Adaptive Filters

This paper introduces several new least mean-square (LMS) algorithms based on error normalization procedure. Different minimization approaches and techniques were used in developing the proposed algorithms. Some of these algorithms are selected and applied to an adaptive noise canceller setup with different stationary noise power levels. Simulation results, carried out using a real speech, clearly demonstrate the superiority of the proposed algorithms over other standard LMS algorithms. Smaller values of steady-state excess mean-square error and better tracking capabilities are obtained using these new algorithms.

The transient behavior of a Polymer Electrolyte Membrane Fuel Cell System (PEMFCS) under an efficient Adaptive Predictive Control with Robust Filter (APCWRF) is analyzed. This control scheme is tested to evaluate its performance when sudden changes in the load occur. It is produced by the demands of the electric motor of a hybrid vehicle, powered by a PEMFC and a supercapacitor bank to fulfil Standard Driving Cycles. The objective of the proposed advanced strategy is to control the oxygen excess ratio in the cathode to improve the system efficiency and to ensure a safe operation for the PEM. Several results through a simulation environment are presented. They are useful for showing the potentiality of the APCWRF for the proposed exigent scenarios .

Many tomographic source localization algorithms used in biomagnetic imaging assume, explicitly or sometimes implicitly, that the source activity at different brain locations are either independent or that the correlation structure between sources is known. Among these algorithms is a class of adaptive spatial filters known as beamformers, which have superior spatiotemporal resolution abilities. The performance of beamformers is robust to weakly coherent sources. However, these algorithms are extremely sensitive to the presence of strongly coherent sources. A frequent mode of failure in beamformers occurs with reconstruction of auditory evoked fields (AEFs), in which bilateral auditory cortices are highly coherent in their activation. Here, we present a novel beamformer that suppresses activation from regions with interfering coherent sources. First, a volume containing the interfering sources is defined. The lead field matrix for this volume is computed and reduced into a few significant columns using singular value decomposition (SVD). A vector beamformer is then constructed by rejecting the contribution of sources in the suppression region while allowing for source reconstruction at other specified regions. Performance of this algorithm was first validated with simulated data. Subsequent tests of this modified beamformer were performed on bilateral AEF data. An unmodified vector beamformer using whole head coverage misplaces the source medially. After defining a suppression region containing the temporal cortex on one side, the described method consistently results in clear focal activations at expected regions of the contralateral superior temporal plane.

While the LMS algorithm and its normalized version (NLMS), have been thoroughly used and studied, and connections between the Kalman filter and the RLS algorithm have bean established, the connection between the Kalman filter and the LMS algorithm has not received much attention. By linking these two algorithms, a new normalized LMS algorithm can be derived that has some advantages to the classical one. Firstly, its stability is guaranteed since it is a special case of the Kalman filter. Secondly, it suggests a new way to control the step size that results in optimum convergence for a large range of input signal powers, that occur in many applications. Finally, it prevents measurement noise amplification that occurs in the NLMS algorithm for low order filters, like the ones used in OFDM equalization systems.

Multichannel speech enhancement has been shown to be an effective method to decrease speech distortion introduced during speech enhancement, especially in environments like MRI (magnetic resonance imaging) which have a distributed noise source. However, these methods suffer from high computational complexity which makes them almost impractical. The use of subband filtering has been suggested to reduce this complexity but the performance of the existing subband methods deteriorate as the number of subbands increases. In this paper we introduce a new multichannel speech enhancement algorithm based on subband adaptive filtering that works for higher number of subbands at a lower complexity. The real-world experiments demonstrate the performance of the new scheme in an MRI room.

The article presents selected objective and subjective methods to assess the quality of voice services in telephone communication. The authors described the impact of the line and acoustic echo phenomena and ambient noise on the functions of telephone communication systems with respect to the speakerphone mode. Additionally, they discussed the possibilities to apply digital technology to improve the quality of voice services in telephone communication and alarm systems.

