Pulse or digital communications – Equalizers – Automatic
Reexamination Certificate
1999-10-07
2002-02-19
Chin, Stephen (Department: 2734)
Pulse or digital communications
Equalizers
Automatic
C375S233000, C375S346000, C375S350000, C708S322000, C708S323000
Reexamination Certificate
active
06349112
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an adaptive equalizer, and more particularly to an adaptive equalizer which can compensate a signal distortion on a transmission path and which uses a delayed decision feedback sequence estimator.
2. Description of the Related Art
Generally, in a digital mobile communication, interference occurs between signal codes on a transmission path due to multi-path. The interference is a factor of transmission performance degradation. As one effectual measure to compensate the interference between the signal codes, equalizers of various types are used such as a maximum likelihood sequence estimation (MLSE) type.
In the digital mobile communication, the transmission path characteristic changes every moment in accordance with the movement of a mobile station. It is an adaptive equalizer that updates an impulse response of the transmission path in accordance with the transmission path characteristic change. At this time, various algorithms such as a least mean square (LMS) algorithm and a recursive least square (RLS) are used as the adaptive algorithm. It should be noted that the maximum likelihood sequence estimation is described by Horikoshi in “Waveform Equalization Technique for Digital Mobile Communication” (pp. 77-92, published from TRICEPS). Also, the adaptive algorithm such as the LMS algorithm is described on pages 33-45 in the same publication.
Conventionally, an adaptive equalizer using the adaptive maximum likelihood sequence estimation is disclosed in Japanese Laid Open Patent Applications (JP-A-Heisei 5-152893 and JP-A-Heisei 5-152894).
FIG. 1
shows a system employing a conventional adaptive equalizer. Referring to
FIG. 1
, after a reception signal is temporarily stored in a memory
60
, the reception signal is supplied to a Viterbi algorithm processing section
70
through a matched filter
61
. The Viterbi algorithm processing section
70
carries out the maximum likelihood sequence estimation to the reception signal and outputs an estimation result. It should be noted that the reception signal has a frame format shown in FIG.
3
and is composed of a known training signal and an unknown data signal.
An impulse response of a transmission path is used in the maximum likelihood sequence estimation by a transmission path estimating section
80
. When the reception signal is supplied from the memory
60
, the transmission path estimating section
80
first determines initial values by use of an adaptive algorithm such that a difference is made small between the reception signal and a replica signal. The replica signal is obtained by convoluting at the receiving end a training signal having a predetermined known value and an impulse response during the reception of the training signal. Also, the transmission path estimating section
80
updates the initial values during the reception of the data signal by use of the adaptive algorithm. The adaptive algorithm functions such that a difference is made small between the reception signal and the replica signal obtained by convoluting the impulse response and the estimation result from the Viterbi algorithm processing section
70
.
The training signal or the estimation result from the Viterbi algorithm processing section
70
are multiplied with the impulse response components h
0n
to h
Ln
by multipliers
82
-
0
to
82
-L through delay elements
81
-
0
to
81
-(L−1). The outputs of the multipliers
82
-
0
to
82
-L are added by an adder
83
. The output of the adder
83
is the replica signal to imitate the reception signal. An error signal is determined by an adder
84
to indicate a difference between the replica signal and the reception signal. A calculating section
85
calculates impulse response estimation values {Ehj} (j=0, . . . , L) using the error signal based on the adaptive algorithm and outputs them to the matched filter
61
and the Viterbi algorithm processing section
70
.
In the above conventional example, the calculating section
85
uses the least mean square algorithm having the function to estimate a plurality of impulse response components using a plurality of parameter correction coefficients. A calculating section
71
has the function to determine optimal correction coefficients based on pathmetric result obtained as the result of the maximum likelihood sequence estimation using the plurality of impulse response components.
By the way, the maximum likelihood sequence estimation type equalizer is an equalizer having the highest ability. However, there is a drawback in a large circuit scale, i.e., it is very calculation-intensive. Therefore, the development of the equalizer is carried forward to reduce a circuit scale without degrading the equalization ability. As one example, there is an equalizer using a delayed decision feedback sequence estimator (DDFSE) in which the maximum likelihood sequence estimator and a decision feedback equalizer (DFE) are combined. Such a delayed decision feedback sequence estimator is described in “NEC Research and Development” (January, 1997, pp. 74-80).
An example of the delayed decision feedback sequence estimation reception apparatus is described in Japanese Patent Application No. Heisei9-158172 (reference 2: corresponding to Japanese Laid Open Patent Application (JP-A-Heisei 11-8573) opened on Jan. 12, 1999).
FIG. 2
shows a schematic structure of the reference
2
. Referring to
FIG. 2
, when a reception signal is supplied, an impulse response is determined by a transmission path characteristic detector
41
during the reception of a training signal. Also, the amplitudes of the impulse response components are determined by an absolute value calculating unit
42
. A summing unit
43
classifies impulse response components into a maximum likelihood sequence estimation region, a decision feedback equalization region and an outside region other than the maximum likelihood sequence estimation region and the decision feedback equalization region. Also, the summing unit
43
determines summations (p, q and r) of the amplitude values for each region. After that, the summing unit
43
calculates the summations p, q and r one after another while shifting each region, to output to a maximum value detector
44
. The maximum value detector
44
carries out the calculation of P/(R+&agr;Q) and outputs timings corresponding to the maximum calculation result to a delayed decision feedback sequence estimator
45
(In this example, &agr;=1/7, and the values P, Q and R are the same as defined above).
The delayed decision feedback sequence estimator
45
determines a maximum likelihood sequence estimation region and a decision feedback equalization region of the impulse response components supplied from transmission path characteristic detector
41
in response to the timing signals which are supplied from the maximum value detector
44
. The delayed decision feedback sequence estimator
45
carries out a sequence estimation using the impulse response components in those regions, and outputs as the maximum likelihood estimation data.
Next, calculation for determining the optimal region of the impulse response components in the maximum value detector
44
will be described.
All the components of the decision feedback equalization region are ideally canceled through the feedback operation, and do not contribute to improvement or degradation of the estimation ability of the sequence estimator. Therefore, the estimation ability is determined based on the ratio P/R of the components of the maximum likelihood sequence estimation region to the components of the outside region. When the ratio is larger, the estimation ability is higher.
However, the decision feedback equalization region cannot be completely canceled due to errors such as a quantization error so that a component is left as a residual distortion. Therefore, it is possible to say that the estimation ability is higher when the ratio of the components of the maximum likelihood sequence estimation region to the ad
Chin Stephen
Ha Dac V.
Ostrolenk Faber Gerb & Soffen, LLP
LandOfFree
Adaptive equalizer compensating signal distortion on... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Adaptive equalizer compensating signal distortion on..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adaptive equalizer compensating signal distortion on... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2952423