ICASSP '98 Main Page General Information
Conference Schedule
Technical Program
Overview
50th Annivary Events
Plenary Sessions
Special Sessions
Tutorials
Technical Sessions
By Date |
||
| May 12, Tue | ||
| May 13, Wed | ||
| May 14, Thur | ||
| May 15, Fri | ||
By Category |
|||
| AE | ANNIV | ||
| COMM | DSP | ||
| IMDSP | MMSP | ||
| NNSP | PLEN | ||
| SP | SPEC | ||
| SSAP | UA | ||
| VLSI | |||
By Author |
||||||
| A | B | C | D | E | ||
| F | G | H | I | J | ||
| K | L | M | N | O | ||
| P | Q | R | S | T | ||
| U | V | W | X | Y | ||
| Z |
||||||
Registration
Exhibits
Social Events
Coming to Seattle
Satellite Events
Call for Papers/
Author's Kit
Future Conferences
Help
Abstract - SSAP8 |
 |
SSAP8.1 |
Optimum Subarray Configurations Using Genetic Algorithms J. Wang, H. Israelsson, R. North (Center for Monitoring Research, SAIC, USA) Subarray configuration is not a trivial problem in array signal processing. A proper subarray configuration is important to improve the detectability of an array. A new searching algorithm, which is based on Genetic Algorithms (GA), for the optimum subarray configuration is proposed in this paper. Our preliminary application to a seismic array has indicated that the new algorithm can search a population of subarrays in a more efficient and robust way. The beamforming gain of the optimum subarray derived by GA is very close to the theoretical gain. Experimental results on signal detections have demonstrated that a beamforming recipe with optimum subarrays can provide further enhanced signal-to-noise ratio (SNR), compared to a recipe without subarray configuration. The approach proposed here can be easily extended to the weight determination problem for the weighted beamforming process by using multi-bit instead of 1-bit representation for each sensor in the chromosome model. |
| SSAP8.2 |
Bayesian Estimation of Abrupt Changes Contaminated by Multiplicative Noise Using MCMC J. Tourneret, M. Doisy, M. Mazzei (ENSEEIHT, France) The paper addresses the estimation of abrupt changes which are contaminated by multiplicative Gaussian noise. The marginal mean a posteriori or marginal maximum a posteriori estimators can be derived for estimating the position of a single abrupt change. However, these estimators have optimization or integration problems for multiple abrupt changes. The paper solves these optimization problems by using Markov Chain Monte Carlo methods. |
| SSAP8.3 |
Locally Optimum Detectors for Deterministic Signals in Multiplicative Noise M. Ghogho, A. Nandi (Strathclyde University, Scotland, UK); B. Garel (INP-ENSEEIHT, France) This paper addresses the problem of detecting deterministic signals in a multiplicative noise model. The multiplicative noise model is appropriate for modelling coherent imaging systems such as SAR and LASER. Lodally Optimum (LO) detectors are derived for any arbitrary multiplicative noise distribution. The gamma and generalized Gaussian distributions are studied in detail. We also introduce an extension of the generalized Gaussian density to include asymmetry. The performance of the LO detectors is studied and compared with that of the linear correlation detector. The paper gives insight into the influence of the tail-length of the noise distribution on the detection performance. |
| SSAP8.4 |
Evaluation of CFAR and Texture Based Target Detection Statistics on SAR Imagery L. Kaplan, R. Murenzi (Clark Atlanta University, USA) In this work, we evaluated the effectiveness of synthetic aperture radar (SAR) target detection algorithms that consist of any number of combinations of three statistics which include two-parameter CFAR, variance, and extended fractal features. The performance of these algorithms were tested at various threshold settings over the public domain MSTAR database. This database contains one foot resolution X-band SAR imagery. Receiver-operating-characteristic (ROC) curves were generated for the seven resulting algorithms. The results indicate that the CFAR statistic is the least effective detection statistic. |
| SSAP8.5 |
On the Use of a General Amplitude PDF in Coherent Detectors of Signals in Spherically Invariant Interference D. Iskander (Queensland University of Technology - Signal Processing Research Center, Australia) The aspects of using a general amplitude probability density function in coherent detectors are investigated. For this purpose, the recently developed Generalised Bessel function K (GBK) distribution is used. The performance of the optimal detector of signals embedded in GBK-distributed interference is compared to the one of the uniformly most powerful invariant detector using extensive Monte Carlo simulations. The results indicate that for small number of integrated pulses the optimal detector outperforms the uniformly most powerful invariant detector by up to 18 dB. It is shown, that this improvement does not vary significantly with changes in the parameters that control the spherical invariance of the GBK distribution. |