Approved for release; distribution unlimited. 17. DIST IIUTION STATEMAEhT (i the miktrace entrogdin Block 20. f iiffllt hum Report) II. SUPOLCENENANY NOTES It. KEY WORDS (Cmehawe an ilee ai* li nerelor an Idsmolp Wy bloked numb~e) 1hie report describes the research activities performed by t eUn vrsity of Southern California for the period I April 1980 to 31 March 1982 er the Contract No. I. N00014 -80-C -0457 with the Office of Naval Rese rch.-e research activities have focussed on the VLSI array processor for signal processing theory and algoritms and the development of parallel computing architec-A solution In today's VLSI research challenge 1lies In a cross-disciplinary research encompassing the areas of mathematics, algoritms, computers, and ap-D0 W ? 43 EDITiON or I wov 4& i ossoLaTe -7LSSFE S'N 062. ~. Old '401SECUIT CLASSIICATION Of TMIS PAG a N .-* * I7

Adaptive signal processing theory was born and has lived by exclusively exploiting the mean square error criterion. When we think of the goal of least squares without restrictions of Gaussianity, one has to wonder why an information theoretic error criterion is not utilized instead. After all, the goal of adaptive filtering should be to find the linear projection that best captures the information in the desired response. In this paper we summarize our efforts to extend adaptive linear filtering to information filtering. We briefly review Renyi's entropy definition, Parzen windows and put them together in a framework to estimate entropy directly from samples (nonparametric). Once this criterion is developed we can train linear or nonlinear adaptive networks for entropy maximization or minimization. We present results on the properties of the Renyi's nonparametric entropy estimator, and show how it performs in chaotic time series prediction.

Robots that are able to acquire an accurate model of their environment are regarded as fulfilling a major precondition of truly autonomous mobile vehicles. To learn a map of the environment, three problems need to be addressed simultaneously, namely exploration, mapping, and localization. In this work, we present an integrated solution to these three problems. To solve the simultaneous localization and mapping (SLAM) problem, our approach uses a highly efficient variant of algorithm proposed by Murphy and colleagues . In this algorithm, a Rao-Blackwellized particle filter (RBPF) is used to efficiently represent the joint posterior about possible maps and trajectories taken by the robot. The key contribution of our approach is an efficient decision-theoretic algorithm for computing vantage points that reduce the expected uncertainty in the RBPF. The approaches mostly related to our work have been presented by Makarenko et al. [3] and Bourgault et al. . They use an Extended Kalman Filter (EKF) to solve the SLAM problem and introduce a utility function which trades-off the cost of exploring new terrain with the potential reduction of uncertainty by measuring at selected positions. A similar technique has been applied by Sim et al. [5], who consider actions to guide the robot back to a known place in order to reduce the pose uncertainty of the vehicle during exploration. In contrast to our work, these approaches assume that the environment contains landmarks that can be uniquely determined during mapping. Our approach, in contrast, learns occupancy grid maps and thus is not restricted to environments with pre-defined landmarks. Compared to previous approaches, the novelty of the work reported here is that our algorithm simultaneously considers the uncertainty in the trajectory and in the map while building accurate occupancy grids. Based on an efficient scheme for computing the uncertainty of the joint posterior, we apply decisiontheoretic framework for choosing appropriate actions. Thereby, we utilize the properties of the Rao-Blackwellization. In brief, the uncertainty of an RBPF is determined by its entropy, which consists of two components. Whereas the first one computes the uncertainty over the trajectory taken by the robot, the second calculates the uncertainty of the individual maps weighted by the likelihood of the corresponding particles. We also demonstrate how the filter can be used to efficiently simulate possible measurements to be obtained along a path through already mapped terrain. To determine potential actions, we consider different action types including actions that collect sensor information about unknown areas as well as actions designed to re-visit known places to reduce the pose uncertainty of the vehicle. To finally select an action, we take into account the potential measurements gathered along the path of the robot as well as the cost for carrying out this action. Our approach has been implemented and tested in the real world and in simulation. Experimental results suggest that our approach leads to a robust exploration behavior that produces highly accurate grid maps. Furthermore, we illustrate the advantages of our action selection technique over previous approaches.

SummaryModels of the environment are needed for a wide range of robotic applications, from search and rescue to automated vacuum cleaning. Learning maps has therefore been a major research focus in the robotics community over the last decades. In general, one distinguishes between metric and topological maps. Metric maps model the environment based on grids or geometric representations whereas topological maps model the structure of the environment using a graph. The contribution of this paper is an approach that learns a metric as well as a topological map based on laser range data obtained with a mobile robot. Our approach consists of two steps. First, the robot solves the simultaneous localization and mapping problem using an efficient probabilistic filtering technique. In a second step, it acquires semantic information about the environment using machine learning techniques. This semantic information allows the robot to distinguish between different types of places like, e. g., ...

