An Optic-fibre Sensor Network System Based On Embedded System

With the computer and network technology development, web-based embedded systems technology has an increasing number of applications to various areas. In this dissertation, based on embedded systems and network technology, we implement a fiber-optic sensor network system based on embedded system. It has improved the traditional sensors shortcomings of single-point detection, single function, formed a multi-point detection, multi-functional, remote control, intelligent optical sensor new work system.The Distributed Optical Fiber Sensor (B-DOFS) system in this dissertation is based on BOTDR. Data acquisition system is included by FPGA and embedded system platform in. We implement acquisition, accumulation, cache, and other operations in FPGA. S3C2410 CPU is used for the embedded platform. An embedded Linux system is running on the embedded platform, and we implement long-distance data transmission and acquisition parameters control in the Linux system. An Embedded Linux driver is implemented to visit the FPGA. And we design a control process to control the server's running and a network process to send data and receive the client's control parameters. Shared memory is adopted to transfer data between control process and network process. We design a distributed data acquisition system, including the some servers (that is the network process in the embedded platform), a repository and some client. We use TCP/IP protocol to achieve the transmission of network. Linux-based multi-threaded technology is used to implement the network service, which can improve the efficiency of the system. We use thread" synchronization to implement a read/write system for the address list of server in repository. In clients, every connection to a server has a control thread, a data transmission thread and unattached collection buffer. Clients adopt thread-slot technology, not only ensure the security parameter transfer between main thread and sub thread, but also make the main thread easier to take scheduling of sub threads.