Tet Hin Yeap

Single longitudinal mode multiwavelength fiber ring lasers

A single longitudinal mode multiwavelength fiber ring laser is demonstrated. The single longitudinal mode multiwavelength operation is obtained using a fiber loop mirror with a saturable absorber and a Lyot-Sagnac filter or a fiber grating array.

Decision feedback recurrent neural equalization with fast convergence rate

Real-time recurrent learning (RTRL), commonly employed for training a fully connected recurrent neural network (RNN), has a drawback of slow convergence rate. In the light of this deficiency, a decision feedback recurrent neural equalizer (DFRNE) using the RTRL requires long training sequences to achieve good performance. In this paper, extended Kalman filter (EKF) algorithms based on the RTRL for the DFRNE are presented in state-space formulation of the system, in particular for complex-valued signal processing. The main features of global EKF and decoupled EKF algorithms are fast convergence and good tracking performance. Through nonlinear channel equalization, performance of the DFRNE with the EKF algorithms is evaluated and compared with that of the DFRNE with the RTRL.

Fuzzy control of spindle torque for industrial CNC machining

Developing a dedicated control system for each and every machining process or machine is costly and time-consuming. Such a practice has obviously undermined the usefulness of many current systems. This paper presents a fuzzy control system that can be used for different machining processes. This system consists of a basic fuzzy logic controller, a fuzzy rule base, and a tuning mechanism used to enhance the adaptability of the system. Industrial tests have been carried out for both end milling and turning processes. The control signal is spindle torque, readily available on many CNC machines. The test results show that the system performs well on both end milling and turning operations and can easily adapt to tool changes as well as workpiece material changes.

A joint wavelet lifting and independent component analysis approach to fault detection of rolling element bearings

Though wavelet transforms have been used to extract bearing fault signatures from vibration signals in the literature, detection results often rely on a proper wavelet function and deep wavelet decomposition. The selection of a proper wavelet function is time consuming and deep decomposition demands more computing effort. This is unsuitable for on-line fault detection. As such, we propose a joint wavelet lifting scheme and independent component analysis (ICA) approach to detecting weak signatures of bearing faults. The optimal envelope spectrum of independent components for signature extraction is selected based on the maximum energy and total energy of each independent component. The performance of the proposed method is evaluated by comparing with several other methods using both simulated and real vibration signals. The results reveal that the proposed method is more effective and robust in extracting bearing fault signatures. The following advantages of the proposed method have also been observed: (a) it is insensitive to wavelet selection and hence is less susceptible to ill selected wavelet function; (b) it is insensitive to the depth of wavelet decomposition, leading to an efficient algorithm; and (c) it takes advantage of ICA in fault detection without using multiple sensors as required in the original ICA.

Applications of random-pulse machine concept to neural network design

Neural networks can reach their true potential only when they are implemented in hardware as massively parallel processors. This paper presents the random-pulse machine concept and shows how it can be used for the modular design of neural networks. Random-pulse machines deal with analog variables represented by the mean rate of random-pulse streams and use simple digital technology to perform arithmetic and logic operations. This concept presents a good tradeoff between the electronic circuit complexity and the computational accuracy. The resulting neural network architecture has a high packing density and is well suited for very large-scale integration (VLSI). Simulation results illustrate the performance of the basic elements of a random-pulse neuron.

VHDL implementation of a turbo decoder with log-MAP-based iterative decoding

Turbo code is one of the most significant achievements in coding theory during the last decade. By concatenating two simple convolutional codes in parallel, it has been shown that transmission systems employing turbo codes could offer near-capacity performance. More importantly, by employing a suboptimal iterative decoding structure with soft-in/soft-out (SISO) maximum a posteriori-probability (APP) decoding algorithm, the near-capacity performance is achievable at a feasible decoding complexity. Given the outstanding performance of turbo code, the challenge now is to implement it into various communication systems at affordable decoding complexity using current very large scale integration (VLSI) technologies. In this paper, we first investigated the existing four different turbo decoding algorithms. Comparisons of both their performances and implementation complexities were performed. Log-maximum a posteriori (MAP) -based turbo decoding was found to offer the best performance-complexity compromise. A register-transfer-level (RTL) 12-bit fixed-point turbo decoder based on Log-MAP algorithm was then designed and simulated using VHDL as the hardware description language. The implemented RTL model was verified by comparing its performances with those obtained from a C-language implementation of the same turbo decoder.

A novel common-mode noise cancellation technique for VDSL applications

xDSL modems operate in frequency bands which coincide with many significant radio-frequency interference sources, particularly commercial AM radio. In these bands, the balance of most twisted-pair cables is low enough to allow substantial interference to transfer to differential mode, disrupting the transmitted information signal. Two novel narrowband and wideband common mode noise cancellation techniques are presented. Simulation results show that both common mode noise cancellation techniques could provide 20 to 30 dB improvement in those radio frequency bands. Construction and testing of a hardware prototype is also presented. With finite-precision effects, initial tests suggest that common mode noise cancellation in the order of 15-20 dB can still be achievable.

Nonlinear state-space modeling using recurrent multilayer perceptrons with unscented Kalman filter

A nonlinear black-box modeling approach using a state–space recurrent multilayer perceptron (RMLP) is considered in this paper. The unscented Kalman filter (UKF), which was proposed recently and is appropriate for state–space representation, is employed to train the RMLP. The UKF offers a derivative-free computation and an easy implementation, compared to the extended Kalman filter (EKF) widely used for training neural networks. In addition, the UKF has a fast convergence rate and an excellent capability of parameter estimation which are appropriate for online learning. Through modeling experiments of nonlinear systems, the effectiveness of the RMLP trained with the UKF is demonstrated.

A wideband crosstalk canceller for xDSL using common-mode information

This letter uses the twisted-pair common-mode signal as the input of an adaptive canceller to remove some wideband crosstalk in a digital subscriber line (xDSL) differential signal. Simulations using simple crosstalk and line balance models show the potential benefits of the canceller to improve the signal-to-noise ratio of an xDSL channel.

Neural Networks and Parkinson's Disease

A closed-loop or recurrent neural network was taught to generate output discharges to reproduce the prototypical activations in agonist and antagonist muscles which produce the displacement of a limb about a single joint. By introducing a generalized decrease in the excitability of the pre-output layer in the network, the network made the displacement more slowly and also showed an inability to maintain a repetitive movement. These concepts can be applied to the human nervous system in the understanding of the physical basis of movement and its disorders. It is suggested that a movement represents the output of a closed-loop network, such as the cortical-basal ganglia-thalamic-cortical motor loop, which iterates repetitively to its end point or attractor. The model provides an explanation of how the state of thalamic inhibition seen in Parkinsons disease physically may produce bradykinesia and the inability to maintain a repetitive movement.