Super-resolution modeling of the indoor radio propagation channel

Results from parametric modeling of the indoor channel frequency response are presented. Minimum-norm and autoregressive (AR) are two super-resolution modeling techniques that are used on frequency domain data to form the theoretical sum-of-sinusoids model for the indoor channel-impulse response. Using these techniques, it is possible to check the validity of assuming Turin´s model for the indoor channel. It is shown that the modeling techniques estimate the indoor channel-impulse response with a time-domain resolution better than that of current measurement systems. A competing model that matches the measured channel frequency response to a sampled finite-impulse response (FIR) filter (as opposed to AR and minimum-norm, which give arbitrary time delays) is presented. It is shown that FIR filter matching produces a parametric model that can be more compact and more accurate in terms of matching the frequency response of the channel than the conventional super-resolution techniques