Multilayer Perceptron Equalizer Research Articles

Reduced Complexity Neural Network Equalizers for Two-Dimensional Magnetic Recording

This article investigates the reduced complexity neural network (NN)-based architectures for equalization over the two-dimensional magnetic recording (TDMR) digital communication channel for data storage. We use realistic waveforms measured from a hard disk drive (HDD) with TDMR technology. We show that the multilayer perceptron (MLP) nonlinear equalizer achieves a 10.91% reduction in bit error rate (BER) over the linear equalizer with cross-entropy (CE)-based optimization. However, the MLP equalizer’s complexity is <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$6.6\times $ </tex-math></inline-formula> the linear equalizer’s complexity. Thus, we propose the reduced complexity MLP (RC-MLP) equalizers. Each RC-MLP variant consists of finite-impulse response (FIR) filters, a nonlinear activation, and a hidden delay line. A proposed RC-MLP variant entails only <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1.59\times $ </tex-math></inline-formula> the linear equalizer’s complexity while achieving a 8.23% reduction in BER over the linear equalizer.

Signal Detection Using Extrinsic Information From Neural Networks for Bit-Patterned Media Recording

To meet the demand for storing large amounts of data, there has been a stronger focus on increasing the recording density of a hard disk drive (HDD). While conventional HDD suffers from thermal effect and superparamagnetic limit, bit-patterned media recording (BPMR) has the potential to resolve these problems and is capable of increasing areal density (AD) beyond 1 terabit per square inch. To increase the AD of BPMR, the distance between islands in both the down- and cross-track directions must be reduced. Consequently, the decreased bit period and track pitch induce more intersymbol interference (ISI) and intertrack interference (ITI), which degrade the bit error rate (BER) performance. In this study, extrinsic information obtained from a multilayer perceptron (MLP) equalizer is used as a priori information in a partial response maximum-likelihood (PRML) detector to improve the detection performance of BPMR. The proposed detection scheme shows better performance compared to both the MLP equalizer and conventional PRML alone. In addition, it was found that the proposed detector provides improved performance when track misregistration (TMR) occurs.

An Efficient ANN Interference Cancelation for High Order Modulation over Rayleigh Fading Channel

High order modulation (HOM) presents a key challenge in increasing spectrum efficiency in 4G and upcoming 5G communication systems. In this paper, two non-linear adaptive equalizer techniques based on multilayer perceptron (MLP) and radial basis function (RBF) are designed and applied on HOM to optimize its performance despite its high sensitivity to noise and channel distortions. The artificial neural network’s (ANN) adaptive equalizer architectures and learning methods are simplified to avoid more complexity and to ensure greater speed in symbol decision making. They will be compared with the following popular adaptive filters: least mean square (LMS) and recursive least squares (RLS), in terms of bit error rate (BER) and minimum square error (MSE) with 16, 64, 128, 256, 512 and 1024 quadrature amplitude modulation (QAM). By that, this work will show the advantage that the MLP equalizer has, in most cases, over RBF and traditional linear equalizers.

Performance Comparison of Equalization Techniques for SI-POF Multi-Gigabit Communication With PAM- M and Device Non-Linearities

In this paper, Gigabit per second transmission over short-range step-index plastic optical fibre (SI-POF) is implemented with multi-level pulse amplitude modulation (PAM- M ) scheme using a laser diode (LD) as the optical source. In particular, the POF channel distortion and the non-linear distortion from the LD are considered. To mitigate these distortions under various PAM- M levels, the bit error rate (BER) performances of the transversal, Volterra and multi-layer perceptron (MLP) based decision feedback equalizers are evaluated and compared. Furthermore, the inverse polynomial technique for non-linear mitigation is also considered. MLP is shown to offer the best BER performance for PAM scheme using more than eight levels especially for a system with high non-linearities. With the MLP equalizer, we record the highest data rate of 3, 7.8, and 18 Gbps for POF lengths of 60, 30, and 10 m, respectively.

A 20-Mb/s VLC Link With a Polymer LED and a Multilayer Perceptron Equalizer

This letter experimentally demonstrates a visible light communication system using a 350-kHz polymer light-emitting diode operating at a total bit rate of 19 Mb/s with a bit error rate (BER) of 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-6</sup> and 20 Mb/s at the forward error correction limit for the first time. This represents a remarkable net data rate gain of ~55 times. The modulation format adopted is ON-OFF keying in conjunction with an artificial neural network classifier implemented as an equalizer. The number of neurons used in the experiment is varied from the set N = {5, 10, 20, 30, 40} with 40 neurons offering the best performance at 19 Mb/s and the BER of 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-6</sup> .

