NLOS Identification and Machine Learning Methods for Predicting the Outcome of 60GHz Ranging System

Millimeter-wave (MMW) signals in 60GHz band have shown immense potential for accurate range estimation with precise time and multipath resolution. Nonline of sight (NLOS) propagation is a primary factor affecting the range precision. To improve range estimation, an Energy detector (ED) based normalized threshold algorithm which employs a Neural network (NN) is developed on the basis of NLOS identification. The maximum curl and standard deviation of the received energy block values are used to determine NLOS environment and the normalized thresholds for different Signal-to-noise ratios (SNRs). The effects of the channel and integration period are evaluated. Performance results are presented which show that the proposed approach provides better precision and is more robust than other solutions over a wide range of SNRs for the CM1.1 and CM2.1 channel models in the IEEE 802.15.3c standard.