一种改进BP神经网络的接收信号强度测距算法

A ranging algorithm of received signal strength based on improved BP neural network

针对传统路径损耗模型测距过多依赖于环境参数A和n的问题,该文在分析BP神经网络模型的基础上,引进了基于蚁群算法优化BP神经网络模型(ACO-BP)的信号衰减模型。利用蚁群算法寻找最优的初始阈值和权值,并将其赋予BP神经网络;将信号强度作为输入值,距离作为输出值对ACO-BP网络进行训练;利用Matlab进行模拟仿真实验。实验结果表明:ACO-BP神经网络比BP神经网络预测距离值的精度平均提高了75%,该算法可应用于无线网络室内定位技术中。

Aiming at the problem that the traditional path loss model relied too much on environmental parameters A and n,based on the research and analysis of BP neural network model,a signal attenuation model based on the ant colony optimization BP neural network model(ACO-BP)had been proposed.Ant colony algorithm was used to find the optimal initial threshold and weight,and they were given to BP neural network.The signal strength was used as the input value and the distance as the output value to train the ACO-BP network.The simulation experiment was carried out by using MATLAB.The accuracy of distance value predicted by ACO-BP neural network was 75%higher than that predicted by BP neural network.This algorithm was used in indoor location technology of wireless network.