Sensor nodes are scattered in a physically spacious area and accordingly powered by batteries instead of being tethered to durable power sources. Generally nodes are assumed to be revoked rather than replenished when they exhaust all the battery power. Previous empirical studies show that the larger portion of power is consumed by communication between nodes. Therefore, in order to expand overall system lifetime, it is crucial to design energy-efficient communication protocols for sensor networks.

Figure 1. Data dissemination services in sensor networks: a two-tiered example.

Currently in this work, we focus on investigating data dissemination in a two-tiered network which is comprised of stationary sensor nodes and mobile data users as shown in Figure 1. For example, in an emergency rescue, rescuers might need to monitor a specific area that they are supposed to search, while approaching that area. Desired data updates would be periodic to keep data fresh, and an area of interest might overlap with another. Such data dissemination applications suggest protocol design criteria like the following:

• Immediate deployment: The protocol should be designed not to require a long-term startup (e.g., network topology construction) after sensor node placement, to get ready for the actual sensor data dissemination.
• Adaptability: The protocol should be scalable to both the number of data sources and the data sink populations, and allow the diversity of user requests in terms of desired update rates and service durations.
• Fast response to data requests: It is desirable that users do not experience a substantial amount of delay after they request sensor data updates.
• Energy-efficiency: Given data update demands, the protocol should be able to satisfy them with lower energy dissipation and ultimately extend the network lifetime.

Collaborators
The members of this work have been collaborating with people at University of Virginia including:

• Sang H. Son
• John A. Stankovic
• Shuoqi Li