Digital Signal Processing

Scientific & Technical

“Digital Signal Processing” No. 4-2012

In the issue:

- detection of tonal components;
- correlation of spectral coefficients;
- coherent accumulation with decimation;
- synchronous detection;
- instantaneous frequency estimation;
- peak factor of OFDM signals;
- signal to noise ratio estimation;
- neural signal processing.

O.V. Ponomareva, A.V. Ponomarev, N.V. Ponomareva
Sliding Parametric DFT in the Tasks of Tonal Component Detection

The paper considers a generalization of the algorithm for a single-bin sliding DFT (SDFT) as a sliding parametric single-bin DFT (SDFT-P) which unlike standard SDFT enables to estimate the spectrum not at fixed frequencies but of a set of frequencies the number of which varies by Q parameter: {2(k+0)/N} where k= 0,N/2-1, 0 < Q < 1, N is a number of signal samples.

Y.S. Radchenko, S.V. Milyaev
On Correlation for Spectral Coefficients of Generalized Fourie Series in Continuous Space

A correlation for spectral coefficient of generalized Fourie series in continuous space within one block and between neighbor image blocks is estimated. Functional relations of standardized coefficients of intra- and interblock correlation of spectral coefficients to a correlation distance value are analyzed. General estimated relations are specified for the case of signal decomposition on Hermite function and Legendre polynomial bases.

B.A. Alpatov, P.V. Babayan, S.A. Smirnov, A.B. Feldman
Measuring Object Positions in Multispectral Image Sequence

A number of developed by the authors algorithms for measuring object positions in multispectral image sequences are considered in this paper. The experiments carried out with the use of nature video subjects show that a combined use of TV and IR band data enables to increase the efficiency of object tracking. To prevent problems arising due to a non-uniform background image and foreign moving objects available it is proposed to take into consideration the above mentioned directly at the stage of measuring object positions.

D.S. Grigoryan
Methods of Signal Data Decimation with Coherent Accumulation in Digital Spectral Analysis with Super-resolution

Phase and amplitude autoregression methods of linear prediction for spectral analysis with data decimation due to preliminary signal processing in bandwidth filters tuned to the energy center of signal frequencies are validated. Due to preliminary signal filtering prior to spectral estimation procedures, coherent accumulation of their energy is achieved. Compared to familiar autoregression methods of spectral analysis this increases the efficiency of super-resolution for sources with similar frequencies within a defraction resolution element.

A.Yu. Glukhov, I.V. Neginsky
Applying Synchronous Detection Method for Polyharmonic Signal Spectrum Estimation

Digital synchronous detection method to estimate absolute amplitudes and phases of strongly nonlinear substances with an arbitrary number of harmonic components is proposed. The possibility of applying this method to be used in real-time systems based on digital signal processors or microcontrollers is considered. The results of experimental verification of the method for a meander- type signal as well as the results of a numerical simulation on test mono- and polyharmonic signals with additive Gaussian noise are given.

R.A. Korotchenko

Method of Seismoacoustic Pulse Selection and Estimation in Computer Processing

Method and sequence of processing algorithms for computer-aided search, selection and estimation of seismoacoustic pulse parameters with a various degree of detailing a wave structure pattern are distributed. Pulse recognition and selection procedures are a success in a wide range of signal-to-noise ratio due to the use of pulse model as low-frequency modulation of background oscillations. Examples for processing real-time signals of seismoacoustic monitoring are presented.

A.G. Zuev, V.A. Uvarov, V.M. Golovkov
Widening the Cumulative Code Class

Methods of information encoding and decoding using pairs of single-channel cumulative cods (A and B) with zero side lobe signals of autocorrelation function (ACF) in even clock periods by a “flight-signal” common to all subscribers are considered. The paper presents the codes of no regular structure with even 6 to 30 digit (capacity) with multiple doubling the latter in a general recurrent formula. The advantage of cumulative codes in question is a wider choice of code length depending on specific conditions of using the code as one improving the validity of information received as well as doubling the efficiency of encoding.

M.A. Dubov, A.L. Priorov

Non-Data-Aided Estimation Method of Signal-to-Noise Ratio and Bit Error Rate for QAM Signals

The paper presents a comparative analysis of methods for measuring signal noisiness seldom described in Russian technical literature. Relatively simple methods for “blind” estimation of signal-to-noise ratio (SNR) for signals with quadrature modulation types (e.g. QAM) are proposed. Modified methods for a precise estimation of bit (symbol) error rate (BER, SER) using some frames and with no need of any standard sequences in a signal. Calculation-base results are also given.

E.N. Potekhin, A.N. Leukhin
Methods of Constructing Biphase and Four-Phase Unimodular Periodic Sequences with Irrational Phases

Method of constructing biphaseunimodular sequences based on periodic optimal binary sequences with a two-level autocorrelation function is proposed. Side-lobe level to sequence-length relation is shown. Method of constructing four-phase sequences with an ideal periodic autocorrelation function (PACF) based on quaternary optimal sequences is also proposed.

Yu.S. Shinakov

Peak Factor of OFDM Signals and Nonlinear Distortion in Radio Equipment of Wireless Access Systems

Using PC simulation and analytic methods this paper shows that nonlinear distortion analysis for power amplifiers of base and mobile stations for access systems with OFDM technology can be carried out by distortion analysis of both test signal envelope and instantaneous signal values. Power amplifier parameters being investigated are undisturbed signal power and signal-to-interference ratio at amplifier output.

A.N. Varnavsky
Heart Ventricle “Surge” Detection

Heart ventricle “surge” detection method based on analyzing electrocardiosignal by using nonlinear integral transformations is described. RS-T segment depression and negative T-wave selection independent of QRS- complex is proposed to be used.

V.A. Romanchuk, V.N. Ruchkin, V.A. Fulin
Modelling Complex Neuroprocessor System

Problems of modeling NM-640x family neuroprocessorbeing developed by scientific and engineering center “Module” as well as problems of modeling complex neuroprocessor-based computing systems are considered. Graphical presentations for models of scalar and vector neuroprocessor unit as highly parallel system are proposed. A model of neuroprocessor-based computing system as a state machine is also proposed. Analytical expressions of modeling results are shown. Case study results are presented.

If you have any question please write: