Hierarchical data dissemination scheme for large scale sensor networks

We introduce three hierarchical data dissemination schemes for reducing energy consumption in sensor networks that dynamically adapt to demands for sensor data. Our schemes are developed in the framework of a general model for wireless sensor networks in which some or all of sink nodes may be mobile. The schemes route data towards sink nodes using a hierarchy of randomly selected dissemination nodes. As dissemination nodes have limited resources, whenever a dissemination node is over-loaded, it inserts another level of dissemination nodes to reduce its load. This is accomplished by dividing the dissemination node's geographical area of coverage into smaller quads by recruiting other sensors. This new level of nodes take over serving the sinks located close to them and thus reduce the load of the dissemination node. The difference between the three schemes is that, the first scheme uses separate hierarchies for each source generating information. The second scheme first aggregates data by creating a level of data aggregating sensor nodes and then uses a common hierarchy to disseminate data to the sinks. The third one combines the ideas implemented in both the first and second schemes and achieves data dissemination. We develop algorithms for dynamic creation of the hierarchy, and maintenance and routing of data to sinks. We show that our schemes scale well and yield 85-95% throughput for large sensor networks with less power consumption compared to existing models.