Blind channel estimation and data detection using hidden Markov models

In this letter, we jointly consider the problems of blind channel estimation (BCE), channel equalization and data detection in underwater visible light communication (UVLC) systems. Specifically, we propose to solve them with a model-driven deep learning (DL) approach. Inspired by the module-based signal processing methodology and communication theory, we devise a new blind receiver architecture called blind detection network (BDNet), which consists of three modules corresponding to BCE, channel equalization, and symbol demapping, respectively. Different from existing BCE schemes, the BDNet does not directly estimate the channel itself, but instead estimates the inverse channel regardless of the scalar ambiguity issue by learning the latent channel features from the received signal only. The UVLC channel effects can then be partially compensated by the zero-forcing equalization module of BDNet. Finally, the symbol demapping module recovers the transmitted symbols with the aid of the learned decision thresholds. Simulation results show the competitive bit error rate performance of BDNet in comparison to conventional pilot-aided schemes.

This paper presents new transceiver framework for the DS-CDMA systems that use the multiple-input multiple-output (MIMO) antennas, with space-time block code (ST-BC). In the transmitter, new hybrid (prefix/postfix) zero-padding sequences combining with the desired user signature is exploited to form the hybrid augmented user signature sequences, with respect to different transmit-antennas. While in the receiver under the GSC framework a new Capon-like semi-blind MIMO-CDMA filterbank detector based on the linearly constrained constant modulus (LCCM) criterion, is derived. Also, the adaptive GSC-RLS algorithm is developed for adaptively implementing the blind LCCM MIMO-CDMA filterbank of the receiver. With the specific design of the hybrid augmented user signature sequences, we are able to resolve phase ambiguity problem of blind channel estimation. Computer simulations verify the merits of the proposed new transceiver framework; it can be employed to achieve better performance, and is more robust against the discrepancy due to Capon channel estimation.

In many practical parameter estimation problems, the statistical properties of the sources can be exploited to improve the quality of the estimates. In this paper we consider the correlation and Kullback matching criteria (CM and KM respectively), which are applied to the problem of blind channel estimation under orthogonal space-time block coded (OSTBC) transmissions. Specifically, it is shown that the special OSTBC structure provides straightforward closed form solutions, which reduce to the extraction of the principal eigenvector of the observation correlation matrix modified by the code matrices and a set of weighting factors. Additionally, we prove that the KM technique is equivalent to the CM approach for low SNRs, and to the relaxed blind maximum likelihood (ML) decoder for high SNRs. Finally, the performance of the proposed techniques is illustrated by means of several simulation examples.

This paper considers the problem of blind channel estimation of multi-channel FIR filters. This is a problem arising in, for example, mobile communication systems using digital signalling. By using the orthogonality property between the noise subspace and the channel matrix, it has been shown in earlier work that the channel matrix is identifiable up to a multiplicative constant. In this article, the asymptotic properties of a subspace method using this orthogonality property is presented. An asymptotically correct weighting matrix is derived, demonstrating an attainable lower theoretical bound using the subspace estimate.

In this paper, we consider the problem of blind channel estimation for single-input single-out (SISO) orthogonal frequency-division multiplexing (OFDM) systems via second order statistics (SOS) only. Based on the assumption that the transmitted symbols are independent and identically distributed (i.i.d), we develop a simple subspace based blind channel estimation technique by utilizing a non-redundant linear block precoding. The proposed method offers improved performance over the existing method using similar precoding technique. Additionally, the design criteria for the precoders by which the performance can be greatly improved is discussed, and the stochastic Cramer-Rao bound (CRB) is derived. The numerical results clearly show the effectiveness of our proposed algorithm

