Experimental Analysis of Wireless Sensor Networks Operating Systems Case Study of Tiny OS

Abstract

While this research will primarily focus on wireless sensor networks, some of the principles, such as those of collaborative information processing and management will be highlighted. Using embedded sensors with TinyOS to solution of different networking application will still be the main concern of this research. Sensor networks consist of (potentially) thousands of tiny, low-power nodes, each of which execute concurrent, reactive programs that must operate with severe memory and power constraints (many applications fit within 16KB of memory, and the core OS is 400 bytes). TinyOS meets these challenges well and has become the platform of choice for sensor network research, and supports a broad range of applications. We provide a descriptive analysis of the system, showing that it supports complex, concurrent programs with very low memory requirements and efficient, low-power operation. The thesis describes how to implement a Wireless Sensor Network using Mica2 mote loaded with TinyOS and use TinyDB to select data that are collecting by sensor on base station. Finally, a simple experiment is presented to compare several operating systems for wireless sensor networks. TinyOS is found to outperformed other operating systems in terms of optimized memory space and energy consumption.