Simultaneous Localization and Mapping (SLAM) is one of the classical problems in mobile robotics. The task is to build a map of the environment using on-board sensors while at the same time localizing the robot relative to this map. Rao-Blackwellized particle filters have emerged as a powerful technique for solving the SLAM problem in a wide variety of environments. It is a well-known fact for sampling-based approaches that the choice of the proposal distribution greatly influences robustness and efficiency achievable by the algorithm. In this paper, we present an improved proposal distribution for grid-based SLAM with Rao-Blackwellized particle filters, which utilizes whole sequences of sensor measurements rather than only the most recent one. We have implemented our system on a real robot and evaluated its performance on standard datasets as well as in hard outdoor settings with few and ambiguous features. Our approach improves the localization accuracy and the map quality, substa...

This paper presents a nonlinear model predictive control (NMPC) strategy that can be used to tackle model predictive control problems that involve relatively simple nonlinear dynamic models, as for example obtained with firstprinciples modeling. The main feature of the proposed NMPC strategy is the usage of a moving horizon estimator (MHE) for the estimation of the states and disturbances (and, if desired, parameters). The closed-loop performance properties of the proposed NMPC strategy are demonstrated by applying it to a model of a municipal solid waste (MSW) combustion plant under a realistic disturbance realization. In addition, a comparison is made with extended Kalman filter (EKF) based NMPC.

The principle of adaptive noise cancellation is to acquire an estimation of the unwanted interfering signal and subtract it from the corrupted signal. Adaptive Noise Cancellation technique is an approach for powerful noise cancellation. In this paper the noise cancellation is performed using adaptive noise cancellers using an adaptive finite impulse response (FIR) filter are presented for the estimation of a transfer function of a noisy channel in a communication link. Optimizations of the Least Mean Square (LMS) versions and Normalized LMS (NLMS), Weiner algorithms are also used to adapt the filter coefficients of the estimated transfer functions in order to minimize the effect of background noise effectively & analyzed their performances with respect to their estimation of weights.

The grid-feeding inverter is the most popular choice for implementing a distributed generator (DG) in a photovoltaic (PV) system. Its role is to deliver active power with zero reactive power while connected to the grid. In most cases, the inverter is disconnected as soon as voltage sag disturbances occur, an operation that is known as an islanding mode. Recently, grid code regulations have been upgraded, allowing the inverter to remain connected to the grid and inject a certain amount of reactive power during voltage sags, provided that it meets grid code requirements. However, this can result in an increase in current injection by the inverter, which may cause overcurrent and damage the inverter if it exceeds the capability of the inverter. To control peak current during voltage sag disturbances, a proposed gridfeeding inverter attempts to detect voltage sag and calculate proportional injected active and reactive power. Once the disturbance has disappeared from the system, the inverters can be restored to normal operation. Prototype experiments have validated the ability of this system to control peak current during voltage sag while protecting the inverters.

It has been observed over last several decades that bandwidth greediness of applications never gets fulfilled. Hence, scientists, researchers, and engineers keep working on new ways of providing higher bandwidth. Recently, a new modulation technique called Orthogonal Frequency Division Multiplexing (OFDM) has been introduced which provides very high data rates. In OFDM the high frequency input signal is modulated over a large number of low frequency sub-carrier signals which are orthogonal to each other. This feature makes it very robust against efficiency degradation at higher frequencies. That is the reason why OFDM is a choice for the modern high and ultra-high data rate communication systems. However, it suffers from high levels of the peak power to the average power also called Peak-to-Average Power Ratio or PAPR. Reducing PAPR in OFDM is a hot research area. There are many schemes available which attempt to reduce PAPR. Some are in fact able to reduce PAPR but not sufficient e...

The S-Transform provides a framework for dataadaptive filters which take advantage of time-frequency localized spectra. These filters basically consist in a data transformation to the time-frequency domain, a dataadaptive weighting of the localized spectra, and a back transformation. We illustrate that the inverse S-Transform of manipulated spectra not necessarily transforms the localized signals as expected from the imposed weighting. The time localization is not directly translated and spurious signals and noise can be generated. We discuss this problem and suggest a new inverse S-Transform which may come helpful to many applications to take more advantage of the time-frequency localization.