In this paper, the problem of blind channel estimation under orthogonal space-time block coded (OSTBC) transmissions is solved by minimizing some distance measure between the theoretical and estimated correlation matrices of the observations. Specifically, the minimization of the Euclidean distance and the Kullback-Leibler divergence leads, respectively, to the Euclidean correlation matching (ECM) and Kullback correlation matching (KCM) criteria. The proposed techniques exploit the knowledge of the source correlation matrix to unambiguously recover the multiple-input multiple-output (MIMO) channel. Furthermore, due to the orthogonality properties of OSTBCs, both the ECM and KCM criteria result in closed form solutions. In particular, the channel estimate is given by the principal eigenvector of a matrix, which is obtained from the estimated correlation matrix of the observations modified by the code matrices and a set of weights. In the ECM case, the weights are fixed and equal to the eigenvalues of the source correlation matrix, whereas the KCM weights depend on both the signal-to-noise ratio (SNR) and the source eigenvalues. Additionally, we show that the proposed approaches are equivalent in the low SNR regime, whereas in the high SNR regime the KCM criterion is asymptotically equivalent to the relaxed blind maximum-likelihood (ML) decoder. Finally, the performance of the proposed criteria is illustrated by means of some numerical examples.

Musicologists and musicians often would like to search by keys in a digital music library. In this paper, we introduce a new key-finding algorithm that can be applied to music in both symbolic and audio formats. The algorithm, which is based on a Hidden Markov Model (HMM), provides two stages of key-finding output; the first one referring to local keys and the second one to the global key. We describe the input, the two output stages, and the parameters of the model. In particular, we describe two configurable parameters, the transition probability distributions, which are based on a matrix of neighbouring keys, and the emission probability distributions, which make use of established key profiles. We discuss the local key-finding capabilities of the algorithm, presenting an example analysis of the Prelude Op. 28 No. 20 in C minor by Chopin, showing the local key regions obtained using different key profiles. We evaluate the global key-finding capabilities of the model, using an existing dataset and six well-known key profiles as different model parameters. Since different key profiles will tend to err or misclassify in different ways and across different pieces, we train an ensemble method with the predictions from all the key profiles (6) through our model. We show that the ensemble method achieves state-of-the-art performance for major and overall keys, however, it still underperforms the state-of-the-art for minor keys.

BACKGROUND A hidden Markov model (HMM) is used for exon prediction on DNA of genes Plasmodium falciparum that has a model structure based on exon region structure in coding sequence (CDS). The objective research was to develop a new structure model to predict exon on DNA-genes of Plasmodium falciparum based on CDS structure using the HMM system. METHODS Model design in CDS, between two exon regions can be found one intron region and the model state number is used for its region. Its state number is used by separating start codon from first exon region and stop codon from the last exon region up to 9. The Viterbi algorithm and the backward-forward method for transition as well as emission states are used for training process. Furthermore, Viterbi and Baum-Welch algorithms are used for the testing process. The correlation coefficient (CC) was used as performance indicator, as the ratio of the estimated state in the output and the original state in the input of the model. RESULTS The simulation results has shown that the CC values depend on the given of the backward-forward transition state values randomly. The model with state number 9 showed the highest average of CC values of 0.7289 for Viterbi algorithm, and is 0.7166 for Baum-Welch algorithm. However, the lowest average of CC values has been found for the model with state number five. Its values are 0.6735 by using Viterbi algorithm and 0.6661 by using Baum-Welch algorithm. CONCLUSION The new structure model based on HMM system was valid to predict exon on DNA-genes of Plasmodium falciparum.

Hidden Markov Models, also known as Markov Switching Models, can be considered an extension of mixture models, allowing for dependent observations. The main problem associated with Hidden Markov Models is represented by the choice of the number of regimes, i.e. the number of generating data processes, which differ one from another just for the value of the parameters. Applying a hierarchical Bayesian framework, we show that Reversible Jump Markov Chain Monte Carlo techniques can be used to estimate the parameters of the model, as well as the number of regimes, and to simulate the posterior predictive densities of future observations. Assuming a mixture of normal distributions, all the parameters of the model are estimated using a well known exchange rate data set.KeywordsMarkov Chain Monte CarloMarkov Chain Monte Carlo MethodDirichlet ProcessMarkov Chain Monte Carlo AlgorithmFinancial Time SeriesThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

