1. |
- Gorokhov, A., et al.
(author)
-
Some Results on Blind Deconvolution Applied to Digital Communication Signals
- 1997
-
In: Proceedings of DSP 97. ; , s. 107-110
-
Conference paper (peer-reviewed)abstract
- Blind deconvolution techniques applied to spatially and/or temporally oversampled signals have recentlyattracted much interest in the research community. This contribution contains an analysis of experimental data collected from an antenna array in a suburban environment. The Noise Subspace (NS) technique of [1] and the Linear Prediction(LP) method of [2, 4] are examined. The real data examples presented demonstrate a case where joint spatial and temporal deconvolution has clear benefits as compared to decoupled processing.
|
|
2. |
- Kristensson, Martin, et al.
(author)
-
Asymptotic Comparison of two Blind Channel Identification Algorithms
- 1997
-
In: Proc. of IEEE Signal Processing Workshop on Signal Processing Advances in Wireless Communications. SPAWC 97. - : IEEE. ; , s. 361-364
-
Conference paper (peer-reviewed)abstract
- In this paper the performance of two second order based blind channel identification techniques is studied. The methods are compared theoretically and the formulas validated practically by simulations. The first method is a well known subspace approach and the second is a covariance matching estimator. This last estimator should attain the lower bound for the asymptotical estimation error covariance of any second order based blindidentification algorithm.
|
|
3. |
|
|
4. |
- Östman, Thomas, et al.
(author)
-
Asynchronous DS-CDMA Detectors Robust to Timing Errors
- 1997
-
In: Proceedings of IEEE 47th Vehicular Technology Conference. VTC97. - : IEEE. ; , s. 1704-1708
-
Conference paper (peer-reviewed)abstract
- An asynchronous DS-CDMA system with timing errors is considered. Two different approaches for dealing with the problem are presented. The first method is based on robustifying the MMSE estimator against timing errors. The second approach models the timing errors as an extra noise term and formulates the best linear unbiased estimator of the data symbols. Simulations of the bit error rate indicate that the two approaches have similar performance. However, the difference to non-robustified algorithms is large.
|
|