

Wireless sensor networks consist of a lot of nodes whose resources such as battery power, CPU and bandwidth are very restricted in general. This paper describes an efficient and scalable method for aggregating and collecting spatially distributed data whose values change in a large scale network. A geographic area where sensors exist is virtually hierarchized in multiple levels in order to construct tree structure. The data of each node are gathered to the cluster-head which is the representative node in each grid. The cluster-head aggregates the data of each level and maintains the summarized data in the network. When end users or applications cast a query to the network, required data is obtained more efficiently by this method compared with nonhierarchical approaches.