Auto-calibrating adaptive array for mobile telecommunications

Wennstrom, M.; Strandell, J.; Oberg, T.; Lindskog, Erik; Rydberg, Anders

doi:10.1109/7.845275

Cited by 2 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Calibration errors was studied by Tsoulos, et al [14][15], where measurements on an adaptive array using a calibration algorithm were presented. Wennström, et al presented an on-line calibration algorithm that is transparent to main antenna operation, continuously tracking the changes in antenna channels [16]. The algorithm use a feedback of the output RF signal, and thus allows the use of, for example, the least mean square (LMS) algorithm for updating the corrected weights.…”

Section: Calibrationmentioning

confidence: 99%

“…To mitigate this residual error, a feedback of the beamformer output signal y(t) could be used to iteratively tune the weights so the weight and calibration errors decreases. A method for this was presented in [16].…”

Section: B Output Signal Powermentioning

confidence: 99%

“…The power of the desired, or useful signal in the beamformer output is described by the first term in (12). The two other terms in (12), and (13) and (16) are assumed to be interferer and noise power terms. Thus, the beamformer output CINR can be expressed as…”

Section: B Output Signal Powermentioning

confidence: 99%

See 2 more Smart Citations

Effects of finite weight resolution and calibration errors on the performance of adaptive array antennas

Wennstrom

Oberg

Rydberg

2001

IEEE Trans. Aerosp. Electron. Syst.

Self Cite

View full text Add to dashboard Cite

Adaptive antennas are now used to increase the spectral efficiency in mobile telecommunication systems. A model of the received carrier-to-interference plus noise ratio (CINR) in the adaptive antenna beamformer output is derived, assuming that the weighting units are implemented in hardware. The finite resolution of weights and calibration is shown to reduce the CINR. When hardware weights are used, the phase or amplitude step size in the weights can be so large that it affects the maximum achievable CINR. It is shown how these errors makes the interfering signals "leak" through the beamformer and we show how the output CINR is dependent on power of the input signals. The derived model is extended to include the limited dynamic range of the receivers, by using a simulation model. The theoretical and simulated results are compared with measurements on an adaptive array antenna testbed receiver, designed for the GSM-1800 system. The theoretical model was used to find the performance limiting part in the testbed as the 1 dB resolution in the weight magnitude. Furthermore, the derived models are used in illustrative examples and can be used for system designers to balance the phase and magnitude resolution and the calibration requirements of future adaptive array antennas.

show abstract

Section: Calibrationmentioning

confidence: 99%

Section: B Output Signal Powermentioning

confidence: 99%