Filter for telecommunications system

Details Filter for telecommunications system Filter for telecommunications system

2000-11-02

2004-02-24

Chin, Vivian (Department: 2682)

Multiplex communications

Communication over free space

Combining or distributing information via time channels

C370S347000, C370S350000

Reexamination Certificate

active

06697351

ABSTRACT:

CROSS-REFERENCE TO RELATED APPLICATION
This application claims priority of European Patent Application No. 99308919.2, which was filed on Nov. 9, 1999.
FIELD OF THE INVENTION
This invention relates to a method and apparatus for filtering signals in a radio telecommunications system.
DESCRIPTION OF THE RELATED ART
In a radio telecommunications base station, the provision of adaptive or smart antennas is known to enhance capacity, mainly because such antennas have good Co-Channel Interference (CCI) rejection capability. However perfect timeslot synchronization is assumed between a desired signal and interfering signals, and small cells must be used to allow the application of standard antenna array processing techniques.
In a system in which timeslots are not synchronized, such as GSM (Global System for Mobile Communications) and EDGE (Enhanced Data Rates for GSM Evolution), or when large cells are employed, such assumptions cannot be made.
FIG. 1
illustrates a plot of amplitude A versus time t of a desired signal D, comprising a tail band T at each end, two pay loads P, and a midamble training sequence S.
FIG. 1
also shows 2 interfering signals, the first I
1
being of substantially greater amplitude than the second I
2
, and there being a break point B between them.
The CCI of
FIG. 1
can be considered as a nonstationary interference, and corresponding nonstationary processing techniques can be applied, as proposed by J Karlsson, “Adaptive antennas in GSM systems with non-synchronized base stations” Licentiate's thesis, Dept of Signals, Sensors and Systems, Royal Inst. Of Technology, Sweden, 1997, and by E Villier, L. Lopes, S Aftelak, “On application of uplink optimum combining to base station reception” in Proc. IEEE 48
th
VTC, pp747-752, Ottawa, 1998. A drawback of such an approach is that the techniques are not effective when a limited volume of data is available, such as the 148 symbols in a GSM timeslot.
In M C Wells, “Increasing the capacity of GSM cellular radio using adaptive antennas”, IEE Proc. Communications, vol 143, no.5, pp 304-310, 1996, it is pointed out that a Spatio-Temporal Filter (STF) can be used to enhance the desired signal D and reject both parts I
1
and I
2
of the asynchronous CCI. However conventional adjustment algorithms such as the Least Squares (LS) estimation may not be effective when the training sequence is concentrated in one part of the burst, as in GSM. It can be seen from
FIG. 1
that the training sequence S overlaps with the interfering signal I
2
, but there is no overlap with I
1
.
A solution would be to spread the training sequence S over the whole burst while keeping the total number of training symbols constant, but this would require a change of the GSM standard, which is not possible.
Another solution would be to use the symbols in the tails T as additional training symbols for CCI rejection, but this may be insufficient because there are not enough symbols.
Another solution, also proposed by Wells is to use a semi-blind algorithm with projection to the Finite Alphabet (FA); the FA property is associated with the whole timeslot of the desired signal and can be used to adjust the coefficients of a STF in asynchronous CCI conditions. This semi-blind solution uses the LS estimation based on only the training symbols as an initialization, and thus the estimator may still suffer from insufficient volume of training data overlapping part of the asynchronous CCI. Referring to
FIG. 1
, in GSM there will be 26 symbols available in the training sequence S, plus 3 from the right hand tail, to train an STF to reject the relatively weak interference I
2
, but only 3 symbols in the left hand tail T to reject the stronger interference I
1
.
A M Kusminskiy and D Hatzinakos, in “Semi-blind estimation of spatio-temporal filter coefficients based on a training-like approach”, IEEE Signal Processing Letters, vol.5 no.9 pp231-233, September 1998, propose a potential method for improving accuracy by finding a solution close to that based on an enlarged number of training symbols displaced to any position within a received slot of data, but a drawback is that complexity may be high because it increases exponentially with the number of additional training-like symbols in a slot.
A need exists for a method and apparatus for filtering signals which overcomes the disadvantages of the known methods.
SUMMARY OF THE INVENTION
In a radio telecommunications system in which each timeslot has a first tail band, a first data payload, a midamble sequence of training symbols, a second data payload, and a second tail band, in accordance with the invention a method and apparatus for filtering signals by determining which of the two ends of a timeslot suffers more from co-channel interference than the other end, selecting as training-like symbols a plurality of symbols in the data payload adjacent the tail band at said end, and utilizing the training-like symbols, the tail band symbols at said end, and the midamble training symbols in an algorithm to reject co-channel interference.
In this specification, a “training-like symbol” means an information symbol of any possible value which can be used to enlarge the available number of training data, i.e. the midamble training symbols and the tail band symbols of FIGS.
1
and
2
.


REFERENCES:
patent: 6002716 (1999-12-01), Meyer et al.
patent: 6567374 (2003-05-01), Böhnke et al.
patent: 0 535 403 (1992-09-01), None
patent: 0 615 352 (1994-03-01), None
patent: WO 97/41702 (1997-11-01), None
A. Kuzminskiy, et al “Semi-Blind Training-Like Estimation Of Spatio-Temporal Filter Coefficients For Finite Alphabet Signals” “9thIEEE Signal Processing Workshop On Statistical Signal and Array Processing” Portland, OR, U.S.A. Sep. 14-16. 1998 pp. 376-379.
A. Kuzminskiy, et al “Semiblind Estimation of Spatio-Temporal Filter Coefficients Based on a Training-Like Approach” “IEEE Signal Processing Letters, Sep. 1998” pp. 231-233.
European Search Report, dated Apr. 17, 2000.

Affiliated with

Also associated with

Rating

Filter for telecommunications system does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Filter for telecommunications system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Filter for telecommunications system will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3325035