Публикация:Корогодин 2019 Adaptive Beamforming Algorithm in Real Numbers Arithmetic
Ilya V. Korogodin, Sergey P. Ippolitov, Ivan V. Lipa Adaptive Beamforming Algorithm in Real Numbers Arithmetic // 2019 Progress in Electromagnetics Research Symposium (PIERS). — 2019.
| BibTeX: |
@article{korogodinPIERS2019STAP,
author = "Ilya V. Korogodin and Sergey P. Ippolitov and Ivan V. Lipa",
title = "Adaptive Beamforming Algorithm in Real Numbers Arithmetic",
journal = "2019 Progress in Electromagnetics Research Symposium (PIERS)",
year = "2019",
language = english
}
|
Аннотация
Controlled reception pattern antennas (CRPA, serpers) are very useful for telecommunication, radar and navigation receivers. They allow the forming of several independent virtual radiation patterns and to control the patterns’ form. As a result, the receiver can gain useful signals and mitigate interferences and multipath signals. The mitigation performance is very high, so, for example, it’s the only known way to improve GNSS receivers’ antijam capability radically (from 40-50 dB to 90 and more dB). CRPA operation method is quite simple. Both the useful signals and the interference signals are captured by several antennas. The signals are considered as relatively narrowband, so, the signal samples can be described as complex numbers with different arguments. After covariance matrix accumulation and inverse matrix problem solving, it is possible to compute complex weights for the signals combination. The combination adds the interference signals antiphasely and mitigates them. The complex representation and the narrowband assumption for the signals are a convenient mathematical model. The model allows describing and implements signal delays as phase shifts. On other hand, it limits mitigated interferences bandwidths and imposes algorithm implementa- tion in complex numbers. The complex representation increases a computational resource overspending. In that case, it is required to use quadrature front-ends, four times more multipliers are needed to calculate the co- variance matrix and implement filters. Inverse matrix problem solving is much more complicated for complex numbers matrices. In the study, we propose an adaptive beamforming algorithm in real numbers arithmetic. It doesn’t use narrowband assumption for the signals. The algorithm operates by finite impulse response filters for the signals alignment, real numbers for the covariance matrix and the signals representation. The real number representation allows decreasing algorithm complexity: number of multipliers, inverse problem solving time and memory consumption.
Галерея
Принт статьи - PDF. Презентация - PDF.
