Applications development on a rule-based WSN middleware

Wireless Sensor Network (WSN) middleware eases the WSN application development by providing an application programming interface (API). Rule-based WSN middleware enables the applications and users to program the behavior of the sensor nodes. REED (Rule Execution and Event Distribution) is such a middleware solution that allows sensor networks to be programmed at run time. In this paper, we propose a method of developing WSN applications that uses finite state machine (FSM) as a bridge between application logics and the rules running on the REED, and demonstrate that for applications, if their behaviors can be described using finite-state machine (FSM), they can be directly described using the rules and thus implemented on the REED; further, we argue that rule-based middleware is useful for implementing bio-inspired mechanisms, such as selforganization, on WSN systems. Two WSN applications are implemented, as examples, on the REED: one is the desynchronization of sensor nodes, and the other is the clustering-based self-organization. This paper is not aimed to study a specific application or control mechanism on WSNs, but to, via two prototype implementations, show that rulebased middleware such as the REED is useful and flexible enough to support the development of WSN applications, especially for bio-inspired mechanisms. Keywords-wireless sensor network; rules; finite state machine