Design And Implementation Of WSN Nodes Programming Language

Wireless sensor networks are networks of compact micro sensors with wireless communication capability. Wireless sensor network technology has been developed rapidly in the past decade. Not only for academic research, but also WSN has bean used in commercial applications. However, the technological progress has not been changed sensor nodes resource-constrained characteristics: progress makes sensor lower price but will not brought memory and energy capacity upgrade. Memory and energy constraints makes building cross platform node applications more difficult, especially the hardware constantly updated. While the increasingly WSN applications needed, how to provide good programming language and programming model to build safe, reliable and easy to maintain application is a urgent problem to resolve.Today, most used node programming language is the C language and nesC language. C programming language is a very popular general language, but does not support to build safe and reusable applications. nesC exposes a programming model that incorporates event-driven execution, and component-oriented application design. But nesC does not support instances with private states, and difficult to use with the ah hoc grammar. Since nodes extremely limited resources, modern object-oriented programming is not applicable due to the cost. After analysis of the main node programming language, programming model, and programming difficult for a node, this thesis proposes a specific new programming language CSpire. While try to provide encapsulate, reliable, easy to use and efficient language, we refer to the object-oriented method and the current high level language grammar. CSpire provides a efficient OO way and a event-subscribe mechanism. Compiler optimization and restrictive Object-oriented characteristics ensure efficiency. During the compiling, data race detect and memory access check code insert make application more reliable.The contributions of this thesis can be summarized as follows: firstly, analysis of the node programming difficulties, and provide a efficient OO programming language; secondly, data race detect improve the reliability of applications; finally, developed a compiler for CSpire, witch optimize the efficiency and generate the target code.