Blind carrier phase tracking with guaranteed global convergence

2006 IEEE 7th Workshop on Signal Processing Advances in Wireless Communications, 2006

In certain digital communication systems, such as those using Tomlinson-Harashima precoding and signal shaping, the received symbols are not equiprobable; rather, symbols with more energy are less likely. It is shown that in such scenarios the performance of standard carrier phase estimators is severely degraded. Using a discrete Gaussian model for the symbol probability distribution, we present the likelihood function for phase estimation, as well as an approximate Maximum Likelihood blind phase estimator. Simulation results show that the proposed estimate is much more robust to non-equiprobable symbol distributions than standard ones.

Journal of Lightwave Technology, 2000

The impact from phase drifts in the different branches of parallel coherent receivers is investigated and it is shown how the spectrum is broadened when the receiver branches are not phase stabilized. Based on this, we propose a blind algorithm for compensating these phase drifts in digital signal processing by minimization of the spectral width. The algorithm performance is then evaluated by numerical simulations of quadrature phase-shift keying data using return-to-zero modulation. It is found that the algorithm is capable of identifying the phases with sufficient accuracy to make the residual effect of the phase mismatches negligible compared to the signal distortion by noise at a bit error rate of 10 −3 .

IEEE Photonics Technology Letters, 2000

Data-aided carrier recovery based on Phase-Lock Loop (PLL) is a popular scheme for tracking the phase of an incoming carrier affected by phase noise. Optimum tracking is achieved by a Kalman filter, that, with multilevel Quadrature Amplitude Modulation (QAM), is implemented by a PLL with variable loop gain. The paper shows that a PLL with fixed gain achieves a performance that is close to the optimal one provided that the phase detector is optimized. Monte Carlo simulations for the mean-square phase error and the symbol error rate of a 256-QAM constellation are provided to validate the analysis.

We introduce a new blind criterion for carrier phase recovery of QAM modulated data, based on the maximization of the average of the q-th power of the phase-compensated received data vector 1 -norm. In order to improve the performance of state-of-the-art higher-order estimators for cross-QAM constellations, which are sensitive to finite precision effects, our focus is on low-order methods (q = 1 and 2). Fixedpoint iterations with good local convergence properties are given for these cases; different existing phase estimators can be used as initializers, depending on the operation environment and application constraints.

— This paper introduces a new receiver for joint blind equalization and carrier phase recovery. The new receiver overcomes several disadvantages available in literature in this area by accomplishing modulus equalization and phase recovery in two independent operations. The proposed structure enables alternative and independent approaches to be taken for the design of the modulus and phase equalizer respectively, and in the context we propose two new algorithms for the modulus equalizer and the phase shifter. In particular, we propose a new constant modulus (CM) and modulus decision directed (MDD) hybrid algorithm for the equalizer. The CM and MDD both influence the adaptation of equalizer weight taps simultaneously. This enhances both the convergence rate and the steady state performance. The selection of step sizes for the modulus equalizer and the phase shifter is discussed. We compare the performance of our receiver with other previous receivers via computer simulations.

IEEE Transactions on Signal Processing, 2003

We address the problem of nondata aided frequency offset estimation for noncircular transmissions over frequencyselective channels in a linear precoder-based communications context. Linear precoding is representative of a downlink direct-sequence code division multiple access (DS-CDMA) system or of an orthogonal frequency division multiplexing (OFDM) system. We observe that twice the frequency offset is a cyclic frequency of the received signal. We thus introduce an estimator relying on the maximization of the empirical cyclocorrelations. We analyze its asymptotic behavior and obtain a closed-form expression for the asymptotic covariance. This enables us to design relevant system parameters. Numerical illustrations are provided and confirm our theoretical assertions. Comparisons with existing methods are also reported.

ISI (Intersymbol Interference) severely degrades the performance of high speed digital communication systems. A popular technique for countering ISI is Adaptive Equalization. Blind/Unsupervized equalization is one class of adaptive equalization where initial training of equalizer's weights is not feasible. Godard's Constant Modulus Algorithm (CMA) was the very first blind equalization algorithm that could be applied for two dimensional communication systems. It requires the transmitted signal's statistics to perform equalization. Picchi & Prati modified the standard decision-directed algorithm and proposed a new algorithm by the name Stop-and-Go. Wei Rao by simply converting the equalizer's output proposed a modified version of the CMA. This paper compares the performance of blind equalization algorithms: CMA, Stop-and-Go decisiondirected algorithm and Wei Rao's modified constant modulus algorithm for 16 and 64-QAM constellations transmitted across a linear band-limited channel. Simulation results indicate impulse response of the cascade of transmission channel and equalizer (after convergence) is similar to that of ideal transmission channel's impulse response. The results also show that stop-and-go outperforms the rest in terms of convergence rate and MSE.

