Location-Based Addressing

Agilla addresses each node by its location rather than its TinyOS ID because the spatial properties of the network are vital to sensor network applications, particularly those written using multiple mobile agents. To operate effectively, mobile agents need to know their location to coordinate with each other, and to correlate sensor data with the location from which they were obtained. Having a sensor reading without its location is not useful. For example, if an agent takes a temperature reading and determines that there is a fire, it cannot simply tell a base station "fire" but rather "fire at (1,2)." By utilizing a location-centric addressing scheme, Agilla provides a unified way of identifying nodes, and allows its primitives to be generalized to operate over regions (e.g., out a tuple into all tuplespaces within a geographic region).


This work is supported by the ONR MURI Project CONTESSA and the NSF under grant number CCR-9970939.