Azzédine Touzni

Fractionally spaced equalization using CMA: robustness to channel noise and lack of disparity

In the noise-free case, the fractionally spaced equalization using constant modulus (FSE-CM) criterion has been studied previously. Its minima were shown to achieve perfect equalization when zero-forcing (ZF) conditions are satisfied and to be able to still achieve fair equalization when there is lack of disparity. However, to our best knowledge, the effect of additive channel noise on the FSE-CM cost-function minima has not been studied. In this paper, we show that the noisy FSE-CM cost function is subject to a smoothing effect with respect to the noise-free cost function, the result of which is a tradeoff between achieving zero forcing and noise enhancement. Furthermore, we give an analytical closed-form expression for the loss of performance due to the noise in terms of input-output mean square error (MSE). Under the ZF conditions, the FSE-CM MSE is shown to be mostly due to output noise enhancement and not to residual intersymbol interference (ISI). When there is lack of disparity, an irreducible amount of ISI appears independently of the algorithm. It is the lower equalizability bound for given channel conditions and equalizer length-the so-called minimum MSE (MMSE). The MMSE lower bound is the sum of the MMSE and of additional MSE mostly due to noise enhancement. Finally, we compare the FSE-CM MSE to this lower bound.

Enhanced 8-VSB transmission for North-American HDTV terrestrial broadcast

We present a new robust backward compatible 8-VSB transmission system for HDTV terrestrial broadcast. The new transmission system allows better reception of signals for portable, mobile and indoor receivers. In the proposed system a scalable fraction of the main service is used to transmit robustly encoded data. The robust service has at least a 6 dB fade margin advantage over the main service. In addition, the data allocated to the robust stream can also be used at the receiver to improve the reception of the normal stream. The robust data is utilized by the normal stream receiver as randomly distributed training symbols which can be used in the equalizer to improve the receiver performance for dynamic multipath channels.

Differential carrier loop estimation for digital vestigial side band modulation

In this paper we address the problem of phase recovery for a digital VSB modulation. Two criteria, a Constant Modulus and MSE, which exploit the properties of VSB signaling in the presence of noise are proposed and analyzed. The extrema of the objective functions and their domains of attraction are derived. Under mild conditions on the statistics of the input sequence, it is shown that the derivative of these functions can be used as a new phase detector in a Phase Locked Loop (PLL) implementation. An effective semi-blind solution which consists of a combination of the MSE and CM phase detectors is introduced and analyzed.

Closed form expression of EMSE for Bussgang equalization with spatio-temporal diversity

Bussgang algorithms are a class of simple stochastic gradient type solutions for blind channel equalization. We investigate the degradation of performance in the source estimation resulting from the stochastic jitter around stationary points of Bussgang algorithms (which correspond to the equalizers of the channel). More precisely, we derive a closed form approximation of the excess mean square error (EMSE) defined as the variance of the jitter which depends on the effects of the channel characteristics, the source distribution and the non-linearity used in the update equation. The analysis is performed in a context of spatiotemporal channel diversity involving a single-input/multiple outputs data model.