The -transform is becoming popular for time-frequency analysis and data-adaptive filtering thanks to its simplicity. While this transform works well in the continuous domain, its discrete version may fail to achieve accurate results. This paper compares and contrasts this transform with the better known continuous wavelet transform, and defines a relation between both. This connection allows a better understanding of the -transform, and makes it possible to employ the wavelet reconstruction formula as a new inverse -transform and to propose several methods to solve some of the main limitations of the discrete -transform, such as its restriction to linear frequency sampling.

Target detection based on the representation of the hyperspectral image (HSI) has drawn the attention of researchers, given its powerful detection performance. The matrix-based approach inevitably loses spatial information and fails to explore the intrinsic multimodal structure of an HSI cube. In this paper, we propose a regularized tensor-based model without altering the data structure. The model assumes that an observed third-order HSI tensor is decomposed into the sum of a Total Variationregularized Low-rank background tensor and a Sparse (TVLrS) target tensor. The two tensors are then represented as the mode-3 product of a third-order tensor called the Tensor Representation Coefficient (TRC), by a matrix representing the spectra dictionary. The model is coined as TVLrS-TRC. Since the construction of a pure dictionary is essential for the background extraction, here, the background dictionary atoms are selected by the kernel spectral angle mapper (KSAM), while the target dictionary construction is done through the target atoms prior information. The tensor nuclear norm (TNN) is utilized to characterize the low rankness of the TRC subspaces, the TV norm is performed on each frontal slice of the TRC to encode its spatial smoothness, and the sparsity of the target TRC is characterized by the ℓ 1 -norm. Extensive experiments on two real hyperspectral data sets demonstrate the advantage of the proposed detector in comparison with several conventional and state-of-the-art target detectors.

The article presents selected objective and subjective methods to assess the quality of voice services in telephone communication. The authors described the impact of the line and acoustic echo phenomena and ambient noise on the functions of telephone communication systems with respect to the speakerphone mode. Additionally, they discussed the possibilities to apply digital technology to improve the quality of voice services in telephone communication and alarm systems.

We analyse and compare three algorithms for "downdating" the Cholesky factorization of a positive definite matrix. Although the algorithms are closely related, their numerical properties differ. Two algorithms are stable in a certain "mixed" sense, while the other is unstable. In addition to comparing the numerical properties of the algorithms, we compare their computational complexity and their suitability for implementation on parallel or vector computers. Only the Abstract is given here. The full paper appeared as . For an application of downdating, see .

Householder reflections applied from the left are generally used to zero a contiguous sequence of entries in a column of a matrix A. Our purpose in this paper is to introduce new rm Householder and TOW hyperbolic Householder reflections which are also applied from the left, but now zero a contiguous sequence of entries in a row of A. We then show how these row Householder reflections can be used to design efficient sliding-data-window recursive least-squares (RLS) covariance algo-

Based on the wavelet transformation and adaptive Wiener Filtering, a new method was presented by the authors to perform the synchronization and detection of the binary data from the Free-Space Optical (FSO) signal . It was shown in that the Haar wavelet with a fixed scale is an excellent choice for this purpose. The output of the filter is zero-mean and is closely related to the derivative of the binary data. In this effort an analysis of the work in [1] is presented to obtain the probability of bit error using a Bayesian ternary hypotheses testing. The analysis also results in determining optimum thresholds for the detection of binary data.

the performance of modern communication system can be reduced by non ideal characteristic of the channel, which is known as inter-symbol interference (ISI).The equalization technique is an efficient method to overcome ISI and improve the characteristics of the system. Adaptive linear filter and various categories of training algorithms are used to imitate different equalizer models. . Here we have simulated a digital communication model having quadrature amplitude modulation technique and additive white Gaussian noise channel implemented in MATLAB where Equalizer plays major role in this model. This paper presents the comparison of performance of linear & non linear adaptive channel equalizer trained by using gradient decent algorithm LMS & its variants such as NLMS, FBLMS, and SRLMS with fixed & variable step size in terms of Bit Error Rate (BER). Inspite of much prior work on this subject we reveal surprising analytical results in terms of well known BER

An image segmentation technique based on fuzzy sets for ultrasonic B-scans is presented. The fuzzy c-means algorithm is adapted to B-scan images. The classes constituting a region are obtained through fuzzy partitioning based on local statistics information. The perfonnance of the method is tested on phantom and tissue images and the results are presented.