| SSAP8.6 |
Statistical Classification of Chaotic Signals C. Couvreur, C. Flamme, M. Pirlot (Faculte Polytechnique de Mons, Belgium) The classification of chaotic signals generated by a low-dimensional deterministic models given a dictionary of possible model is considered. The proposed classification methods rely on the concept of "best predictor" of signal. A statistical interpretation of this concept based on the ergodic theory of chaotic system is presented. A sort of "bootstrapping" estimator of the statistical properties is introduced. The method is validated by numerical simulations. Directions for future research are suggested. |
| SSAP8.7 |
Detection of Spectrally Equivalent Parametric Processes Using Higher Order Statistics M. Coulon, J. Tourneret (ENSEEIHT/GAPSE, France); A. Swami (Army Research Lab, USA) The paper addresses the problem of detecting two spectrally equivalent parametric processes (SEP's) : the noisy AR process and the ARMA process. Higher order statistics (HOS) are shown to be effective for detection. Two HOS based detectors are derived and compared. The first detector studies the singularity of a HOS-based Yule-Walker matrix. The second detector filters the data by an AR filter estimated from the data; the residual HOS are then shown to be effective for the SEPP detection problem. |
| SSAP8.8 |
Multiple Hypothesis Modulation Classification Based on Cyclic Cumulants of Different Orders P. Marchand (CEPHAG - ENSIEG, Domaine Universitaire, France); C. Le Martret (CELAR, France); J.-L. Lacoume (CEPHAG - ENSIEG, France) A multiple hypothesis modulation QAM classification task is adressed in this paper. The classifier is designed within the rigorous framework of decision theory. A characteristic feature is extracted from the signal, and is compared to the possible theoretical features in the maximum likelihood sense. This feature is composed of a combination between fourth-order and squared second-order cyclic temporal cumulants. No assumption about the power of the signal is made. It is shown that this uncertainty about the power of the signal does not affect the decision rule. As an application, we present simulated performance in the context of 4-QAM {\it vs} 16-QAM {\it vs} 64-QAM classification. |
| SSAP8.9 |
Optimal Sensor Scheduling for Hidden Markov Models J. Evans, V. Krishnamurthy (University of Melbourne, Australia) Consider the Hidden Markov model where the realization of a single Markov chain is observed by a number of noisy sensors. The sensor scheduling problem for the resulting Hidden Markov model is as follows: Design an optimal algorithm for selecting at each time instant, one measurement provided by one out of the many sensors. Each measurement has an associated measurement cost. The problem is to select an optimal measurement scheduling policy, so as to minimize a cost function of estimation errors and measurement costs. The problem of determining the optimal measurement policy is solved via stochastic dynamic programming. Numerical results are presented. |
| SSAP8.10 |
A General Maximum Likelihood Framework for Modulation Classification D. Boiteau (Centre d'Etudes de Systemes et Techniques Avancees, France); C. Le Martret (Centre d'Electronique de l'Armement, France) This paper deals with modulation classification. First, a state-of-the-art is given which is separated in two classes: the pattern recognition approach and the Maximum Likelihood (ML) approach. The we propose a new classifier called the General Maximum Likelihood Classifier (GMLC) based on an approximation of the likelihood function. We derive equations of this classifier in the case of linear modulation classification and apply them to the M PSK / M' PSK problem. We show that the new tests are a generalization of previous ones using the ML approach, and don't need any restriction on the baseband pulse. Moreover, the GMLC provides a theoretical foundation for many empirical classification systems including those systems that exploit cyclostationary property of modulated signals. |
| < Previous Abstract - SSAP7 | SSAP9 - Next Abstract > |