Digital Signal Processing

Russian
Scientific & Technical
Journal


“Digital Signal Processing” No. 2-2013

15 International Scientific Research Conference
"Digital Signal Processing and its Application - DSPA-2013"

In the issue:

- discrete transformations;
- digital filtering;
- wavelet techniques;
- analysis of scalegramms;
- algorithm of finding maximal singular value;
- demodulation of radio signals;
- modulation recognition.



T
HE ONE-DIMENSIONAL VECTOR STATIONARY LINEAR DIGITAL FILTERS
Gadzikovsky V.I., gadzikovsky_vi@mail.ru


The methods of description and block diagrams of the one-dimensional vector digital filters including the generalized linear noise model of L. Jackson are considered. The evaluation of the noise influence and the calculation of required number of digits of digital signal processors with a fixed point are carried out.

Keywords: one-dimensional vector digital filter, fixed-point processor, digital filter.

METHOD FOR REAL SEQUENCE DISCRETE FOURIER TRANSFORM FAST EVALUATION
Ponomareva O.V., Ponomarev A.V., Ponomareva N.V., ponva@mail.ru

The paper presents a method and an algorithm for fast computation of the discrete Fourier transform (DFT), based on the properties of parametric discrete Fourier transform (DFT-D). A comparison of the proposed fast Fourier transform algorithm with existing algorithms is presented. The efficiency of the proposed approach is considered.

Keywords: discrete Fourier transform, parametric discrete Fourier transform, fast Fourier transform, actual sequence, even / odd symmetry of spectrum, complex conjugate symmetry.

OPTIMIZATION OF SHORT-TIME FOURIER AND CONTINUOUS WAVELET TRANSFORMS FOR SPECTRAL ANALYSIS OF MUSICAL SIGNALS
Aliyev R.M., RomanAliyev@gmail.com

The method to improve time-frequency accuracy during spectral analysis of musical signals is offered. The method optimizes the parameters of spectral transforms and includes three stages - spectral modeling of sound samples, synthesis of music spectrums, and comparison of the synthesized and computed spectrums. Experiments with classical music databases presented the optimum window size in short-time Fourier transform and the optimum wavelet compactness in continuous wavelet transform are explored.

Keywords: time-frequency analysis, short-time Fourier transform, wavelet transform, musical signal.

THE ENCEPHALOGRAMS ANALYSIS BASED ON CHAINS RESEARCH OF SCALEGRAMMS LOCAL MAXIMUMS
Turovsky Ya.A., Kurgalin S.D., Semenov A.G., yaroslav_turovsk@mail.ru


It is shown that a new type of observable information can be extracted from electroencephalograms (EEGs) by analysis of chains structure of scalegramms local maximums which are obtained by use of continuous wavelet transform. High efficiency of this method as a tool for the study of brain activity is demonstrated.

Keywords:
wavelet-analysis, biomedical information, electroencephalogram.

THE DYNAMIC OF POWER INDEXES OF LOCAL MAXIMUMS CHAINS OF THE WAVELET-COEFFICIENTS FOR BIOMEDICAL SIGNALS
Turovsky Ya.A., Kurgalin S.D., Semenov A.G., yaroslav_turovsk@mail.ru

The method of local maxima chains (LMCs) determination of the continuous wavelet transform coefficients for different types of signals and analysis of their characteristics with use of Morlet wavelet function as basic function is presented. Various ways of LMCs detection and approaches to research of their properties are demonstrated, advantages and lacks are noted. The dynamics typologization of the energetic values of local spectra maxima forming LMCs is carried out and its results are presented. The energy averaging algorithms of LMCs are proposed. Three ways of these averages are presented. Five types of LMCs energy maxima dynamic behavior both for model signals and for real EEGs is revealed.

Keywords: wavelet-analysis, electroencephalogram, local spectrum, chain of local maxima.

