Only 14 pages are availabe for public view
The use of wireless sensor networks to protect sensitive facilities or international borders has recently attracted more and more attention. It has become a high-priority issue in many countries. In addition to the physical fences built for stopping illegal intruders from crossing the border, smart fencing has been proposed to extend intrusion detection capabilities. Event detection is a central component in numerous wireless sensor network (WSN) applications. In spite of this, the area of event description has not received enough attention. The majority of current event description approaches rely on using precise values to specify event thresholds. However this crisp values cannot adequately handle the imprecise sensor readings.
The Fuzzy Logic System (FLS), is known to be robust and has excellent immunity to external disturbances, can tolerate the unreliable and imprecise sensor readings, and much more intuitive and easier to use. In addition to Fuzzy Logic System, a hybrid event detection algorithm is also introduced and added to the proposed system to enhance the capability of noise reduction in the sensor output signals. Therefore, in this Thesis, Event-detection algorithm based on two layers, 12 bits resolution, fuzzy Logic system (FLS) is modeled using Matlab Simulink, and then designed with the same aspects using VHDL as a proposed design for each node in the wireless sensor Network, which significantly improves the accuracy of the event detection. Each sensor node has an acoustic signal sensor and 3-axis acceleration sensor to improve the precision of the detection system, as well as reducing false alarm rate specially in a noisy environment. Then, the simulation results from both models using Matlab and the VHDL design, have been compared and provided that the FPGA-based fuzzy controller is very close to the software-based controller using MATLAB. Finally, the proposed system has been tested and verified to show the detection ratio of the entire system, which is about 81.2%, the suggestions of improving this testing result is also illustrated in this thesis.