The production of computer-generated video presents a number of difficulties not encountered with motion pietures. Interlaced scanning and the color subcarrier of NTSC video are responsible for special problems such as interline flicker, and chroma aliasing. As in motion pictures, temporal aliasing is also an issue. A renderer can sample and filter a moving image in an arbitrary manner and is not constrained to simply imitate the behavior of a television camera. This paper explores several different spatiotemporal antialiasing filters and how they affect the quality of video animation.

The production of computer-generated video presents a number of difficulties not encountered with motion pictures. Interlaced scanning and the color subcarrier of NTSC video are responsible for special problems such as interline flicker, and chroma aliasing. As in motion pictures, temporal aliasing is also an issue. A renderer can sample and filter a moving image in an arbitrary manner and is not constrained to simply imitate the behavior of a television camera. This paper explores several different spatiotemporal antialiasing filters and how they affect the quality of video animation.

In this letter, a new block adaptive decision feedback equalizer is developed. Both the feedforward (FF) and the feedback (FB) filters are updated once every K K K sample time intervals, with K K K being the block length. It should be noted that this block adaptation is done in such a way that the resulting filters, and the decisions as well, are identical to those computed by the conventional sample by sample LMS-based decision feedback equalizer (LMS-DFE). The new algorithm offers substantial computational savings as compared to the sample-by-sample LMS-DFE with no loss in performance. The new block DFE turns out to be particularly suitable for applications requiring long equalizers.

In this letter, a new block adaptive decision feedback equalizer is developed. Both the feedforward (FF) and the feedback (FB) filters are updated once every K K K sample time intervals, with K K K being the block length. It should be noted that this block adaptation is done in such a way that the resulting filters, and the decisions as well, are identical to those computed by the conventional sample by sample LMS-based decision feedback equalizer (LMS-DFE). The new algorithm offers substantial computational savings as compared to the sample-by-sample LMS-DFE with no loss in performance. The new block DFE turns out to be particularly suitable for applications requiring long equalizers.

Time-varying system behavior learning is crucial in the current landscape of machine learning, where statistical behavior often lacks precision. This approach effectively models many real-world scenarios characterized by their time-varying non-stationary nature. The focus has shifted towards adaptive systems for these applications, as they demonstrate superior performance compared to traditional fixed-weight systems that have been in use for a longer period. The introduction of multilayer adaptation has further enhanced the effectiveness of signal processing in this context. A single layer of adaptation utilizing a recursive or steepest-gradient filter is insufficient to ensure both rapid convergence and optimal error fidelity during the learning process, particularly in unknown time-varying environments. Therefore, a multilayer adaptation approach that combines least mean squares (LMS) and recursive least squares (RLS) algorithms has been proposed, incorporating a mixing parameter [5], [6], [7] that plays a pivotal role in the learning process. This multilayer adaptation has been shown to significantly improve learning capabilities, achieving approximately 4 to 5 dB better noise fidelity compared to existing single-layer adaptation methods.

However, when the channel state information (CSI) is unknown or unexpected, it might be difficult to use and alter the parameters of an adaptive channel equalizer, even though it has self-adjusting characteristics. In some cases, the off-base filter decision leads to an inefficient yield. An adaptive equalizer cannot simultaneously attain high convergence speed and the lowest MSE with more significant following capabilities. It is possible to formulate the two contradictory questions for flexible channels as lowering estimate noise and errors and altering the parameters under consideration. One adaptive filter in a channel equalizer is insufficient to accomplish the two opposing objectives listed above. The ensuing parameter diversity of the channel requires a quick adaptive filter. The disparity between the input and evaluated values is swiftly adjusted by the fast filter. To lower the estimate noise and inaccuracy in the channel, a significant filter is required. The convergence rate determines the rate at which the channel merges into its ultimate state. The faster rate of convergence of a flexible system is frequently a desirable property. Robust channel equalization should need a trade-off between convergence rate and least mean square error to satisfy the competing objectives of the channel equalizer. In a channel equalizer, the flexible conspire should be narrowly targeted using the least mean square error. Consequently, a multilayer adaptation-based machine learning (ML) strategy combining the recursive least squares (RLS) and least mean squares (LMS) algorithms has been created, incorporating a mixing parameter that is crucial to the equalizing process. This multilayer adaptation has been shown to significantly improve equalizing capabilities, resulting in noise fidelity that is higher than that of existing single-layer equalization methods.

