Maximum likelihood (ML) amplitude and noise variance estimation without having to jointly estimate the frequency and the phase and based only on information from the noisy received signal magnitude, is studied for a single sinusoid in complex additive white Gaussian noise. This estimation problem is equivalent to the classic problem of parameter estimation for the Rician distribution. While solving the likelihood equation is impossible in general, we propose a new approach based on a large argument approximation. For the case with known noise variance, a closed-form ML amplitude estimator is obtained, which outperforms the conventional root-mean-square estimator. For the case with unknown noise variance, the closed-form joint amplitude and noise variance estimators obtained do not require prior knowledge of one another.

The blind quality evaluation of screen content images (SCIs) and natural scene images (NSIs) has become an important, yet very challenging issue. In this paper, we present an effective blind quality evaluation technique for SCIs and NSIs based on a dictionary of learned local and global quality features. First, a local dictionary is constructed using local normalized image patches and conventional -means clustering. With this local dictionary, the learned local quality features can be obtained using a locality-constrained linear coding with max pooling. To extract the learned global quality features, the histogram representations of binary patterns are concatenated to form a global dictionary. The collaborative representation algorithm is used to efficiently code the learned global quality features of the distorted images using this dictionary. Finally, kernel-based support vector regression is used to integrate these features into an overall quality score. Extensive experiments involving the proposed evaluation technique demonstrate that in comparison with most relevant metrics, the proposed blind metric yields significantly higher consistency in line with subjective fidelity ratings.

This paper aims at providing explicit, closed-form solutions to the error probability performance analysis of digital communications over fading channels. This is achieved by first deriving a family of new upper bounds on the Gaussian Q-function Q(x), which is given by a sum of products of the exponential function and c/x where c is a constant. The bounds obtained can be made arbitrarily tight as the number of summation terms increases, and thus, can be used to approximate Q(x) accurately. Their applications to the performance analysis over fading are then presented to highlight the significance of the bounds derived. Both short-term fading and combined short-term and long-term fading are considered. It is shown that the new bounds can lead to better performance than the popular exponential-type upper bounds.

This paper is concerned with performance analysis (in terms of bounds and approximations to the average symbol error probability (ASEP)) of a product and power of Gaussian Q-functions over Nakagami-m fading. The results are valid for arbitrary product/power order. This is done by first deriving a family of new, simple lower bounds on the Gaussian Q-function, which is obtained as a sum of products of an exponential function and cx where c is a constant. These lower bounds can be made arbitrarily tight as the number of summation terms increases, and thus, can be used to approximate the Gaussian Q-function accurately. Their applications to the evaluation of the ASEP are then presented. Some advantages of the results derived here over those given in the literature are briefly discussed.

We consider a differential amplify-and-forward transmission scheme for a two-user cooperative system with a single relay. We propose a new analytical approach based on first principles, and obtain a closed-form expression for the exact bit error probability (BEP) in Rayleigh fading. This result involves only a single exponential integral function. It applies generally to the case where the channel statistics on different links are non-identical. As expected, the BEP decreases monotonically with the source transmission energy, or the relay transmission energy. The explicit BEP result allows us to study optimum energy allocation between the source and the relay, given a total transmission energy. By lower bounding the exact BEP expression via Jensen's inequality, we obtain analytically a closedform approximate optimum energy allocation solution between the source and the relay. Numerical results show that the BEP bound and the approximate optimum energy allocation solution are very tight.

Automatic repeat request (ARQ) schemes with cyclic redundancy check (CRC) codes are designed specifically for additive white Gaussian noise channels based on average performance measures. Average performance measures are known to be inadequate for high data rate wireless communication where the channel gain varies randomly in time, because they do not reflect the poor QoS in deep fades. The instantaneous accepted packet error outage (IAPEO) probability is a more meaningful performance measure. Therefore, we examine the IAPEO of conventional ARQ and observe that, to achieve a certain IAPEO performance, the system must operate above a minimum signal-to-noise ratio (SNR), which may be too high to achieve. This motivates us to propose ARQ with channel gain monitoring (CGM), in which a packet is only demodulated and checked for errors when the channel gain exceeds a threshold; otherwise retransmission is requested. We show that it does not have a minimum SNR requirement. The average performance measures of ARQ-CGM with selective-repeat, go-back-n and stop-and-wait retransmission protocols, i.e. the average accepted packet error probability, throughput and goodput, are also obtained as functions of the pilot length and the CRC code length. The allocation of bandwidth and energy between channel estimation and CRC coding is studied.

The probability of instantaneous accepted packet error outage (IAPEO) is proposed as a performance measure for packet transmission with automatic-repeat-request (ARQ) schemes over wireless channels. It is defined as the probability that the instantaneous accepted packet error probability (IAPEP) exceeds an IAPEP threshold. We obtain a closed-form upper bound expression for IAPEO of a pure ARQ over Rayleigh fading, and observe that, in order to satisfy a system design requirement of maximum tolerable IAPEO, the system must operate above a minimum SNR value. We next propose an ARQ scheme by incorporating channel gain monitoring (ARQCGM), such that the IAPEO requirement can be satisfied at any SNR value with the right channel gain threshold. The IAPEO performance of ARQ-CGM with selective repeat retransmission protocol is related to the conventional performance measures, i.e. average accepted packet error probability, throughput and goodput.

Using the convexity property of the exponential function, we obtain a family of exponential lower bounds on the Gaussian Q-function using the Jensen's inequality. The tightness of the bounds can be improved by increasing the number of exponential terms. The coefficients of the exponentials are constants, allowing easy averaging over the fading distribution using the moment generating function method. This method is also applied to the symbol error probability of M-ary phase shift keying and M-ary differential phase shift keying over additive white Gaussian noise and fading channels. The tightness of the bounds is demonstrated.

We propose to use the probability of instantaneous symbol error outage (ISEO) as a performance measure for digital communication over wireless channels. It is defined as the probability that the instantaneous symbol error probability (ISEP) exceeds an ISEP threshold. We analyze the impact of imperfect channel state information (CSI) on the ISEO probability over Rayleigh fading channels. A simple, but tight, closed-form upper bound on the ISEO probability is obtained as a function of the channel estimation mean square error (MSE). It is shown that the ISEO performance improves rapidly with decreasing MSE, when MSE drops below a certain value, which is determined partly by the ISEP threshold chosen. As CSI is commonly obtained by pilot-symbol-assisted channel estimation in packet transmission, we obtain the optimum allocation of pilot and data energy in a frame that minimizes the ISEO probability, which also minimizes the instantaneous packet error outage probability.

In this letter, we extend the baseband impulse noise modeling over the subscriber loop conducted over the German Telephone Network to the passband scenario. We apply an in-phase (I) and quadrature (Q) resolution of the passband impulse noise and model these components as a generalized error distribution. We use simulation to verify our postulation. Closed-form symbol error rate (SER) expressions for M-ary quadrature amplitude modulation over this passband impulse noise channel are then derived. From the computed SER results, we conclude that the system performance is not sensitive to the dependency factor between the I and Q noise components.