High SNR Conditions Research Articles

On Maximum-Likelihood Detection for Non-coherent FH/BFSK in Multitone Jamming

We derive the optimum maximum-likelihood (ML) non-coherent detector for FH/BFSK systems in the environment of multitone jamming and additive white Gaussian noise with no aid of jamming state indication and the resultant signal-plus- jammer tone, which, however, were required in K. C. Teh’s previous work [1]. Because of the optimum ML detector’s huge complexity, a novel complexity-reduced scheme which is called joint-suboptimum maximum-likelihood (JSML) detector is proposed. Simulation results show the superiority of the proposed optimum ML and JSML detectors over other existing diversity combining approaches. The performance analysis in terms of bit error probability (BEP) for the JSML detector with single diversity in n = 1 and n = 2 band multitone jamming (BMJ) is provided and further validated by simulation results. In moderate and high SNR conditions, though the proposed JSML detector performs slightly worse than the optimum ML detector, it does achieve much complexity reduction.

Blind signal separation for recognizing overlapped speech.

Utilization of blind signal separation method for the enhancement of speech recognition accuracy under multi-speaker is discussed. The separation method is based upon the ICA (Independent Component Analysis) and has very few assumptions on the mixing process of speech signals. The recognition experiments are performed under various conditions concerned with (1) acoustic environment, (2) interfering speakers and (3) recognition systems. The obtained results under those conditions are summarized as follows.(1) The separation method can improve recognition accuracy more than 20% when the SNR of interfering signal is 0 to6dB, in a soundproof room. In a reverberant room, however, the improvement of the performance is degraded to about 10%.(2) In general, the recognition accuracy deteriorated more as the number of interfering speakers increased and the more improvement by the separation is obtained.(3) There is no significant difference between DTW isolated word discrimination and HMM continuous speech recognition results except for the fact that saturation in improvement is observed in high SNR condition in HMM CSR.

Comparing maximum likelihood estimation and constrained Tikhonov-Miller restoration

The authors have compared the performance of the EM-MLE and ICTM restorations applied to confocal images. Both methods greatly reduce diffraction-induced distortions of confocal images. Due to their nonlinearity, both are able (partially) to restore data of missing frequencies. From the authors' experiments, it is clear that for their test objects, the EM-MLE algorithm performs much better than ICTM. The EM-MLE algorithm produces better results under all the conditions the authors tested, and with respect to all 3 performance measures (I-Divergence, MSE, GDT) the authors used. Only for high SNR conditions, the MSE performance of ICTM approaches the EM-MLE results. It must be noted that this conclusion is only valid for the type of objects the authors used in their experiments (sparse objects); it may well be that for more dense objects, the situation is different. The poor ICTM performance shows that its functional is not well suited for images distorted with Poisson noise. The authors did not find artifacts such as ringing in the results of either algorithm. The restoration results on the cylindrical objects show, however, that the EM-MLE algorithm has a tendency to reconstruct an image that is sharper and smaller than the original object. This aspect of EM-MLE should be investigated thoroughly. Greander's method of Sieves (1991) seems promising for regularizing the EM-MLE algorithm. Finally, to reduce the computational burden of ICTM and EM-MLE, methods to speed up these algorithms should be investigated more fully.

Incorporating a priori information into MUSIC-algorithms and analysis

Constrained MUSIC and beamspace MUSIC are similar algorithms in that they both require a priori information about signal directions and they both involve linear transformations on the data. Constrained MUSIC uses precise information regarding the directions of a subset of the signal directions to improve the direction estimates for the remaining signals. Beamspace MUSIC uses approximate knowledge regarding all the signal directions to reduce computational complexity and improve breakdown properties. These two methods can be combined, resulting in constrained beamspace MUSIC. We also perform asymptotic analysis of constrained and unconstrained MUSIC demonstrating that (asymptotically) improved subspace estimates always result from the use of constraints, and (asymptotically) the variance of constrained MUSIC is less than that of unconstrained MUSIC under either high coherence, large numbers of sensors, or high SNR conditions. As a part of this analysis, we study the effects of coherence on MUSIC and derive best/worst case coherences in terms of the variance of MUSIC. We also demonstrate that those conditions where the variance of MUSIC is predicted to be less than that of constrained MUSIC generally correspond to conditions where MUSIC is in breakdown (and constrained MUSIC is not). So, unconstrained MUSIC actually does not achieve its theoretically predicted advantage in those cases. While constrained MUSIC requires precise information about the known signal to improve performance when the unknown signal is very near, it can also offer performance advantages with only approximate knowledge if the unknown and known signals are not too close to each other.

Gaussian mixture density modeling of non-Gaussian source for autoregressive process

A new approach is taken to model non-Gaussian sources of AR processes using Gaussian mixture densities that are known to be effective for approximating wide varieties of probability distributions. A maximum likelihood estimation algorithm is derived for estimating the AR parameters by solving a generalized normal equation, and a clustering algorithm is used for estimating the parameters of Gaussian mixture density of the source signals. The correlation matrix of the generalized normal equation is not Toeplitz but is symmetric and in general positive definite. Higher order statistics of skewness and kurtosis are used for identifying the source distribution as being Gaussian or non-Gaussian and, consequently, determining the parameter estimation technique between the conventional method and the proposed method. Experiments on non-Gaussian source AR processes demonstrate that under high SNR conditions (SNR/spl ges/20 dB), the proposed algorithm outperforms the conventional AR estimation algorithm and the cumulant-based algorithm by an order-of-magnitude reduction of average estimation errors. The proposed algorithm also has very low estimation errors with short data records. Finally, a maximum likelihood prediction method is formulated for non-Gaussian source AR processes that has shown potential in achieving higher efficiency signal coding than linear predictive coding. >

Comparison of time‐ and frequency‐domain techniques for wind velocity estimation using multiple‐receiver MF radar data

Multiple‐receiver MF radar returns from the mesosphere are used to investigate the relationship between spaced antenna (SA), radar interferometry (RI), and imaging Doppler interferometry (IDI) wind estimation techniques. Our results show that frequency‐domain (RI and IDI) and time‐domain (SA) techniques yield almost identical results under high SNR conditions suitable for SA full correlation analysis (FCA).