Ahmed Masmoudi

A Maximum-Likelihood Channel Estimator for Self-Interference Cancelation in Full-Duplex Systems

Operation of full-duplex systems requires efficient mitigation of the self-interference signal caused by the simultaneous transmission/reception. In this paper, we propose a maximum-likelihood (ML) approach to jointly estimate the self-interference and intended channels by exploiting its own known transmitted symbols and both the known pilot and unknown data symbols from the other intended transceiver. The ML solution is obtained by maximizing the ML function under the assumption of Gaussian received symbols. A closed-form solution is first derived, and subsequently, an iterative procedure is developed to further improve the estimation performance at moderate-to-high signal-to-noise ratios (SNRs). We establish the initial condition to guarantee the convergence of the iterative algorithm to the ML solution. In the presence of considerable phase noise from the oscillators, a phase noise estimation method is proposed and combined with the ML channel estimator to mitigate the effects of the phase noise. Illustrative results show that the proposed methods offer good cancelation performance close to the Cramer-Rao bound (CRB).

A Non-Data-Aided Maximum Likelihood Time Delay Estimator Using Importance Sampling

In this paper, we present a new time delay maximum likelihood estimator based on importance sampling (IS). We show that a grid search and lack of convergence from which most iterative estimators suffer can be avoided. It is assumed that the transmitted data are completely unknown at the receiver. Moreover the carrier phase is considered as an unknown nuisance parameter. The time delay remains constant over the observation interval and the received signal is corrupted by additive white Gaussian noise (AWGN). We use importance sampling to find the global maximum of the compressed likelihood function. Based on a global optimization procedure, the main idea of the new estimator is to generate realizations of a random variable using an importance function, which approximates the actual compressed likelihood function. We will see that the algorithm parameters affect the estimation performance and that with an appropriate parameter choice, even over a small observation interval, the time delay can be accurately estimated at far lower computational cost than with classical iterative methods.

Closed-Form Expressions for the Exact Cramér–Rao Bounds of Timing Recovery Estimators From BPSK, MSK and Square-QAM Transmissions

In this paper, we derive for the first time analytical expressions for the exact Cramer-Rao lower bounds (CRLB) for symbol timing recovery of binary phase shift keying (BPSK), minimum shift keying (MSK), and square QAM-modulated signals. It is assumed that the transmitted data are completely unknown at the receiver and that the shaping pulse verifies the first Nyquist criterion. Moreover the carrier phase and frequency are considered as unknown nuisance parameters. The time delay remains constant over the observation interval and the received signal is corrupted by additive white Gaussian noise (AWGN). Our new expressions prove that the achievable performance holds irrespective of the true time delay value. Moreover, they corroborate previous attempts to empirically compute the considered bounds thereby enabling their immediate evaluation.

Channel Estimation and Self-Interference Cancelation in Full-Duplex Communication Systems

This paper presents a two-stage self-interference (SI) cancelation for full-duplex multi-input–multi-output (MIMO) communications systems. By exploiting the SI channel sparsity, a compressed-sensing-based SI channel estimation technique is developed and used in the first SI-cancelation radio-frequency (RF) stage to reduce the SI power prior to the receiver low-noise amplifier (LNA) and the analog-to-digital converter (ADC) to avoid overloading. Subsequently, a subspace-based algorithm is proposed to jointly estimate the coefficients of both the residual SI and intended channels and transceiver impairments for the second SI-cancelation baseband stage to further reduce the residual SI. Unlike other previous works, the intended signal is taken into consideration during the estimation process to reduce the overhead. It is demonstrated that the SI channel coefficients can be perfectly estimated with no knowledge of the intended signal, and only a few training symbols are needed for ambiguity removal in intended-channel estimation. Simulation results show that the proposed algorithms outperform the least square (LS) algorithms and offer the remarkable signal-to-residual-SI-and-noise ratio (SINR) approaching the signal-to-noise ratio (SNR).

A Maximum Likelihood Time Delay Estimator in a Multipath Environment Using Importance Sampling

In this paper, we present a new implementation of the maximum likelihood criterion for the estimation of the time delays in a multipath environment and then extend it to the estimation of the time difference of arrival when the transmitted signal is unknown. The new technique implements the concept of importance sampling (IS) to find the global maximum of the compressed likelihood function in a modest computational manner. It thereby avoids traditional complex multidimensional grid search or initialization-dependent iterative methods. Indeed, one of the most interesting features is that it transforms the multi-dimensional search inherent to multipath propagation into a much simpler one-dimensional optimization problem in the delays dimension. Moreover, it guarantees convergence to the global maximum, contrarily to the popular iterative implementation of the maximum likelihood criterion by the well known expectation maximization (EM) algorithm. Comparisons with some other methods such as the EM algorithm, MUSIC and accelerated random search (ARS) demonstrates the superiority of the proposed IS-based multipath delay estimator in terms of estimation performance and complexity.

