Algorithm for adaptive noise filtering in digital antennae arrays of satellite communication

Аuthors

Chistyakov V. A.

Compani «Information satellite systems of academician M.F. Reshetnev», 52, Lenin str., Zheleznogorsk, Krasnoyarsk region, 662972, Russia

e-mail: vitalys1367@mail.ru

Abstract

Currently, no radio equipment can do without an antenna system, ranging from primitive radios to complex space systems. So on modern communication spacecraft, complex antenna arrays (AR) are installed, which differ in their configuration, a set of elements and a number of other important parameters. The main trend of space technologies development is noise immunity of communication channels in conditions of both peace and wartime. The article proposes an algorithm for adaptive filtering of useful signals against the background of various disturbances, which allows ensure necessary interference protection of communication channels during messages transmission

The adaptive algorithm is based on the method of direct inversion of the signals correlation matrix at the input of the antenna system. The main idea of the algorithm consists in obtaining an optimal weighting factors vector that allows changing the amplitude-phase distribution of the antenna array in such a way that deep sags are formed in the directional pattern (DP) in the direction of the interference.

The optimal weighting factors vector estimation computing is being performed by direct inversion of the interference correlation matrix. This method implicate creating a sample estimate of the correlation matrix using a pack of training vectors.

The advantages of this algorithm include a high rate of convergence of the weight coefficients vector and a rather deep formation of sags in the DP of AR in the direction of interference. Nonetheless, this algorithm is difficult to implement, in the case when the correlation matrix is degenerate, which makes its inversion impossible. This problem occurs when the signals from the outputs of the antenna array elements are linearly dependent.

However, there are two ways to solve this problem.

The first method involves application of the so-called diagonal load, which allows increase the difference between the maximum and minimum values of the correlation matrix, allowing thereby solve the singularity problem. Also, this method is noteworthy in that it allows evaluating the correlation matrix when the number of training pack is less than the number of the antenna array elements.

The second method involves increasing the number of training pack so that their number is at least twice the number of the antenna array elements.

The process of adaptive filtering is represented by the example of a flat rectangular 25-element array antenna, which input receives useful and interfering signals. The narrow-band signals with FM4 modulation, acting within the main lobe of the directional pattern, are used as the useful signals. The interference, in turn, is represented as a white Gaussian noise with a wide spectrum. In consequence of the algorithm modeling, the author has formed an adaptive antenna array directional pattern, and presented a useful signal spectrum passed the adaptive processing, which allowed filtering-out the interfering signals components.

Thus, the performed simulation has confirmed theoretical calculations of the adaptive filtering algorithm for the direct inversion of the correlation matrix, which makes it attractive for applicaiton in modern digital antenna systems.

However, the main factors affecting the technical characteristics of the adaptive filtering algorithm should be noted as well.

The first factor affecting the high characteristics of the above-described algorithm is, in the first place, its own noise caused by analog and digital equipment included in the antenna system.

In particular, the internal noise of the analog-digital path is determined by the jitter of the aperture, jitter from sampling, and differential nonlinearity.

The second factor is the length of the sample of training packs. As was said above, the correlation matrix of the input signals might be degenerate, which would lead to an inaccurate estimation of the weight coefficients. Thus, to avoid this situation, it is necessary to increase the number of training pack.

The third factor, causing the operation errors of the adaptive filtering algorithm is the decorrelation of the signals in the receiving channels of the digital antenna array. This is due primarily to the nonidentity of the amplitude-phase characteristics of the antenna array channels, as well as the moments of signals sampling discrepancy in them. Thus, decorrelation of signals leads to the fact that the non-diagonal elements of the correlation matrix of the input signals decrease in modulus, which causes an incorrect estimate of the weight coefficients.

Keywords:

adaptive filtration, correlation matrix, directional pattern, signal spectrum, antenna array, radio source, noise immunity

References

1. Ratynskii M.V. Adaptatsiya i sverkhrazreshenie v antennykh reshetkakh (Adaptation and super-resolution in antenna arrays), Moscow, Radio i svjaz, 2003, 200 p.

