Digital Signal Processing

Scientific & Technical

“Digital Signal Processing” No. 2-2012

14 International Scientific Research Conference
“Digital Signal Processing and its Application - DSPA-2012”

In the issue:

- discrete transformations;
- wavelet techniques;
- voice conversion;
- multirate processing;
- signal detection and reception;
- adaptive filtering;
- additive sonar system.

O.V. Ponomareva
Fast Parametric Discrete Fourie Transformation of Real Sequences

Using standard complex FFT-P with ?=1/2 to real x(n), n=0,N 1 sequences on the one hand results in calculating N/2 spectrum points at negative frequencies with no additional information for x(n) signal spectrum (excessive computation). On the other hand a desired memory capacity is twice as much as the required one (excessive memory costs). To eliminate the above drawbacks FFT-P algorithms of real sequences are proposed thereby including both memory and calculation costs. FFT-P algorithms of real sequences with ?=1/2 is also proposed.
O.V. Ponomareva, e-mail:

N.I. Oreshko, V.V. Geppener, D.M. Klionsky
Using Harmonic Wavelets in Oscillating Signal Processing

The paper deals with using harmonic wavelets with reference to a wide class of oscillating signals. New approaches to solve the problems of determining the boundaries of typical processes in a signal and denoising the signals by harmonic wavelets based on a modified thresholding are considered. General matrix view for a harmonic wavelet transformation is deduced. Special criterion for estimating the effectiveness of the given wavelets are introduced.
N.I. Oreshko, e-mail:
V.V. Geppener, e-mail:
D.M. Klionsky, e-mail:

I.S. Azarov, A.A. Petrovsky
Real-Time Voice Conversion System with Text-Independent Training Based on Hybrid Parametric Speech Representation

The paper presents a voice conversion technique based on a hybrid parametric speech representation. The conversation function is estimated by using speech samples of source and target speakers with an arbitrary text content. The effectiveness of a proposed technique is estimated by experimental implementation of voice conversion system.
I.S. Azarov, e-mail:
A.A. Petrovsky, e-mail:

O.V. Mandrikova, I.S. Solovyev
Wavelet-Based Approach for Geomagnetic Data Processing and Analysis

The paper deals with creating automated systems to analyze geomagnetic data, determine and identify local features, caused by solar activity. The authors propose wavelet-based algorithms to reveal in automatic mode local structures forming a recorded geomagnetic signal to analyze and identify these structures as well. The proposed methods and approaches are approved by the Earth’s magnetic field data obtained in Paratunka observatory (village Paratunka, Kamchatka region).
O.V. Mandrikova, e-mail:
I.S. Solovyev, e-mail:

O.V. Mandrikova, Yu.A. Polozov
Method of Allocating Abnormal Features in Critical Frequency Data of the Ionosphere based on Combination of Wavelet Transformations and Neural Works

A method of automatic allocating the abnormal features occurring in a recorded time series of ?oF2 critical frequency of the ionosphere during abnormal seismic or solar activity is proposed. The method is based on a combined use of wavelet transformation and neural networks. Wavelet-based algorithms for allocating the abnormal features and estimating their parameters are developed. Determining and analyzing the specific time series components are carried out by a combined use of wavelet transformation and neural networks. Method approbation is performed with ?oF2 data obtained in observatory “Paratunka” (village Paratunka, Kamchatka region).
O.V. Mandrikova, e-mail:
Yu.A. Polozov, e-mail:

A.A. Mikheev, P.A. Blinov, N.S. Tochilina

Multirate Processing in Electrocardiosignal Filtering

One of the effective methods for eliminating additive noise in electrocardiosignal is considered. Multirate electrocardiosignal processing is proposed to be used. Decimation variants of electrocardiosignal sampling frequency and pulse filter performance are presented.
N.S. Tochilina, e-mail:

A.M. Paramonov
Receiving CQPSK signals by Frequency Discriminator

Spectral characteristics of CQPSK signals are considered. Methods for forming and demodulating of such signals are described. Digital realization of incoherent CQPSK signal demodulator using frequency discriminator is proposed. BER-SNR relations for this demodulator are estimated by a computer simulation.
A.M. Paramonov, e-mail:

A.M. Paramonov

Two-channel Crosspolarization Interference Canceller

One of the available block diagrams for a digital two-channel crosspolarization interference canceller (XPIC) is given. Mathematical model of radio system with orthogonal polarization and two-channel reception is described. The effectiveness of cancelling crosspolarization interference while receiving QAM signals is estimated by a computer simulation.
A.M. Paramonov, e-mail:

A.A. Gorshenkov, Yu.N. Klikushin, V.Yu. Kobenko
Algorithm for Detecting, Measuring and Classifying Signal Trends

An algorithm for solving the problems of detecting, measuring and classifying signal trends is proposed. The algorithm is based on calculating the relative distance between sample implementation of the signal tested and two reference sequences formed out of the original signal by its forward and reverse sorting. Reference sequences set left and right boundaries for a trend measurement range of a signal being tested.
Yu.N. Klikushin, e-mail:
V.Yu. Kobenko, e-mail:

V.I. Koshelev, V.A. Belokurov

Synthesis and Analysis of Adaptive Multiframe Algorithm for Detecting a Small Target

Algorithm for detecting a radar signal with eliminating range ambiguity based on its multiframe accumulation and adaptive amplitude estimation is synthesized and analyzed. Hardware requirements to implement this algorithm are determined.
V.I. Koshelev, e-mail:
V.A. Belokurov, e-mail:

A.I. Demidov, N.N. Zalogin, R.Sh. Komochkov, S.S. Mosolov, A.V. Sknarya, S.A. Toschov, E.V. Tutynin
Active Sonar System with Ultra Wide Band Probing Signals

Some problems of realizing wide-band technique in native sonar by collaborative works of RAS IRE named after V.A. Kotelnikov and Public Company “NIIP”.

V.A. Mesterhazy
Program Integer-Valued Integration of ADC Signals

Problems of integrating digital signals in fast-acting real-time microprocessor control systems are solved. Algorithms and program implementation in assembly language are presented. The examples of various integration methods realized in microprocessor systems being a success in energy complexes of our country are given.
V.A. Mesterhazy, e-mail:,

V.G. Bartenev, A.Yu. Gordeev
New Method of Forming Weights for Adaptive Filtering of Unclassified Observation Samples

Techniques of adaptive filtering for unclassified observation samples based on weight cross control while forming these weights by Prony and Steiglitz-McBride methods are considered. The results of unclassified observation sample processing with a useful signal and two-component correlated interference by using the above mentioned approaches to form filter weights are given. Processing results of the same unclassified observation samples but using autoregressive approach to form weights are presented as well.
V.G. Bartenev, e-mail:
A.Yu. Gordeev, e-mail:

If you have any question please write: