Font Size: a A A

The Design And Application Of Wireless Sensor Network Nodes

Posted on:2012-09-27Degree:MasterType:Thesis
Country:ChinaCandidate:M Y ZhuFull Text:PDF
GTID:2178330332986096Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
The research purpose of the paper is to use infrared and vibration sensors to collect the signal from people when they go through the sensors. Use vibration and magnetic sensors to collect the signal from vehicles. Sensor nodes do the signal processing and data fusion in order to recognize whether it is from people or vehicles. Then sensor nodes transmit the results to the central node by wireless network. Finally, the central node reports it to the computer. The significance of the paper is to do the pre-study for wireless sensor network applied in traffic monitoring in the future.The general design mothod of wireless sensor network is as follow:It is made up of many nodes in the form of mobile and self-organizing network. It includes two kinds of nodes, many sensor nodes and one central node. There are three kinds of sensors:Infrared, vibration and magnetic sensors. It is a static network. Sensor nodes transmit data to a central node by the way of single-hop network. Central node communicates with sensor nodes in the way of combining polling with CSMA-CA(Carrier Sense Multiple Access with Collision Avoidance).To simplify the design, all the nodes have the same hardware architecture. Whether nodes are connected by the sensor or not and the software running on the node can be used to distinguish between the central node and sensor nodes. There are no data acquisition and recognition algorithms in central node software but in sensor nodes software. The functions of control, management and uploading data to PC are added to central node. Low-power, integrated and modular designs are used in hardware and software of nodes to ensure the system reliability and scalability.Hardware design process is divided into five steps:architecture design, component selection, schematic design, PCB layout and PCB test. Nodes are based on the system architecture of modular, including wireless transceiver module, serial port communication mcdule, sensor module, data collection and processing module, and power supply module.Software design is divided into three parts:MCU program, serial port debugging software, MATLAB program (Both serial port debugging software and MATLAB program are run on the computer). MCU program is the carrier of network protocol and recognition algorithm. It is subdivided into:main program, drivers and initialization part, network part, data collection part and identification algorithms part. Serial debugging software is downloaded from Internet. MATLAB program receives data through the serial port, which are uploaded from nodes. and then display and save the data. According to the saved data, do the recognition algorithm programming. Do the initial recognition algorithm verification and spectrum analysis in MATLAB, and it is easy for the program to migrate to MCU.Test the hardware and software after design. Hardware test is to guarantee its quality. Observe the Welding of components. Through the test program, initialize the hardware to detect its quality. Software test is to guarantee its functions. Detect the establishment of network, network maintenance, data transmission and the accuracy of recognition results. Finally, alignment test is the application test in practical environment.Generally, the communication of nodes is good and the accuracy of recognition algorithm is over 80%. Basically it meets the needs of application requirements. In the paper the innovation points as follows:The communication mode combines polling with CSMA-CA. Besides, just after node has started, the anti-interference algorithm automatically detects environmental noises, and modifies the threshold. The recognition algorithm includes the method of extreme value point detection and frequency calculation without using time.
Keywords/Search Tags:wireless sensor network, recognition algorithm, spectrum analysis, test
PDF Full Text Request
Related items