2. Dzhigan V.I. Adaptivnaya fil'tratsiya signalov: teoriya i algoritmy (Adaptive filtration of signals: theory and practice), Moscow, Tekhnosfera, 2013, 528 p.

3. Lipovtsev Yu.V., Tret'yakova O.N. Osnovy vysshei matematiki dlya inzhenerov (Fundamentals of Higher Mathematics for Engineers), Moscow, Vuzovskaja kniga, 2009, 484 p.

4. Kester U. Analogo-tsifrovoe preobrazovanie(Analog-to-digital conversion), Moscow, Tekhnosfera, 2007, 1016 p.

5. Wu JD. Audio quality improvement of vehicular hands-free communication using variable step-size affine-projection algorithms, International Journal of Wavelets, Multiresolution and Information Processing, 2010, vol. 8, no. 6, pp. 875 - 894.

6. Ahmad M.S., Kukrer O., Hocanin A. A 2-D recursive inverse adaptive algorithm, Signal, Image and Video Processing, 2010, vol. 1, pp. 1 - 6.

7. Paulo S.R. Diniz. Adaptive filtering: algorithms and practical implementation, New York, Springer US, 2013, 652 p.

8. Zimin A.S., Krinitskii G.V. Trudy MAI, 2012, no. 51, available at: http://trudymai.ru/eng/published.php?ID=29151

9. Lewis F.L. Optimal estimation: with an introduction to stochastic control theory, New York, Wiley, 1986, 369 p.

10. Paleologu C., Benesty J., Ciochina S. Sparse adaptive filters for echo cancellation, CA, Morgan & Claypool, 2011, 124 p.

11. Dixit S., Nagaria D. LMS adaptive filters for noise cancellation, International Journal of Electrical and Computer Engineering, 2017, vol. 7, no. 5, pp. 2520 - 2529.

12. Shchesnyak S.S., Popov M.P. Adaptivnye antenny (Adaptive antennas), Saint Petersburg, VIKKA im. A.F. Mozhajskogo, 1995, 611 p.

13. Sally L. Wood, John R. Treichler, Michael J. Ready. Analysis of post-demodulator adaptive filter behavior to compensate for multipath channels for FSK signals, IEEE 13th Digital Signal Processing Workshop, 2009, pp. 337 - 341.

14. Manokhin A.E. Trudy MAI, 2013, no. 67, available at: http://trudymai.ru/eng/published.php?ID=41530

15. Richard W. Middlestead. Digital communications with emphasis on data modems, New York, Wiley, 2017, 824 p.

16. Bei X., Tamal B. Partial update least-square adaptive filtering, CA, Morgan & Claypool, 2014, 115 p.

17. Grigor'ev L.N. Tsifrovoe formirovanie diagrammy napravlennosti v fazirovannykh antennykh reshetkakh (Digital direction pattern forming in phased antenna arrays), Moscow, Radiotekhnika, 2010, 141 p.

18. Harry L. Van Trees. Optimum Array Processing, New York, Wiley, 2002, 1400 p.

19. Grant P.M., Kouen K.F.N. Adaptivnye fil'try (Adaptive filters), Moscow, Mir, 1988, 392 p.

20. Johnson Don H., Dudgeon D. Array signal processing, Englewood Cliffs, Prentice Hall, 1993, 537 p.

21. Grigor'ev V.A., Shchesnyak S.S., Gulyushin V.L., Raspaev Yu.A., Lagutenko O.I. Adaptivnye antennye reshetki (Adaptive antenna arrays), Saint Petersburg, Universitet ITMO, 2016, 118 p.

22. Van Veen B.D., Buckley K.M. Beamforming: A versatile approach to spatial filtering, IEEE ASSP Magazine, 1988, vol. 5, no. 2, pp. 4 - 24.

Download