Advanced Markov Chain Monte Carlo Methods for Iterative (Turbo) Multiuser Detection
نویسندگان
چکیده
Recently, Markov Chain Monte Carlo (MCMC) sampling methods have evolved as new promising solutions to both multiuser and multiple-input multiple-output (MIMO) detection problems. Approaches based on Gibbs sampling as a special type of MCMC methods are well suited due to their good trade-off between performance and complexity. However, it is known that detection methods based on Gibbs sampling show a performance degradation in the high signal-to-noise ratio (SNR) regime. We propose an improved version of a soft-input soft-output algorithm, where this degradation effect is considerably mitigated. Employing the algorithm for turbo multiuser detection in overloaded code-division multiple-access (CDMA) systems yields excellent performance in comparison to other known detection schemes while offering moderate computational complexity.
منابع مشابه
Turbo Multiuser Detection with Markov Chain Monte Carlo Methods Employing Forced State Transitions
Recently, Markov Chain Monte Carlo (MCMC) sampling methods have evolved as promising solutions to both multiuser and multiple-input multiple-output (MIMO) detection problems. Approaches based on Gibbs sampling as a special type of MCMC methods are well suited due to their good trade-off between performance and complexity. However, it is known that detection methods based on Gibbs sampling may s...
متن کاملBlind turbo multiuser detection for long-code uplink CDMA with unknown interferences
We consider the problem of demodulating the multiuser symbols in an uplink long-code CDMA system in the presence of unknown out-cell multiple-access interference and narrowband interference. A novel blind Bayesian multiuser detector is derived based on the Bayesian inference of all unknown quantities. The Gibbs sampler, a Markov chain Monte Carlo (MCMC) procedure, is then used for Bayesian comp...
متن کاملAdaptive Bayesian multiuser detection for synchronous CDMA with Gaussian and impulsive noise
We consider the problem of simultaneous parameter estimation and data restoration in a synchronous CDMA system in the presence of either additive Gaussian or additive impulsive white noise with unknown parameters. The impulsive noise is modeled by a two-term Gaussian mixture distribution. Bayesian inference of all unknown quantities is made from the superimposed and noisy received signals. The ...
متن کاملMonte Carlo Bayesian Signal Processing for Wireless Communications
Many statistical signal processing problems found in wireless communications involves making inference about the transmitted information data based on the received signals in the presence of various unknown channel distortions. The optimal solutions to these problems are often too computationally complex to implement by conventional signal processing methods. The recently emerged Bayesian Monte...
متن کاملTurbo equalization for GMSK signaling over multipath channels
A novel receiver is derived for Gaussianminimum shift keying (GMSK) signals through a multipath channel. A nonlinear signal model is derived which avoids the linear approximation in the conventional finite impulse response(FIR) system model. A Bayesian equalizer based on the Gibbs sampler, a Markov chain Monte Carlo (MCMC) procedure, is developed for joint channel estimation and symbol detectio...
متن کامل