Energy-Based Timing Estimation and Artificial Neural Network Based Ranging Error Mitigation in mm-Wave Ranging Systems Using Statistics Fingerprint Analysis

摘要

Non-line-of-sight (NLOS) and dense multipath problems are the major challenges for the millimeter wave (mm-wave) indoor ranging systems. To acquire time of arrival (TOA) estimate accurately in such a poor environment, an improved statistics fingerprint analysis algorithm for energy-based timing estimation and artificial neural network (ANN) based ranging error mitigation is presented in this paper. The developed algorithm can obtain TOA estimate accurately by measuring the kurtosis, skewness, standard deviation, minimum slope, and gradient of the received mm-wave pulses. ANN is employed to mitigate the ranging error based on the obtained nonlinear regression between the thresholds and the analyzed characteristics of mm-wave pulses. The presented numerical simulation results indicate the proposed algorithm can achieve significant performance improvements in both line of sight and NLOS channels of the IEEE 802.15.3c standard, as compared to conventional algorithms.