Beam forming method
View Patent ↗A radio communications system has subscriber stations and a base station that is provided with an antenna system with a plurality of antenna elements that emit a downlink signal in a weighted manner with coefficients of a weighting vector. In order to form a beam, a plurality of weighting vectors is determined at the subscriber station and the determined weighting vectors are transmitted to the base station. In a subsequent working phase, the subscriber station selects a dominant vector among the determined weighting vectors and transmits an indicator of the selected weighting vector to the base station.
1. A method for beamforming in a radio communications system having a plurality of subscriber stations and a base station which has an antenna device with a plurality of antenna elements A 1 to A M , comprising:
transmitting a downlink signal to a subscriber station from each antenna element;
for each of the antenna elements, weighting the downlink signal transmitted thereby with a weighting coefficient w i , i=1, . . . , M, the weighting coefficients forming a current weighting vector;
in an initialization phase of the subscriber station:
producing a first spatial covariance matrix from the downlink signal,
determining eigen vectors of the first spatial covariance matrix, and
transmitting the eigen vectors from the subscriber station to the base station and using the eigen vectors as first weighting vectors, which correspond respectively to radiation directions of the antenna elements;
in a working phase:
selecting one of the first weighting vectors at the subscriber station as a currently dominant weighting vector,
transmitting an identifier of the currently dominant weighting vector from the subscriber station to the base station, and
using the currently dominant weighting vector to weight the downlink signal in a subsequent time slot;
in selected time slots, sending information on components of the first weighting vector from the subscriber station to the base station, the information on components of the first weighting vector being sent instead of the identifier of the currently dominant weighting vector; and
performing the initialization phase at greater time intervals than the working phase.
2. The method as claimed in claim 1 , wherein
the initialization phase is repeatedly performed,
differences are determined between the weighting coefficients of a first weighting vector determined in a current initialization phase and weighting coefficients of a first weighting vector determined in a preceding initialization phase, the differences being determined at the subscriber station,
the differences are transmitted to the base station, and
the differences are added to corresponding weighting coefficients of the first weighting vector determined in the preceding initialization phase to thereby reproduce the weighting coefficients of the first weighting vector at the base station for the current initialization phase.
3. The method as claimed in claim 2 , wherein
a second spatial covariance matrix is produced cyclically in the working phase,
eigen vectors are determined for the second spatial covariance matrix, and
the eigen vector which has the largest eigen value is selected and used as the weighting vector.
4. The method as claimed in claim 3 , wherein
each antenna element periodically emits a training sequence which is orthogonal to the training sequences of the other antenna elements, and
the first weighting vectors are determined on the basis of the training sequences received by the subscriber station.
5. The method as claimed in claim 4 , wherein
the number of first weighting vectors is two, and
the identification of the dominant weighting vector is transmitted in each time slot allocated to the subscriber station.
6. The method as claimed in claim 5 , wherein
the number of first weighting vectors is 2 n ,
the identification of the dominant weighting vector comprises n bits, and
the n bits are divided with portions of the n bits being transmitted respectively in time slots allocated to the subscriber station.
7. The method as claimed in claim 6 , wherein the current weighting vector is a linear combination of the first weighting vectors.
8. The method as claimed in claim 7 , wherein
a plurality of downlink signals, which each have a different space time block coding, are produced from a symbol sequence which is intended for the subscriber station, and
each of the downlink signals is weighted with a different current weighting vector.
9. The method as claimed in claim 8 , wherein the current weighting vector is chosen from the first weighting vectors when an LOS transmission path exists between the base station and the subscriber station.
10. The method as claimed in claim 9 , wherein, before completing a process of determining the first weighting vectors, the current weighting vector, which is used for transmitting the downlink signal during a time slot, is defined on the basis of the previously defined weighting vectors.
11. The method as claimed in claim 1 , wherein
the initialization phase is repeatedly performed,
the subscriber station determines mathematical signs of the differences between the weighting coefficients of a first weighting vector determined in a current initialization phase and weighting coefficients of a first weighting vector determined in a preceding initialization phase,
the mathematical signs are transmitted to the base station, and
the coefficients of the first weighting vector determined in the preceding phase are incremented or decremented by one unit depending on the corresponding mathematical sign.
12. The method as claimed in claim 1 , wherein
the subscriber station receives taps of the downlink signal corresponding to the antenna elements of the antenna device, and
the first spatial covariance matrix is produced individually for each tap of the downlink signal.
13. The method as claimed in claim 1 , wherein
a selected portion of the eigen vectors of the first spatial covariance matrix are used as the first weighting vectors, and
the eigen vectors which are selected and used are the eigen vectors which have the largest eigen values.
14. The method as claimed in claim 1 , wherein the first spatial covariance matrix is averaged over a plurality of time slots of the downlink signal.
15. The method as claimed in claim 1 , wherein
a second spatial covariance matrix is produced cyclically in the working phase,
eigen vectors are determined for the second spatial covariance matrix, and
the eigen vector which has the largest eigen value is selected and used as the weighting vector.
16. The method as claimed in claim 1 , wherein
each antenna element periodically emits a training sequence which is orthogonal to the training sequences of the other antenna elements, and
the first weighting vectors are determined on the basis of the training sequences received by the subscriber station.
17. The method as claimed in claim 1 , wherein
the number of first weighting vectors is two, and
the identification of the dominant weighting vector is transmitted in each time slot allocated to the subscriber station.
18. The method as claimed in claim 1 , wherein
the number of first weighting vectors is 2 n ,
the identification of the dominant weighting vector comprises n bits, and
the n bits are divided with portions of the n bits being transmitted respectively in time slots allocated to the subscriber station.
19. The method as claimed in claim 18 , wherein the dominant weighting vector is used for beamforming in the time slot immediately following its transmission.
20. The method as claimed in claim 18 , wherein
a bits are allocated to the subscriber station per time slot, and
the n bits are inserted into n/a time slots.
21. The method as claimed in claim 1 , wherein the number of time slots used to transmit weighting vector information is varied as a function of speed of movement of the subscriber station.
22. The method as claimed in claim 1 , wherein the current weighting vector is a linear combination of the first weighting vectors.
23. The method as claimed in claim 22 , wherein
the current weighting vector is a linear combination of the first weighting vectors, and
in the operating phase, the subscriber station transmits to the base station information relating to coefficients of the linear combination.
24. The method as claimed in claim 23 , wherein the information indicates at least one of a phase and a magnitude of one of the coefficients of the linear combination.
25. The method as claimed in claim 22 , wherein the coefficients of the linear combination of first weighting vectors have magnitudes respective to magnitudes of respective eigen values.
26. The method as claimed in claim 1 , wherein
a plurality of downlink signals, which each have a different space time block coding, are produced from a symbol sequence which is intended for the subscriber station, and
each of the downlink signals is weighted with a different current weighting vector.
27. The method as claimed in claim 1 , wherein the current weighting vector is chosen from the first weighting vectors when an LOS transmission path exists between the base station and the subscriber station.
28. The method as claimed in claim 1 , wherein, before completing a process of determining the first weighting vectors, the current weighting vector, which is used for transmitting the downlink signal during a time slot, is defined on the basis of the previously defined weighting vectors.
29. The method as claimed in claim 28 , wherein the previously defined weighting vectors each have one, and only one, non-vanishing component.