Search

Since it is practically difficult to generate and propagate an impulse, often a system is excited by a narrow time domain pulse. The output is recorded and then a numerical deconvolution is often done to extract the impulse response of the object. Classically, the fast Fourier transform technique has been applied with much success to the above deconvolution problem. However, when the signal to noise ratio becomes small, sometimes one encounters instability with the FFT approach. In this paper, the method of conjugate gradient is applied to the deconvolution problem entirely in the time domain. The method converges for any initial guess in a finite number of steps. Also for the application of the conjugate gradient method the time samples need not be uniform like FFT. Computed impulse response utilizing this technique has been presented for measured incident and scattered fields from a sphere and a cylinder.

The authors provide necessary and sufficient conditions for bispectral and trispectral factorization for processes with finite-extent cumulant sequences. These conditions are derived entirely in terms of the cumulant sequences. They use the fact that for a finite-extent cumulant sequence factorability is equivalent to finding a finite-order moving-average process with an identical cumulant sequence. In principle the results can be extended to polyspectra of even higher orders. An interesting result of the investigation is that there exist processes generated by nonlinear mechanisms that are factorable. >

This paper demonstrates through numerical simulation the performance of moderate (8x10 3 ) block length Low Density Parity Check (LDPC) codes on a High Frequency (HF) multipath fading channel utilizing Orthogonal Frequency Division Multiplexing (OFDM). Some results are also shown for the Additive White Gaussian Noise (AWGN) and Rayleigh channels. The simulation parameters were chosen to compare results with the standard OFDM 39-tone HF waveform. LDPC codes are linear block codes that have outstanding error correction and detection capabilities.

In this paper we propose an adaptive routing using a fuzzy system. The traffic in the network is re-routed to nodes, which are less congested, or have spare capacity. Based on a set of fuzzy rules, link cost is dynamically assigned depending upon the present condition of the network. Distance vector algorithm, which is one of the shortest path routing algorithms is used to build the routing tables at each node in the network. The proposed fuzzy system determines the link cost given the present congestion situation measured via the delays experienced in the network and the offered load on the network. Delay in the links, was estimated by the time taken for the test packets to travel from the node to its neighbors. The delay information collected by the test packets and the number of packets waiting in the queue, are the two inputs to the fuzzy system. The output of the fuzzy system is the link cost. This algorithm was applied on a simulated NSFNET, the USA backbone, as well as to another test network with a different topology. Robustness and optimality of the proposed fuzzy system was tested by simulating various types of load patterns on these networks. Simulation studies showed that the performance of the fuzzy system was very close to or better than the best performance of the composite metric under different load conditions and topologies.

Discrete-time linear systems that possess scale-invariance properties even in the presence of continuous dilation were proposed by Zhao and Rao. The paper presents results of subsequent investigation characterizing self-similarity properties of discrete-time signals synthesized by these systems. It is shown that white noise inputs to these linear scale invariant systems produce self-similar outputs regardless of the marginal distribution of the noise. We investigate this with different types of inputs and in most instances the outputs are fractional Gaussian and self-similar. This is confirmed by generating the fractional Gaussian noise from the fractional Brownian motion and comparing its characteristics with the system output. For heavy tailed input distributions, the output is also heavy-tailed and self-similar. It is also shown that it is possible to synthesize statistically self-similar signals whose self-similarity parameters are consistent with those observed in network traffic.

It is now empirically documented that data traffic over networks of various types exhibit fractal or self-similar behavior in many instances. Accurate analysis of traffic density and estimation of buffer size must take into account this self-similar nature. Researchers have investigated procedures for generating self-similar signals to model the traffic. Approaches based on the discrete wavelet transform (DWT) are among those that have been proposed. The basis for using the DWT is that it possesses certain scale-invariance properties and scale-invariance provides the foundation for characterizing self-similarity. However, self-similar processes generated with the DWT demonstrates self- invariance to dyadic scaling factors. Zhao and Rao have proposed novel models for purely discrete-time self-similar processes and linear scale-invariant (LSI) systems based on a new interpretation of the discrete-time scaling (equivalently dilation or contraction) operation which is defined through a mapping between discrete and continuous time. They show that it is possible to have continuous scaling factors through this operation although the signal itself is discrete-time. In this paper, we demonstrate application of these LSI systems to the synthesis of data whose self-similarity parameters match those observed in network traffic. Both theoretical development and experimental results are provided.

The problem of separating multiple, independent code division multiple access (CDMA) sources at a single antenna element receiver in the presence of inter symbol interference, co-channel interference and additive white Gaussian noise is addressed. The channels are not known and separation process is required to be blind and adaptive. The system is assumed to consist of multiple users. Only a single user separation is desired. The CDMA channel is assumed to be a multiple-input single-output dispersive channel. The user's codes are assumed to be orthogonal and known to the receiver. The adaptation process is blind and uses the well-known constant modulus (CM) algorithm. The CM approach uses a modulus-corrected version of the output signal in place of the training signal needed in nonblind equalization techniques. The CM algorithm minimizes the deviation of this modulus from a constant in a mean squared error sense. The minimization is done using an LMS-type algorithm.© (1999) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

