Mucahit K. Uner

Comparison of covariance estimators for nongaussian multipulse detectors

In recent years, multi pulse nongaussian signal model has been one of the most frequently used signal model on radar detection. A suboptimum receiver for that model has been already developed. In order to realize that detector the correlation properties of the underlying clutter have to be estimated. Although there are several estimators proposed for that, they are not widely used. One of the main reasons of this is that there is not enough comparison analysis between these estimators. So the suitable estimator for selected application is not known. In this paper these correlation estimators are compared under complex clutter scenarios.

Spectrum based covariance estimation for nongaussian multipulse detectors

Multipulse detection under nongaussian clutter is addressed. GLRT based detectors have already been developed for this problem. These detectors require the correlation estimation of the local environment for practical use. Some ML based estimators have been developed for this purpose. Unfortunately the exact ML estimate does not exist for the compound gaussian clutter model considered in this study. Only approximations exist. In this work this problem is handled with a different approach. Correlation estimation problem is reformulated as a local doppler spectrum estimation problem. After the Doppler spectrum estimated, correlation estimate is obtained from spectrum-correlation relation. Proposed estimators are compared with the ML based estimators and it is shown that the performances of the proposed estimators are equivalent or superior to the ML based estimators in many conditions.

Estimation of radar threshold in environments with unknown distribution

Radar threshold is a parameter that affects radar performance directly by causing a trade off between detection and false alarm probability. While designing radar systems, it must be set accurately in order to have reliable decisions about target detection. In this study, a radar threshold estimation method based on Generalized Pareto Distribution (GPD) is presented. The performance of this method for environments with different distributions is analyzed through simulations.

Performance of cell averaging constant false alarm rate (CA-CFAR) processors for M-correlated sweeps in multiple target environments

In this paper, the detection performance of CA-CFAR processor is analysed for M-correlated sweeps in Gaussian clutter environment which contains interfering targets. The performance analyses are done for different number of sweeps, M, and sweep-to-sweep correlation coefficient, R, with Swerling-I target model in homogeneous and non-homogeneous environments.

CFAR detection based on decision fusion of order statistic CFAR and order statistic clutter map CFAR processors

In automatic target detection radars, adaptive constant false alarm rate (CFAR) processors, which determine the threshold, are examined. Two techniques are used for achieving the required CFAR. The first technique determines the adaptive threshold using range neighbouring cells which are located in sliding window. The second one is based on clutter map. In this technique, adaptive threshold is determined using the samples of test cell from previous scans which are assumed to contain no target. In this study, new techniques based on combining the two techniques in the data fusion system (DFS), are proposed. Analytical equations for homogeneous and nonhomogeneous environment are derived and performance analysis are carried out.

A stop criteria for genetic algorithms applied to distributed CA-CFAR radar systems

Search space is an infinite set in decimal genetic algorithms (GA). In optimization problems where decimal genetic algorithms are used, the stop criteria such as the maximum number of iterations or cost function having no more difference than a specified value may not guarantee that the computed solution is equal to the real solution. The proposed stop criteria has been applied to the problem of distributed CA-CFAR radar systems and it has been observed that the computed solution is equal to the real solution.

Adaptive Censored CFAR Performance in Nonhomogeneous Correlated Clutter

In this work, performance analysis of CFAR radar processors in nonhomogenous partially correlated Gaussian background is studied using binary integration detectors. In nonhomogeneous background, a new algorithm is proposed to find the homogeneous region that the test cell is located. CFAR analysis is performed in this region and the probability of false alarm results are compared with the design value.

Turbo Decoding on the TMS320C6416 Digital Signal Processor

Channel coding plays a crucial role in modern communication systems. Forward error correcting (FEC) channel codes provide improvements in the bit error rate performance by mitigating channel perturbations. Turbo codes approach the Shannon limit closer than other FEC codes. In this paper, the error performance of a turbo coded system has been analyzed on a digital signal processor (DSP). By using a 720 MHz spectrum digital TMS320C6416 DSP starter kit (DSK), the error performance curves and coding gains of a 3GPP turbo coded system are obtained in an additive white Gaussian noise (AWGN) channel for different interleaver sizes. In the receiver, the on-chip turbo decoder co-processor (TCP) of the DSP has been employed for decoding

Image fusion for computed tomography and gamma camera

Medical imaging systems are widely used in diagnosis due to their relatively higher technology and informative output. All imaging systems have their advantages with respect to other imaging systems. However, an imaging system can yield single format images, either anatomical or physiological. Image fusion is mostly useful in this situation where combined information is requested. The image fusion is used in a lot of areas as an operation that represents more information in one fused image than individual images separately. The method, that is used to get more efficient images using image fusion form different image sources, is used in factory automation, satellite images, military and civilian objects, as well as medical areas. Popular medical imaging techniques CT (computed tomography) and gamma camera have high performance in their respective areas. Image fusion using images from both systems gathers important aspects of each image and helps the doctors to get more efficient and faster diagnoses. In this work, we aim to fuse two images obtained from CT and Gamma camera systems operating in Hacettepe University, Departments of Radiology and Nuclear Medicine. Three images are obtained from each modality, which belong to three different slices of the same subject. From these three image pairs, fused images are obtained. The fused results were approved by the radiologists. In the future, a graphical user interface is planned to be developed.

Wavelet based image compression

In this work, the embedded zerotree wavelet (EZW) coding and the set partitioning in hierarchical trees (SPIHT) coding methods are studied. The effect of the arithmetic coding on the performance of the EZW and SPIHT coding methods is analysed. The performance of these two coding methods are compared with the well-known Joint Photographic Experts Group (JPEG) compression methods performance.