Optics Express

 Users may download and print one copy of any publication from the public portal for the purpose of private study or research.  You may not further distribute the material or use it for any profit-making activity or commercial gain  You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

IEEE Transactions on Signal Processing, 2008

This paper introduces a novel, not data aided, phaseoffset estimator for quadrature amplitude modulated (QAM) signals. Contrarily to near-efficient existing phase acquisition techniques, this estimator does not require a preliminary gain adjustment stage while its accuracy preserves the slope of Cramér-Rao bound for medium-high signal-to-noise ratio (SNR) ranges, where it typically outperforms existing blind estimators, with significant improvement for dense and cross QAM constellations. Moreover, it needs only a very rough estimate of the SNR. Like other gain-control-free blind phase-offset estimators, it measures the amount of the cyclic shift by which the (four-folded) phase probability density function (pdf) is rotated under an unknown phase-offset. Estimation of the phase-offset-induced cyclic shift is conducted first by measuring the received data phase pdf by a canonical phase histogram procedure, then by estimating the phase-offset-induced cyclic shift through a cyclic cross correlation-based procedure between the measured phase histogram and a reference phase pdf evaluated within the zero phase-offset hypothesis. Actually, the estimation procedure is presented in a generalized version that considers a tomographic projection of the bidimensional (magnitude/ phase) pdf of suitable nonlinear transformations of the received data. The tomographic projection performs a magnitude weighing on the pdf, and this, in turn, results in an improved overall estimation accuracy, as shown by theoretical analysis and numerical simulations here performed to assess the estimator performance.

IEEE/SP 13th Workshop on Statistical Signal Processing, 2005, 2005

Two new iterative methods for blind phase estimation of QAM signals are presented. The algorithms seek the maxima of certain cost functions derived from the dispersion of the de-rotated data with respect to a diamond-shaped contour, and their distinctive feature is the inclusion of a sign nonlinearity at every iteration. A connection between one of these methods and the standard fourth-power method is presented. Simulations with cross-QAM constellations show that, when properly initialized, the new schemes present lower variance than the fourth-power method, and may even outperform Cartwright's eighth-order method.

2008 5th International Multi-Conference on Systems, Signals and Devices, 2008

Multi-path effect in HF communication not only changes the amplitude, but changes the phase of receiving signal as well. The rapid and random movement of ionospheric causes the oscillation of the phase. It could be viewed as the Doppler shift of highfrequency carrier caused by ionospheric. The shift of frequency should be corrected because it brings difficulty in recovering the carrier. The non-blind method needs much quantities of training (or pilot frequency) sequences at the cost of system recourse (bandwidth). However, the blind estimation will wildly encounter the vague phase. In this paper, a semi-blind parameter estimation method is introduced which uses few training sequences to distinguish the carrier. In this way, the carrier's phase and timing parameter could be recovered rapidly. Besides, this algorithm is easy to implement in hardware platform.

IEEE Signal Processing Letters, 1996

Abstruct-In this letter we present a method for phase recovery in QAM communication systems based on higher order statistics. Under the assumption that the QAM signals have independent, identically distributed (i.i.d.) in-phase and quadrature components, we derive the relation between the phase error and the fourth-order cumulant of the output. Numerical performance of the approach is demonstrated by computer simulation.

Conference Record of the Thirty-Third Asilomar Conference on Signals, Systems, and Computers (Cat. No.CH37020)

This paper considers the problem of blind (non-data aided) symbol clock and frequency offset est,imation in the context of time-selective fading communication channels. The performance of estimat,ors based on t.he cyclic correlation is analyzed and optimality issues are addressed. Closed-form expressions for the Cram&-Rao bound are developed under the Gaussian assumption. Extensions of the algorithm to staggered modulation (e.g., OQPSK) are studied. In this case, synchronization generally requires fourth-order cyclic statist,ics

Asia Communications and Photonics Conference, 2012

CITATIONS 0 READS 7 5 authors, including: Some of the authors of this publication are also working on these related projects: EU ICONE View project

1993

In this paper, a blind carrier recovery technique aided by modulation identification is presented. This technique can also be interpreted as a modulation identification method that is robust against phase and frequency offsets. The performance of the proposed technique is studied with respect to the maximum frequency offset and observation window size.