Modeling City-Canyon Pedestrian Radio Channels Based on Passive Sounding in In-Service Networks.
Xuefeng Yin,Meng Tian,Luxia Ouyang,Xiang Cheng,Xuesong Cai,Li Tian,Jiajing Chen,Pinlu Yang
DOI: https://doi.org/10.1109/tvt.2015.2506180
IF: 6.8
2015-01-01
IEEE Transactions on Vehicular Technology
Abstract:Recently, a measurement campaign for characterizing time-variant radio channels in city-canyon environments was conducted along a 2-km-long pedestrian shopping street in Shanghai. A passive sounding approach was adopted where a software-defined-radio platform is used as a receiver (Rx) to record downlink signals transmitted by base stations (BSs) in an in-service Universal Mobile Telecommunications System (UMTS). Channel impulse responses are calculated from the data received in common pilot channels, and a space-alternating generalized expectation-maximization (SAGE) algorithmis applied to estimate delays, Doppler frequencies, and complex attenuations of multipath, which are further grouped into clusters characterized by time-evolving parameters. Based on the observations from a total of 70 BSs, stochastic models are established for composite channel parameters, cluster-level characteristics, and the time variabilities of these parameters. The novelties of the models lie in the passive data acquisition, which leads naturally to superior model ergodicity and higher fidelity than the standard models constructed based on active sounding. Furthermore, the models proposed here also exploit a new channel categorization method that takes into account the realistic impact of system configurations, such as the usage of multiple remote radio units in a single BS.