This paper analyzes the implementation of an algorithm into a FPGA embedded and distributed target location method using the Received Signal Strength Indicator (RSSI). The objective is to show a method in which an emb...This paper analyzes the implementation of an algorithm into a FPGA embedded and distributed target location method using the Received Signal Strength Indicator (RSSI). The objective is to show a method in which an embedded feedforward Artificial Neural Network (ANN) can estimate target location in a distributed fashion against anchor failure. We discuss the lack of FPGA implementation of equivalent methods and the benefits of using a robust platform. We introduce the description of the implementation and we explain the operation of the proposed method, followed by the calculated errors due to inherent Elliott function approximation and the discretization of decimal values used as free parameters in ANN. Furthermore, we show some target location estimation points in function of different numbers of anchor failures. Our contribution is to show that an FPGA embedded ANN implementation, with a few layers, can rapidly estimate target location in a distributed fashion and in presence of failures of anchor nodes considering accuracy, precision and execution time.展开更多
针对无线传感器网络中直接使用接收信号强度指引RSSI(Received Signal Strength Indicator)定位会出现因接收信号强度随机性变化而导致的测距粗糙、定位不稳定的普遍现象,结合实际项目定位精度要求,在实验的基础上提出一种提高RSSI定位...针对无线传感器网络中直接使用接收信号强度指引RSSI(Received Signal Strength Indicator)定位会出现因接收信号强度随机性变化而导致的测距粗糙、定位不稳定的普遍现象,结合实际项目定位精度要求,在实验的基础上提出一种提高RSSI定位精度的功率匹配算法PMA(Power Match Algorithm)。实验首先通过测定RSSI与距离的关系,建立测距模型,然后在此基础上建立四边定位静态数据库,最后进行现场测试和误差分析。实验结果表明,该算法能适应RSSI测距信号强度变化不稳定性特点,定位平均误差约为0.07 m。展开更多
文摘This paper analyzes the implementation of an algorithm into a FPGA embedded and distributed target location method using the Received Signal Strength Indicator (RSSI). The objective is to show a method in which an embedded feedforward Artificial Neural Network (ANN) can estimate target location in a distributed fashion against anchor failure. We discuss the lack of FPGA implementation of equivalent methods and the benefits of using a robust platform. We introduce the description of the implementation and we explain the operation of the proposed method, followed by the calculated errors due to inherent Elliott function approximation and the discretization of decimal values used as free parameters in ANN. Furthermore, we show some target location estimation points in function of different numbers of anchor failures. Our contribution is to show that an FPGA embedded ANN implementation, with a few layers, can rapidly estimate target location in a distributed fashion and in presence of failures of anchor nodes considering accuracy, precision and execution time.
文摘针对无线传感器网络中直接使用接收信号强度指引RSSI(Received Signal Strength Indicator)定位会出现因接收信号强度随机性变化而导致的测距粗糙、定位不稳定的普遍现象,结合实际项目定位精度要求,在实验的基础上提出一种提高RSSI定位精度的功率匹配算法PMA(Power Match Algorithm)。实验首先通过测定RSSI与距离的关系,建立测距模型,然后在此基础上建立四边定位静态数据库,最后进行现场测试和误差分析。实验结果表明,该算法能适应RSSI测距信号强度变化不稳定性特点,定位平均误差约为0.07 m。