USING THE SPARSE APPROXIMATION AND CLUSTERING METHODS FOR THE TIME SERIES STRUCTURE DESCRIPTION OF ACOUSTIC EMISSION
Afanaseva A.A., Lukovenkova O.O., Marapulets Yu.V., Tristanov A.B., marpl@ikir.ru, alextristanov@mail.ru


The sparse approximation method with novel dictionary has been used for time-frequency analysis of acoustic emission. The proposed dictionary is based on Berlage's functions. The algorithm for the acoustic impulses detection, based on the methods of sparse approximation and clustering is suggested.

Keywords: sparse approximation, time-frequency analysis, acoustic emission, clustering.

EFFICIENT ALGORITHM FOR FINDING THE MAXIMUM SINGULAR VALUE FOR SOLVING THE PROBLEM OF RANDOM SIGNAL DETECTION
Ratynsky M.V., Petrov S.V., m3v5r7@inbox.ru, petrovsv@list.ru

An algorithm for finding the maximum singular value of a rectangular complex matrix for solving the problem of random signal detection from an external source against the background of the intrinsic noise of the antenna-receiver system is proposed. The algorithm is based on the use of the iterative power method applied to the matrix of input signals samples of detector received from the outputs of the antenna array elements. This eliminates the step of forming the sample estimate of the input signals correlation matrix, that allows obtaining a significant gain in the amount of computation that under certain task settings can reach 5 to 10 times and more.

Keywords:
detection of a random signal, the maximum singular value, the power method, the rectangular complex matrix.

DEMODULATION OF V-BLAST SPATIALLY MULTIPLEXED RADIO SIGNALS WITH MULTIPLE MANIPULATIONS
Mukhin I.A., Nemirovsky M.S.,
ilyamukhin@gmail.com

The energy efficiency comparison of different methods of collective demodulation in V-BLAST networks with use of multiple manipulations depending on the characteristics of radio waves propagation paths is produced. All paths are classified on the basis of the real paths model offered in authors' work [1]. Data on the effectiveness of collective demodulation various methods (including relatively little-studied relaxation methods) obtained for each path class by simulation in MATLAB. Basically discusses options with QAM. A certain place is given to options with multiple PSK, because, as shown in this article, the difference in power of used radio signals negatively affects the efficiency of demodulation in V-BLAST networks.

Keywords: MIMO, V-BLAST, multiple antenna systems, spatial multiplexing, demodulation.

MODULATION RECOGNITION: WHAT CAN WE FIND OUT NOT KNOWING ANYTHING A PRIORI
Beljaeva M.B. maria.beljaeva29@gmail.com


Modulation recognition problem is considered in the case of greatest possible uncertainty: no signal characteristics are known a priori, no noise distribution and signal-to-noise ratio are known also. The only assumption made is the symbol rate interval which may be as wide as desired. Vertical boxes algorithm (VBA) for non-cooperative automatic discrimination between digital signals with amplitude information (AI) from those without AI is presented. VBA also distinguishes between some other classes of signals. Some numerical simulations illustrate the VBA effectiveness.

Keywords: modulation, modulation recognition, vertical boxes algorithm.

"DIGITAL SIGNAL PROCESSING" IN THE NEW SYSTEM OF HIGHER PROFESSIONAL EDUCATION
Alla I. Solonina,
as-io@yandex.ru

Article describes basic and more fundamental education in Digital Signal Processing (DSP), its place inside the architecture of the contemporary system of higher professional education and inside university architecture, as well as its content and methods of learning based on computer modeling.

Keywords: methods of teaching, digital signal processing, system of higher professional education.

HARMONIZATION OF THE RUSSIAN STANDARDS OF THE HI-SPEED INTERFACES
D.A. A
minev., A.V. Kondrashov

Issue of design of hi-speed information system in Russia is studied. Classification of hi-speed interfaces is offered. Righthold and design organizations of standards is reviewed. Íarmonization prepositions are formulated.

Keywords: hi-speed interfaces, standards, harmonization of standards.


If you have any question please write: info@dspa.ru