In this paper, the fault detection (FD) problem of uncertain discrete-time switched systems under the constrained switching law and external disturbances is addressed. In the multiple Lyapunov functions (MLF) framework with H -/H ∞ performance index, a robust fault detection filter (RFDF) is designed in order to generate a residual signal which is sensitive to the faults and robust against the disturbances. The design method is formulated to be solved by linear matrix inequality (LMI) technique. The main contribution in this paper is to improve the performance of the FD for switched systems by considering switch detection logics. The design of the RFDF is based on the local performance index of each subsystem. An adaptive threshold is set based on both, the bounds of disturbances and the individual performance index for each subsystem.

The paper presents a hearing aid (HA) system based on a low-delay cochlear filter bank derived from the discrete cochlear model. The spectral gain shaping method (SGSM) that includes dynamic range compression, hearing loss compensation and noise reduction is applied in perceptually matched frequency bands. The acoustic feedback cancellation is implemented as an off-the-forward path scheme and does not introduce an additional delay to the forward path. The cancellation itself is made by adaptive filtering of non-uniform subband signals obtained from an oversampled warped cosine-modulated filter bank.

Radio frequency seeker filter design for homing guidance requirements are highly demanding and challenging. Seeker data is highly noisy and is characterized by correlation, non-guassian due to target glint and modulated by radar cross section fluctuations. Over and above there will be periodic data loss due to eclipsing. For generating smooth homing guidance commands an efficient filter having minimum lag and maximum seeker noise attenuation capability is required. This paper presents an interacting multiple model seeker filter with extended Kalman filter as mode-matched filter which operates in closed homing guidance loop to generate required guidance commands to intercept maneuvering air-breathing targets. The performance of the seeker filter is evaluated using simulation of different interceptor-target engagement scenarios with different target maneuver profiles.

A novel approach to vehicle velocity estimation is proposed. To be cost effective, the estimation is based solely on wheel speed measurement without any additional information on vehicle acceleration. This is made possible by employing an adaptive nonlinear filter scheme which proves to be simple and efficient. Simulation results from field-testing data show that this method results in accurate and smooth vehicle velocity estimation and it is practically feasible.

An important application of Active Noise Control (ANC) is to reduce noise produced by machinery in modern living places. We present an ANC system that is designed to cancel narrowband noise generated from the compressor of an HVAC (Heating, Ventilation, and Air-Conditioning) system in real time. To achieve cost effectiveness of the system, we propose a singlechannel Feedback ANC implementation, which uses only a single speaker and a single microphone. The ANC algorithm is implemented on a NI 7833R Reconfigurable I/O (RIO) FPGA board. A leakage factor is required to ensure stability of the system because of the inherent quantization error in fixed-point implementation of ANC algorithms. To alleviate the effect of quantization error, the fixed-point length is increased from 24 bits to 32 bits, to improve the performance. By correctly choosing the leakage factor, the Leaky Feedback ANC implementation allows for stable reduction of the HVAC compressor noise ranging from 4.59 to 10.73 decibels (dB).

In this paper, a simple and efficient low complexity fast converging partial update normalized LMS (PNLMS) algorithm is proposed for the decision feedback equalization. The proposed implementation is suitable for applications requiring long adaptive equalizers, as is the case in several high-speed wireless communication systems. The proposed algorithm yields good bit error rate performance over a reasonable signal to noise ratio. In each iteration, without reducing the order of the filter, only a part of the filter coefficients are updated so that it reduces the computational complexity and improves speed of operation. The NLMS algorithm can be considered as a special case and slightly improved version of the LMS algorithm which takes into account the variation in the signal level at the filter output and selects a normalized step size parameter which results in a stable as well as fast converging adaptive algorithm. The frequency domain representation facilitates, easier to choose step size with which the proposed algorithm convergent in the mean squared sense, whereas in the time domain it requires the information on the largest eigen value of the correlation matrix of the input sequence. Simulation studies shows that the proposed realization gives better performance compared to existing realizations in terms of convergence rate.

The design and simulation of winner-take –all Current Mode (WTA) circuit is proposed. Inputs and outputs of the Circuits are current and voltage respectively, which makes the circuit appropriated forlow voltage neural hardware computation. The circuit was designed and simulated using Cadence gpdk090. The proposed novel current mode CMOS WTA operates at 0.7v with resolution of 0.001nA and a dynamic range from input current 10nA---1μA, which are much better than existing realizations. It is also observed that dynamic range can be raised further with increasing of power supply. Simulation results along withappropriate mathematical relations are reported.

