针对井下无线信号传播环境发生改变需要重新采集RSS(received signal strength)指纹数据的问题,提出一种基于支持向量回归(SVR)的定位误差修正算法。该算法在离线阶段利用支持向量回归机,建立待训练点的RSS值和定位结果与定位误差之间...针对井下无线信号传播环境发生改变需要重新采集RSS(received signal strength)指纹数据的问题,提出一种基于支持向量回归(SVR)的定位误差修正算法。该算法在离线阶段利用支持向量回归机,建立待训练点的RSS值和定位结果与定位误差之间的非线性关系,在线阶段利用该模型计算RSS样本的定位误差,并修正定位结果。实验结果表明,在离线训练点数量增加前后,该修正算法比指纹匹配算法的定位误差分别减少了22%与38%。展开更多
This work extends the use of wavelet-based denoising as an alternative processing scheme to improve measured mobile-radio channel power delay profiles. It has already been reported that, when applied on real domain da...This work extends the use of wavelet-based denoising as an alternative processing scheme to improve measured mobile-radio channel power delay profiles. It has already been reported that, when applied on real domain data (amplitude only), denoising provides mainly a qualitative improvement. Here, phase content was also considered, leading to significant qualitative and quantitative improvement of the processed profiles. Signal-to-noise ratios and dynamic ranges improvements as high as 50 dB have been observed.展开更多
文摘针对井下无线信号传播环境发生改变需要重新采集RSS(received signal strength)指纹数据的问题,提出一种基于支持向量回归(SVR)的定位误差修正算法。该算法在离线阶段利用支持向量回归机,建立待训练点的RSS值和定位结果与定位误差之间的非线性关系,在线阶段利用该模型计算RSS样本的定位误差,并修正定位结果。实验结果表明,在离线训练点数量增加前后,该修正算法比指纹匹配算法的定位误差分别减少了22%与38%。
文摘This work extends the use of wavelet-based denoising as an alternative processing scheme to improve measured mobile-radio channel power delay profiles. It has already been reported that, when applied on real domain data (amplitude only), denoising provides mainly a qualitative improvement. Here, phase content was also considered, leading to significant qualitative and quantitative improvement of the processed profiles. Signal-to-noise ratios and dynamic ranges improvements as high as 50 dB have been observed.
