Ansgar Scherb

Performance Analysis of Maximum-Likelihood Semiblind Estimation of MIMO Channels

Iterative channel estimation and data detection is a useful method to improve the channel estimation quality without sacrificing the bandwidth efficiency. Since both the known training symbols (non-blind) and the unknown data symbols (blind) are used for channel estimation, corresponding techniques are referred to as semiblind. If the channel estimator and data detector are both optimal in the sense of maximum-likelihood criterion, we may call the algorithm as maximum-likelihood (ML) semiblind channel estimation (SBCE). This paper deals with ML-SBCE for frequency-flat multi-input multi-output systems with focus on the channel estimation mean squared error (MSE) analysis. Through semi-analytical efforts, we showed that ML-SBCE is biased at low SNR and tends to be unbiased at high SNR. The reasons of biasing are the erroneous data detection and the correlation between the noise and the detection errors. Besides, we showed that the MSE performance of ML-SBCE is also influenced by the noise-error correlation. Based on these analyses, possibilities to compensate the biasing as well as improve the MSE performance is pointed out

On phase correct blind deconvolution exploiting channel coding

In this paper we derive an algorithm, which estimates the channel blindly exploiting the statistical dependencies of the transmitted signal caused by channel coding. An additional feature of this algorithm is that in contrast to most blind deconvolution algorithms phase correct estimates can be obtained. The error performance of the proposed algorithm depends on the characteristics of the channel code. If the code has appropriate properties, which is true for some convolutional codes as well as for several block codes, e.g. especially low-density parity check codes (LDPC), the proposed algorithm performs similarly or slightly better in comparison to higher order statics based algorithms.

Blind identification and equalization of LDPC-encoded MIMO systems

We propose a new algorithm which blindly identifies and equalizes a MIMO system, where all sources are independently protected against errors by an LDPC code. The proposed method exploits statistical dependencies caused by the channel code. In contrast to most common blind source separation algorithms, the new method does not suffer from a permutation ambiguity. Furthermore, if the channel code is asymmetric, the suggested method delivers phase correct estimates of the channel and the corresponding equalizer. The performance of the presented method is evaluated by numerical results.

Cramer-Rao lower bound for semiblind channel estimation with respect to coded and uncoded finite-alphabet signals

Within this paper we derive the Cramer-Rao lower bound (CRLB) for semiblind channel estimation with respect to coded or uncoded finite alphabet source signals in finite impulse response systems. Since the obtained solution incorporates a high dimensional integral, which can only be solved numerically, we approximate the CRLB for low and high signal to noise ratio (SNR). We also examine the SNR range where the CRLB crosses over from the low to the high SNR approximation. It is shown that the crossover range depends on the modulation index and code characteristics. Combining the approximations for high and low SNR and estimating the crossover range yields an overall approximation, which can easily be calculated also for more complex scenarios. It is shown that the approximation meets the true CRLB very well.

On phase correct blind deconvolution of flat MIMO channels exploiting channel encoding

In this paper, we propose a new algorithm separating blindly the sources of a flat MIMO-communication link, where all sources are independently channel encoded. To this end the proposed method exploits statistical dependencies caused by the channel code for blindly estimating the channel and simultaneously adjusting a linear equalizer. In contrast to most common blind source separation methods the suggested method completely resolves the phase and permutation ambiguity of the estimated channel impulse response and the equalizer. The performance of the presented method is evaluated by numerical results.

Pilot aided channel estimation for short-code DS-CDMA

This paper investigates pilot based channel estimation techniques for short code DS-CDMA in a multipath environment. It is shown that the conventional correlative estimation algorithm is biased and very sensitive to multiuser interference Therefore, we suggest two advanced estimation algorithms. The first one eliminates the inherent path crosstalk and yields an unbiased estimate while still suffering from MUI. The second approach suppresses the influence of the MUI by incorporating a linear interference suppression into the channel estimation algorithm. It is shown that significant performance improvements can be achieved with the proposed algorithms especially under extreme near-far conditions.

Non-Coherent LDPC Decoding on Graphs

Within this paper graph based non-coherent decoding algorithms for LDPC encoded systems are proposed. We study a class of LDPC codes suitable for non-coherent detection. On the basis of the related graphs a non-coherent decoding algorithm with variable trade-off between computational complexity and BER performance is derived. The proposed scheme is also capable to deal with pilot symbols if available. Finally, the excellent performance of the proposed methods is verified by simulations

Iterative near---far resistant channel estimation by using a linear minimum mean squared error detector

Channel estimation techniques for CDMA system need to combat multiple access interference (MAI) to improve the estimation performance. The linear MMSE detector has certain advantages with respect to the near–far problem and can be used to develop a channel estimation algorithm. In this paper, an efficient iterative method for near–far resistant single-user mobile radio channel estimation in slow fading multi-path direct sequence code division multiple access (DS-CDMA) channels is presented. Computer simulation results demonstrate that a significant performance improvement can be achieved with the proposed method especially under extreme near–far conditions.

Blind Equalization of Frequency Selective MIMO Systems via Statistical and Trellis-based Methods

In this paper we present a composite algorithm for blind sequence estimation in frequency selective multiple-input multiple-output systems. The proposed algorithm combines methods based on second and higher order statistics with a trellis-based approach. Moreover, the computational complexity of the presented method can be reduced by pre-processing the observed data with a blindly designed linear impulse shortening filter. As verified by simulations low computational effort has not necessarily led to a deterioration of performance.

Comparison of methods for iterative joint data detection and channel estimation

This paper compares iterative deterministic and Markov chain Monte Carlo algorithms approximating the maximum likelihood of joint data detection and channel estimation with respect to the quality of an initial channel estimate. The quality of the initial channel estimate is measured by the normalized mean squared error between estimated and true channel. The deterministic method does not take the instantaneous quality of the channel estimation or of the current data estimate into account and might get trapped in a local maximum of the likelihood function, whereas the Monte Carlo methods theoretically almost converge to the joint maximum likelihood. Based on simulation results, it is shown that a performance gain can be achieved by applying the second class of algorithms at the expense of slower convergence speed.