We propose a new blind equalization technique based on second and fourth order cumulants. The algorithm proposed will equalize the second and fourth order cumulants of the input and output sequences. It is entirely driven by statistics, only requiring knowledge of the variance (power) of the input signal. Because of insensitivity of the higher-order cumulants to Gaussian processes, the algorithm performs well with additive Gaussian noise. Simulation examples are presented in which the proposed technique is compared with other existing equalization algorithms.© (1995) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Differential pulse code modulation (DPCM) is a widely used technique for both lossy and lossless compression of images. In this paper, the effect of using a nonlinear predictor based on artificial neural networks (ANN) for a DPCM encoder is investigated. The ANN predictor uses a 3-layer perceptron model with 3 input nodes, 30 hidden nodes, and 1 output node. The back-propagation learning algorithm is used for the training of the network. Simulation results are presented to compare the performance of the proposed ANN-based nonlinear predictor with that of a global linear predictor as well as an optimized minimum-mean-squared-error (MMSE) linear predictor. Preliminary computer simulations demonstrate that for a typical test image, the zeroth-order entropy of the differential (error) image can be reduced by more than 15% compared to the case where optimum linear predictors are employed. Some future research directions are also discussed.

A two-channel, quadratic, nonlinear process driven by sinusoidal signals is considered. Equations for estimating the parameters of the process are derived from colored Gaussian noise contaminated observations of the process. The derivation is based on the bispectrum and cross-power spectrum of the two channels. The resulting equations are nonlinear and iterative techniques are needed to solve for the parameters. >

We introduce the iteratively clustered interpolation algorithm to compute a structured printer inverse look-up table from irregularly sampled experimental color data. The algorithm is based on a gradient optimization method, with initial points generated through an iterative technique. Experimental results with a digital color printer are provided to illustrate the algorithm.

Previous approaches to the problem of computing scattering by conducting bodies have utilized the well-known marching-on-in-time solution procedures. However, these procedures are very dependent on discretization techniques and sometimes lead to instabilities as the time progresses. Moreover, the accuracy of the solution cannot be verified easily, and usually there is no error estimation. In this paper we describe the conjugate gradient method for solving transient problems. For this method, the time and space discretizations are independent of one another. The method has the advantage of a direct method as the solution is obtained in a finite number of steps and also of an iterative method since the roundoff and truncation errors are limited only to the last stage of iteration. The conjugate gradient method converges for any initial guess; however, a good initial guess may significantly reduce the computation time. Also, explicit error formulas are given for the rate of convergence of this method. Hence any problem may be solved to a prespecified degree of accuracy. The procedure is stable with respect to roundoff and truncation errors and simple to apply. As an example, we apply the method of conjugate gradient to the problem of scattering from a thin conducting wire illuminated by a Gaussian pulse. The results compare well with the marching-on-in-time procedure.

Since it is impossible to generate and propagate an impulse, often a system is excited by a narrow time-domain pulse. The output is recorded and then a numerical deconvolution is often done to extract the impulse response of the object. Classically, the fast Fourier transform (FFT) technique has been applied with much success to the above deconvolution problem. However, when the signal-to-noise ratio becomes small, sometimes one encounters instability with the FFT approach. In this paper, the method of conjugate gradient is applied to the deconvolution problem. The problem is solved entirely in the time domain. The method converges for any initial guess in a finite number of steps. Also, for the application of the conjugate gradient method, the time samples need not be uniform, like FFT. Since, in this case, one is solving the operator equation directly, by passing the autocorrelation matrix computation, storage required is 5N as opposed to N2. Computed impulse response utilizing this technique has been presented for measured incident and scattered fields.

While the bispectrum has been used for the reconstruction of non-minimum phase 1-D signals not much work of a similar nature has been done for multidimensional signals. Here we present a technique for reconstructing 2-D non-minimum phase signals from samples of their bispectra. The reconstruction procedure involves recovering the magnitude of the Fourier transform of the signal from the bispectrum magnitude and its phase from the bispectrum phase. By using the bispectrum the ambiguity regarding phase is considerably removed (up to a linear phase factor) when compared to the spectral factorization approach.

The problem of reconstructing a moving object from multiple snapshots contaminated by noise arises in many imaging applications. Many techniques have been proposed for noise elimination that rely either on measurements of the autocorrelation and/or power spectrum of the observations, or on the assumption that the additive noise is white. The power spectrum is affected by additive noise, and in many situations, the noise is spatially or temporally correlated. The above techniques are sensitive to deviations from assumptions. The bispectrum is identically zero for random processes with symmetric distributions regardless of spatial or temporal correlations. This property along with its ability to retain phase and magnitude information, have led researchers to propose bispectral techniques for estimating parameters of random signals in noise. The bispectrum is also insensitive to translational motion. If these properties are to be taken advantage of to solve the moving object (deterministic signal in noise) reconstruction problem, it is necessary to obtain good estimates of the bispectrum of the object from the noisy observations. In order to do this it is necessary to restrict bispectrum estimation to a certain region of the frequency plane. A consequence of this is that several techniques proposed for bispectral analysis of random signals cannot be used. The paper develops new approaches which enable signal reconstruction from bispectrum measurements made over the restricted region. Simulations of application of these techniques to moving object reconstruction, data transmission over channels with jitter and noise, and image restoration, show that they are more robust with respect to the statistics of the contaminating noise than methods based on autocorrelation.