Audio sur\eillance tapes are prime candidates for spccch enhancement hecause of the Inany degradations and sources of interference that mask the spccch signals on such tapes. I n this paper. the authors describe ways to cancel intcrferencc \+ hen an a\.ail:itdc reference hignal is not s\nchronized with the sur\~eillance recording. for example. \+hen the reference is obtained later frcm a phonofraph record or an air check recording from a broadcast wurce. As a qwcific esaniple. we discuss our experiences processing a \virctap recording used in an actu:rl court case. We transformed the rcfcrence 5ign;rl to reflect room and transmission effect> and then wtltracted the resulting sccondar! signal from the primary intercept signal. thus enhancing the speech of the dc>ircd t;rlLcrs hy remo\ing interfering sounds. Before the secondary signal could he subtracted. the signals had to he aligned properly in time. The intercept signal \\.as subjected to time-scale modificirtiims m;ds necestry t-ry the virr!ing phonograph and tape recorder kpeeds. While thcse spced dit1crenct.s are usually small enough not to affcct the perceived quality. they acfvcrwl\ ;rffcct the abilit! to cancel interference automaticall!.. In working \vith recording devices. \\c took into accoirnt four f;rcrors that affcct the signal quality: the frequency rcsponw. nonlinc;rr Jiitortion. noiic. and speed v;rri;rtions. 7-he t\\o methods that \\.ere most succes>ful for cn1l;tnccmcnt were the Ie:rst-nic~in-squares ( L M S ) ;rdapti\.e cancellation and spectr;rl whtracti~m.

This letter studies the tracking performance of a stochastic gradient-based adaptive algorithm, namely the maximum correntropy criterion algorithm, where a random walk is used to model the non-stationarity. In our analysis, we use the energy conservation argument to derive expressions for the steady-state excess mean square error (EMSE). We consider two different cases for measurement of noise distribution including the Gaussian noise and general non-Gaussian noise. For the Gaussian case, we derive a fixed-point equation that can be solved numerically to find steady-state EMSE value. For the general non-Gaussian case, we derive an approximate closedform expression for EMSE. For both cases, unlike the stationary environment, the EMSE curves are not increasing functions of step size parameter. We use this observation to find the optimum step size learning parameter for general non-Gaussian case. The validity of the theoretical results are justified via simulation results.

Adaptive ¯lters have wide range of applications in areas such as echo or interference cancellation, prediction and system identi¯cation. Due to high computational complexity of adaptive ¯lters, their hardware implementation is not an easy task. However, it becomes essential in many cases where real-time execution is needed. This paper presents the design and hardware implementation of a variable step size 40 order adaptive ¯lter for de-noising acoustic signals. To ensure an area e±cient implementation, a novel structure is being proposed. The proposed structure eliminates the requirement of extra registers for storage of delayed inputs thereby reducing the silicon area. The structure is compared with direct-form and transposed-form structures by adapting the ¯lter coe±cients using four di®erent variants of the least means square (LMS) algorithm. Subsequently, the ¯lters are implemented on three di®erent ¯eld programmable gate arrays (FPGAs) viz. Spartan 6, Virtex 6 and Virtex 7 to ¯nd out the best device family that can be used to implement an Adaptive noise canceller (ANC) by comparing speed, power and area utilization. The synthesis results clearly reveal that ANC designed using the proposed structure has resulted in a reduction in silicon area without incurring any sig-ni¯cant overhead in terms of power or delay.

In the literature, Query-adaptive multiple instance learning for video instance retrievalObject-based image retrieval, in which a user is only interested in some object in the images. While in this paper, by considering both region level and image information at the same time, a novel method based on multigraph multiinstance learning algorithm method using SURF features is proposed. Two graphs are constructed in this method, and the relationship between each image and its segmented regions are introduced in order to get optimized framework. Moreover, this method is further extended to video retrieval. By enhancing the relationships between all the video shots, images, and segmented regions, such that it deals with the case when train1 International Journal of Pure and Applied Mathematics Volume 118 No. 17 2018, 523-536 ISSN: 1311-8080 (printed version); ISSN: 1314-3395 (on-line version) url: http://www.ijpam.eu Special Issue ijpam.eu

