Pulse or digital communications – Receivers – Particular pulse demodulator or detector
Reexamination Certificate
2000-12-27
2004-07-13
Corrielus, Jean B. (Department: 2631)
Pulse or digital communications
Receivers
Particular pulse demodulator or detector
C708S704000, C708S704000
Reexamination Certificate
active
06763075
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to symbol estimation in a mobile communication system.
BACKGROUND
In high data rate mobile communication, system performance is degraded heavily by intersymbol interference ISI. The equalisation technique is one of the main issues for a receiver to combat ISI. It is well established that maximum likelihood sequence estimation MLSE, implemented by the Viterbi algorithm can provide optimum performance in terms of sequential error event probability. This technique however has great computational complexity which hinders it from use for channels with a long delay spread. In particular, when the technique is implemented using a trellis structure, the complexity of the trellis structure increases exponentially with channel length. Channel length is a time period which is defined as the influence depth (time duration) of each symbol. With a symbol transmitted at time t
0
, the symbol will be heard most significantly between times t
1
to t
2
(subsequent to t
0
). The channel length is considered to be t
2
−t
1
.
It is an aim of the present invention to reduce the complexity of symbol estimation, particularly but not exclusively for implementing symbol estimation in channels with a long delay spread using a trellis structure.
SUMMARY OF THE INVENTION
According to one aspect of the present invention there is provided a method of estimating symbols transmitted between a mobile station and a base station in a communication system comprising:
receiving via a communication channel a set of received signal samples which have travelled via different transmission paths, each signal sample conveying a symbol component;
estimating from the set of received signal samples a soft-output decision for the symbol which combines over a number of symbol components an estimated value for each symbol component with a likelihood parameter which indicates a level of reliability associated with that estimated value; and
using the soft-output decision as a feedback element to modify a subsequent set of received signal samples prior to estimating a soft output decision for the subsequent set of received samples.
According to another aspect of the present invention there is provided a symbol estimation circuit for use in a mobile communication system for estimating symbols, the circuit comprising:
a receiver arranged to receive via a communication channel a set of signal samples which have travelled via different transmission paths, each signal sample conveying a symbol component;
a estimator for estimating from the set of received signal samples a soft-output decision for the symbol which combines over a number of symbol components an estimated value for each symbol component with a likelihood parameter which indicates a level of reliability associated with that estimated value; and
a feedback path for feeding back the soft-output decision to modify a subsequent set of received signal samples prior to estimating a soft-output decision for the subsequent set of received samples.
The symbol estimation technique can be used recursively either within signal bursts or between signal bursts. That is, in a communication system wherein a sequence of signal bursts are received by the communication channel, a soft-output decision for each symbol in the signal burst may be estimated and used as a feedback element prior to estimation of a subsequent symbol in the same signal burst. Alternatively, a soft-output decision for a preceding signal burst can be used for a subsequent signal burst. The former is more likely to be reliable.
Estimation of the soft-output decision for the symbol can be carried out in a trellis equaliser in which state transitions are effected via a set of transition branches. The states for the trellis equaliser can initially be defined by channel taps of a channel impulse response estimated from the receive signal samples.
In the described embodiment, a first set of the channel taps is utilised for setting the initial states in a trellis equaliser and a second set of the channel taps are treated as interference and used to modify the input signal prior to estimating a subsequent soft-output decision.
A minimum phase channel impulse response can be generated using an all path prefilter.
In the technique according to the following described embodiments of the invention, a suboptimum soft-output algorithm SSA is used as a feedback decision as a tradeoff between optimality and implementation reality. The algorithm can be implemented using a trellis structure. The computational power needed for the described suboptimum soft-output algorithm is in the same range as that of using a Viterbi algorithm, but the memory needed is much less. Moreover, the algorithm generates a soft-output decision rather than a hard decision and the inventors have noted that soft decision feedback can give better performance than hard decision feedback for a low signal to noise ratio domain.
For a better understanding of the present invention and to show how the same may be carried into effect reference will now be made by way of example to the accompanying drawings.
REFERENCES:
patent: 6269124 (2001-07-01), Nagayasu et al.
patent: 6302576 (2001-10-01), Ono et al.
patent: 0843444 (1998-05-01), None
Kurt Metzger; “A new Approach to Sequential Detection”, European Transactions on Telecommunications and Related Technologies, vol. 5, No. 2, Mar. 1, 1994, pp. 235-244.
Jianjun Wu and A. Hamid Aghvami; “A new Adaptive Equalizer with Channel Estimator for Mobile Radio Communications”, IEEE Transactions On Vehicular Technology, vol. 45, No. 3, Aug. 1996, pp. 467-474.
International Search Report for PCT/EP98/05481.
International Search Report for PCT/EP98/04040.
Jarvela Mikko
Zhengdi Qin
Corrielus Jean B.
Nokia Corporation
Squire Sanders & Dempsey L.L.P.
LandOfFree
Symbol estimation using soft-output algorithm and feedback does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Symbol estimation using soft-output algorithm and feedback, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Symbol estimation using soft-output algorithm and feedback will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3251858