In order to control and manage highly aggregated Internet traffic flows efficiently, we need to be able to categorize flows into distinct classes and to be knowledgeable about their different behaviours. We consider the problem of classifying BGP (border gateway protocol) level prefix flows into a small set of homogeneous classes. We argue that a simple two states hidden Markov model (HMM), even if not sufficient to describe a flow, is sufficient to distinguish between flows and to help in classifying them to homogeneous classes. We propose a classification method based on modelling windows of flow observations using the hidden Markov model and classifying them based on parameters of the model. We use a classical EM algorithm for estimating all model parameters as well as the flow membership probabilities - the probability that a flow belongs to any given class. One of our key contributions is a new and relatively fast method for Internet flow classification. The method is fast as it can classify flows over 30 minutes, i.e. it needs only 6 reports generated at 5 minute intervals (as is usually done for SNMP reports) opening the way for online flow classification and monitoring. We show that our method is powerful in that it is capable of examining macroscopic flows while simultaneously making fine distinctions between different traffic classes. We demonstrate that our scheme can address issues with flows being close to class boundaries and the inherent dynamic behaviour of Internet flows.

The discrete channel hidden Markov models indicate the temporal statistical characteristics of the generated error sequences by fading channels with memory precisely. In this paper first the optimal order of the hidden Markov models is estimated using Baum Welch algorithm (BWA) with comparing statistics such as average loglikelihood and autocorrelation function for error modelling of IEEE 802.16/WiMAX standard. As a perfect alternative to HMM error models, a new efficient stochastic swarm intelligent-based particle swarm optimization (PSO)-HMM is suggested for this standard to find the global and optimal solution. Comparing the proposed method with the conventional (BWA) by performing numerical simulations using error sequences of various lengths, the results indicate the superiority of the PSO approach even when the target sequences are as short as 2000 in length; moreover, the appropriateness of the PSO-HMM is investigated for different sampling periods. It can be concluded that the PSO with two states arrives at high level of accuracy in contrast to HMM with much higher number of states. For validating the proposed method, the burst error statistics of the PSO-HMM with respect to the autocorrelation function and error free interval distribution of the original error sequence are presented. The simulation results indicate that the PSO-HMM has better performance to find the optimal model parameters more stable than (BWA).

Publication in the conference proceedings of EUSIPCO, Lausanne, Switzerland, 2008

We propose an application of variational Bayesian inference to the problem of blind estimation of single-input multiple-output (SIMO), finite-impulse-response (FIR) channels. The blind channel estimation problem has been extensively studied well over a decade. Nevertheless, there are some types of channel that cannot be successfully estimated by existing blind methods. One example of such a channel type is frequency-selective, exponentially-weighted Rayleigh fading channel. Our blind estimation approach to it is to incorporate prior knowledge on channel weighting in the form of channel prior distribution, then obtain the approximation to the posterior distribution of the channel using the variational Bayesian framework. Despite the error due to the mean field approximation involved in the variational framework, the derived algorithm has been verified by simulation to provide much improvements over the existing ones in a statistical sense.

We propose a simple, fast, deterministic pre-fitting approach which derives the Baum–Welch algorithm initial values directly from the input data. Such pre-fitting has the purpose of improving the fitting time for a given Hidden Markov Model (HMM) while maintaining the original Baum–Welch algorithm as the fitting one. The fitting time is improved by avoiding the Baum–Welch algorithm sensitiveness through the generation of parameters closer to the global maximum likelihood. Furthermore, by keeping the original Baum–Welch algorithm as the fitting one, we guarantee that all related methods will continue to work properly. On the other hand, the pre-fitting generates the HMM parameters directly derived from time-series data, without any data transformation, using an [Formula: see text] operation.

Credit scoring become an important task to evaluate an applicant by a banker. Many models and tools are available for making initial lending decisions. This paper presents an Hidden Markov Models (HMMs) for modeling credit scoring problems. Baum-Welch algorithm - an iterative process for estimating HMM parameters are often used to developed such models and improve the pattern recognition for many problems. We introduce HMM/Baum-Welch initial model selection: a tool developed to test the impact of choosing initial model to train Baum-Welch process. Experiments results show that the performance of learned models depend on different way of generating the initial models used in credit scoring.