Instead of using single kernel, different approaches of using multiple kernels have been proposed recently in kernel learning literature, one of which is multiple kernel learning (MKL). In this paper, we propose an alternative to MKL in order to select the appropriate kernel given a pool of predefined kernels, for a family of online kernel filters called kernel adaptive filters (KAF). The need for an alternative is that, in a sequential learning method where the hypothesis is updated at every incoming sample, MKL would provide a new kernel, and thus a new hypothesis in the new reproducing kernel Hilbert space (RKHS) associated with the kernel. This does not fit well in the KAF framework, as learning a hypothesis in a fixed RKHS is the core of the KAF algorithms. Hence, we introduce an adaptive learning method to address the kernel selection problem for the KAF, based on competitive mixture of models. We propose mixture kernel least mean square (MxKLMS) adaptive filtering algorithm, where the kernel least mean square (KLMS) filters learned with different kernels, act in parallel at each input instance and are competitively combined such that the filter with the best kernel is an expert for each input regime. The competition among these experts is created by using a performance based gating, that chooses the appropriate expert locally. Therefore, the individual filter parameters as well as the weights for combination of these filters are learned simultaneously in an online fashion. The results obtained suggest that the model not only selects the best kernel, but also significantly improves the prediction accuracy.

This manusaipt has been reproduced from the mtcrofdm master. UMI films the text directly from the ClIigiMI or ClOp)' submitted. ThJs, some thesis and dissertation copies ate in typewriter face, whir. others may be from any type of computer printsr. The quality of this reproduction is dependent upon the quality of the copy submitted. 8mken« indistinct print, CDIored or poor quality illustrations and photographs, print ~rough, substandard margns. and improper alignment can adversely affect reproduction. In the unlikely event that the aIthor did not send UMI a complete manusaipt and there are missing pages, these 'Nill be noted. Also, if lXlauthorized copyright material had to be removed, a nota will indicate !he deletion. Oversize materials (e.g., maps. drawings. charts) are reproduced by sectioning the original, beginning at the upper M!ft-Nncl corner and continuing from left to right in equal sections wfth small over1aps. Photographs induded in the original manusctipt have been reproduced xerographically in this copy. Higher quality e" x 9" black and white photographic prints are avaita~for.-,y photographs or ilustrations appearing WI this copy for an additional charge. Contact uur di'ectIy to OI'der.

fMRI is used to investigate brain functional connectivity after removing nonneural components by General Linear Model (GLM) approach with a reference ventriclederived signal as covariate. Ventricle signals are related to lowfrequency modulations of cardiac and respiratory rhythms, which are nonstationary activities. Herein, we employed an adaptive filtering approach to improve removing physiological noise from BOLD signals. Comparisons between filtering approaches were performed by evaluating the amount of removed signal variance and the connectivity between homologous contralateral regions of interest (ROIs). The global connectivity between ROIs was estimated with a generalized correlation named RV coefficient. The mean ROI decrease of variance was 52% and 11%, for adaptive filtering and GLM, respectively. Adaptive filtering led to higher connectivity between grey matter ROIs than that obtained with GLM. Thus, adaptive filtering is a feasible method for removing the physiological noise in the low frequency band and to highlight resting state functional networks.

Smart grid (SG) is widely considered as the next generation power grid. Recently, wireless sensor networks (WSNs) have shown its advantages for SG by achieving low-power transmissions. However, SG communication technologies are subjects to ineffective spectrum allocation issues. Cognitive radio wireless sensor network (CRWSN) has been identified as a new technique for SG to face these challenges. CRWSN is a new research paradigm where WSN and cognitive radio are combined. In this paper, WSN with cluster tree topology is proposed as a compatible choice for SG design. We introduce CRWSN into SG using cluster tree topology to improve the sensing performance and solve the rarity of spectrum by opening the underused licensed bands to secondary users. We adopt cluster distributed cooperative sensing as a method of spectrum sensing. We later evaluate the performance of this method under two distributed policies; OR and AND rules based on energy detection.

In this paper, we present a new spam filter which acts as an additional layer in the spam filtering process. This filter is based on what we call a representative vocabulary. Spam e-mails are divided into categories in which each category is represented by a set of tokens which form a Representative Text (RT). Tokens are strings of characters (words, sentences, or some times meaningless strings of characters). This RT is used to compute a resemblance ratio with incoming e-mails. With this ratio we decide whether the incoming e-mail is a spam. This filter was implemented and integrated to Spamihilator software. Some experimental and interesting results will be presented.