Residual self-interference after cancellation in full-duplex systems

We investigate the signal-to-residual-interference ratio (SIRout) in a full-duplex transceiver with analog self-interference cancellation in consideration of three major sources of imperfection: (i) self-interference channel estimation error, (ii) quantization error in the receiver analog-to-digital converter (ADC), and (iii) quantization error in the digital-to-analog converter (DAC) used to generate the self-interference replica. In particular, we first derive the Cramér-Rao lower bound on the variance of the self-interference channel estimation error, and use it to further develop a closed-form expression of the SIRout. The resulting SIRout expression facilitates a study of the limit of a full-duplex system and determines the minimum required resolution for the ADC and DAC in order to meet a given performance. The expression reveals that, with a sufficiently high number of bits, the effects of ADC and DAC are negligible, but the cancellation performance is limited by the thermal noise and, in the best case, we can obtain a SIRout equal to the received signal-to-thermal-noise ratio (SNR).

A Maximum-Likelihood Channel Estimator in MIMO Full-Duplex Systems

This paper focuses on the channel estimation for residual self-interference cancellation at the baseband in a full-duplex transceiver. In particular, we analyze and develop a semi-blind maximum-likelihood algorithm to jointly estimate both the residual self-interference channel and intended signal channel based on the perfectly known transmitted symbols from its own transmitter, and both known pilot and unknown data symbols sent from the other intended transmitter. We first derive a closed-form solution for the channel estimate, and subsequently develop an iterative procedure to improve the estimation performance of the closed- form approach at high SNR. The iterative algorithm is guaranteed to converge to the ML solution when properly initiated. Simulation results show that, with a modest complexity, the proposed algorithm can offer good channel estimation MSE that follows well the Cramer-Rao bound (CRB), and good cancellation performance for a large SNR range.

Self-interference cancellation for full-duplex MIMO transceivers

Full-duplex operation requires effective self-interference (SI) cancellation that in turn needs reliable SI channel estimation. In this paper, we develop two estimation algorithms suitable for a 2-stage SI cancellation structure. By exploiting the sparsity of the SI channel, we first derive a compressed sensing-based SI channel estimation algorithm to be used in the first SI cancellation stage at radio-frequency (RF) to reduce the SI. We then develop a subspace-based algorithm to jointly estimate the residual SI channel, the intended channel and the transmitter nonlinearities for the second SI cancellation stage at baseband. Including the intended received signal in the estimation process is the main advantage of the proposed algorithm as compared to previous works that assume it as additive noise. Simulation results show that the proposed algorithms outperform the least-square (LS) algorithm and offer higher signal-to-residual-interference-and-noise ratio (SINR) over a large received signal-to-noise ratio (SNR) range.

A digital subspace-based self-interference cancellation in full-duplex MIMO transceivers

This paper addresses the problem of digital self-interference (SI) cancellation in full-duplex systems. Under practical transmitter imperfections, the received SI is affected by transmitter nonlinearities and propagation channel, which need to be estimated in order to cancel the SI. The proposed estimation method is based on subspace decomposition. The major detriment of subspace technique is the need of oversampling or multisensor receiver to obtain a nondegerate noise subspace. We modify the traditional subspace techniques by exploiting the covariance and the pseudo-covariance of the received signal. This enables us to increase the dimension of the received signal without resulting to oversampling or multisensor receiver. The different parameters are estimated, up to an ambiguity term, without any knowledge of the intended signal. We develop a joint detection and ambiguity identification procedure that requires a considerably smaller number of pilots than standard training-based methods. Simulation results show that the proposed algorithm can properly estimate the SI channel coefficients and the nonlinear parameters without any pilot symbol from the intended transmitter.

A new importance-sampling-based non-data-aided maximum likelihood time delay estimator

In this paper, we present a new non-data-aided (NDA) maximum likelihood (ML) time delay estimator based on importance sampling (IS). We show that a grid search and lack of convergence from which most iterative estimators suffer can be avoided. It is assumed that the transmitted data are completely unknown at the receiver. Moreover the carrier phase is considered as an unknown nuisance parameter. The time delay remains constant over the observation interval and the received signal is corrupted by additive white Gaussian noise (AWGN). We use importance sampling to find the global maximum of the compressed likelihood function. Based on a global optimization procedure, the main idea of the new estimator is to generate realizations of a random variable using an importance function, which approximates the actual compressed likelihood function. We will see that the algorithm parameters affect the estimation performance and that with an appropriate parameter choice, even over a small observation interval, the time delay can be accurately estimated at far lower computational cost than with classical iterative methods.