Publication detail

Use of the Analog Neural Networks in the Adaptive Antenna Control Systems

TOBEŠ, Z., RAIDA, Z.

Original Title

Use of the Analog Neural Networks in the Adaptive Antenna Control Systems

English Title

Use of the Analog Neural Networks in the Adaptive Antenna Control Systems

Type

journal article - other

Language

en

Original Abstract

In the paper, original control system of adaptive antennas, which is based on Kalman filter, is presented and compared with earlier approaches to this problem. The designed control circuit eliminates some disadvantages of the control circuits based on the classical Kalman neural network and the Wang one, and enables a real time processing of quickly changing signals processed by adaptive antennas. Especially, the dependence of the convergence rate on ratio of eigenvalues and the risk of instability are significantly reduced.

English abstract

In the paper, original control system of adaptive antennas, which is based on Kalman filter, is presented and compared with earlier approaches to this problem. The designed control circuit eliminates some disadvantages of the control circuits based on the classical Kalman neural network and the Wang one, and enables a real time processing of quickly changing signals processed by adaptive antennas. Especially, the dependence of the convergence rate on ratio of eigenvalues and the risk of instability are significantly reduced.

Keywords

Analog recurrent neural networks, Kalman filter, adaptive antenna, convergence rate

RIV year

2002

Released

01.09.2002

Pages from

14

Pages to

21

Pages count

8

BibTex


@article{BUT40904,
  author="Zdeněk {Tobeš} and Zbyněk {Raida}",
  title="Use of the Analog Neural Networks in the Adaptive Antenna Control Systems",
  annote="In the paper, original control system of adaptive antennas, which is based on Kalman filter, is presented and compared with earlier approaches to this problem. The designed control circuit eliminates some disadvantages of the control circuits based on the classical Kalman neural network and the Wang one, and enables a real time processing of quickly changing signals processed by adaptive antennas. Especially, the dependence of the convergence rate on ratio of eigenvalues and the risk of instability are significantly reduced.",
  chapter="40904",
  number="3",
  volume="11",
  year="2002",
  month="september",
  pages="14",
  type="journal article - other"
}