Performance evaluation of interference cancellation using a generalized-array-manifold model

In this paper, a novel interference cancellation (IC) algorithm is introduced which is applicable to improving the signal to interference power ratio (SIR) for the signals of a specific user by estimating the channels accurately. In this algorithm, a so-called generalized array manifold (GAM) generic channel model is adopted for parameter estimation which makes use of the Firstorder Taylor-series expansion to approximate the contribution of individual dispersive components in the channel to the received signal. Since the GAM model takes into account the influence of unresolvable distributed components, the IC algorithm is superior in terms of higher SIRs to conventional methods that reconstruct the signals of a desired user for dominant specular-path components. Monte-Carlo simulation results show that the SIR improvements vary with the delay and Doppler frequency spreads of a propagation channel. Besides, experimental results obtained by processing the measurement data collected in an in-service Universal Mobile Telecommunication System (UMTS) network demonstrate that by using the proposed IC algorithm and under the assumption that a channel consists of single dispersive component, which is usually the case where rich scattering occurs in the vicinity of a user, the SIR can be improved significantly when the delay and Doppler frequency spreads of channel are within certain ranges.