Research On And Implementation Of Compiling Optimization Tools In Wireless Sensor Networks

Wireless sensor networks have a significant potential in diverse applications but raise lots of difficulties for application development. The difficulty in programming sensor networks is not only due to its extremely constrained resources, such as low computation ability and small memory size, but also the high reliable requirement for its long running time and running environment. Comparing with application development for normal platform, the developers should care more about the running time or the RAM and ROM usage of applications for sensor nodes, and meet kinds of difficulties.The research work of this thesis is concentrated on the compiling tools of sensor node application development. The contributions can be summarized as follows: firstly, design the new programming language CSpire and its building system after analysis current sensor node programming languages and their build systems. The CSpire application is also optimized for its resource allocation. Secondly, based on some static stack analysis method of excitable file, point out a common analysis process and extend the method to different platforms and systems and improve its accuracy. And finally, integrate the build system and compiling optimization tools to SenSpire IDE.