Neural-Network-Based NLOS Identification of Angular Clusters at 60 GHz

Abstract

The work in this paper identifies the nature of individual angular clusters as line-of-sight (LOS) or non-line-of-sight (NLOS) in indoor millimeter-wave channels. The proposed technique utilizes the channel knowledge that is readily available from a beam training process in directional antenna-based communications. In particular, the behavior of five different channel metrics, namely the angular covariance, the time-domain, and frequency-domain channel kurtosis, the mean excess delay, and the RMS delay spread, is analyzed using maximum likelihood ratio and artificial neural network. A noticeable difference between LOS and NLOS clusters is observed and assessed for identification. Hypothesis testing shows errors as low as 0.02-0.003 in simulations and 0.04-0.07 in measurements at 60 GHz in indoor shortrange environments.