Enhancement of New Channel Equalizer Using Adaptive Neuro Fuzzy Inference System

The rapidly increasing need for information communication requires higher speed data transmission over the existing channels. The data rate over these channels is limited mainly by Inter Symbol Interference (ISI). Channel equalizers are used to reduce the effect of ISI. In this paper, a new equalizer based on Adaptive Neuro-Fuzzy Inference System is presented. The performance of the proposed equalizer is evaluated for both linear as well as non-linear channels in terms of bit-error rate for different noise powers. Simulation results show that the proposed equalizer has better Bit Error Rate (BER) performance compared to multi-layer perceptron and least mean square equalizers. However, its BER performance is slightly poorer than that of radial basis function network and optimal Bayesian equalizer but is better in terms of structural complexity. Keywords: Channel equalizer, Hybrid learning algorithm, Intersymbol interference, Membership function, optimal Bayesian equalizer. The problem of equalization is treated in two different ways. Firstly, it may be interpreted as an inverse filtering problem, so that the combination of the channel and equalizer must behave like ideal channel .The conventional equalizers such as Linear, Decision-feedback, and Fractionally- spaced equalizers are based on this approach. But their performance is poor especially when the channel response is non-linear. Secondly, the problem of equalization may be considered as a classification problem. All equalizers using neural networks are based on the latter approach.. The capabilities of neural networks for equalization of simple channels are described by Gibson, Siu and Cowan. Multi-layer perceptron (MLP) equalizers are superior to conventional linear and decision feedback equalizers in terms of probability of error, but their practical applications are severely restricted due to difficulties such as very long training time, indeterminate nature of training time, and lack of a methodology for architecture selection. Radial basis function network (RBFN) equalizer has received a great deal of attention because of its structural simplicity and more efficient learning compared to MLP. Moreover, there exists structural equivalence between RBFN equalizer and optimal symbol spaced Bayesian equalizer . However, the structure of RBFN equalizer depends on channel length, and thus requires a large number of centers in the hidden layer, which increases the computational complexity. In this paper, we propose a new channel equalizer based on adaptive neuro-fuzzy inference system, i.e., an adaptive fuzzy inference system whose parameters are adapted using neural network learning algorithms. Herein hybrid learning algorithm is used for optimizing parameters of ANFIS. The performance of the proposed equalizer is evaluated in terms of bit-error rate (BER) for different noise powers in the channel. Both linear as well as non-linear channels are considered for performance evaluation. The BER of proposed equalizer is compared with conventional least mean square (LMS) equalizer, MLP equalizer, RBFN equalizer, and optimal Bayesian equalizer. Simulation results show that proposed equalizer has better BER performance compared to MLP and LMS equalizers, but slightly poorer than RBFN and optimal Bayesian equalizers. However, the proposed equalizer has smaller structure than RBFN equalizer, thus provides a better compromise between the performance and structural complexity.

Conditional distribution learning with neural networks and its application to channel equalization

We present a conditional distribution learning formulation for real-time signal processing with neural networks based on an extension of maximum likelihood theory-partial likelihood (PL) estimation-which allows for (i) dependent observations and (ii) sequential processing. For a general neural network conditional distribution model, we establish a fundamental information-theoretic connection, the equivalence of maximum PL estimation, and accumulated relative entropy (ARE) minimization, and obtain large sample properties of PL for the general case of dependent observations. As an example, the binary case with the sigmoidal perceptron as the probability model is presented. It is shown that the single and multilayer perceptron (MLP) models satisfy conditions for the equivalence of the two cost functions: ARE and negative log partial likelihood. The practical issue of their gradient descent minimization is then studied within the well-formed cost functions framework. It is shown that these are well-formed cost functions for networks without hidden units; hence, their gradient descent minimization is guaranteed to converge to a solution if one exists on such networks. The formulation is applied to adaptive channel equalization, and simulation results are presented to show the ability of the least relative entropy equalizer to realize complex decision boundaries and to recover during training from convergence at the wrong extreme in cases where the mean square error-based MLP equalizer cannot.

Using recurrent neural networks for adaptive communication channel equalization

Nonlinear adaptive filters based on a variety of neural network models have been used successfully for system identification and noise-cancellation in a wide class of applications. An important problem in data communications is that of channel equalization, i.e., the removal of interferences introduced by linear or nonlinear message corrupting mechanisms, so that the originally transmitted symbols can be recovered correctly at the receiver. In this paper we introduce an adaptive recurrent neural network (RNN) based equalizer whose small size and high performance makes it suitable for high-speed channel equalization. We propose RNN based structures for both trained adaptation and blind equalization, and we evaluate their performance via extensive simulations for a variety of signal modulations and communication channel models. It is shown that the RNN equalizers have comparable performance with traditional linear filter based equalizers when the channel interferences are relatively mild, and that they outperform them by several orders of magnitude when either the channel's transfer function has spectral nulls or severe nonlinear distortion is present. In addition, the small-size RNN equalizers, being essentially generalized IIR filters, are shown to outperform multilayer perceptron equalizers of larger computational complexity in linear and nonlinear channel equalization cases.