JPH0770947B2 - Adaptive receiver - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an adaptive receiver, and more particularly to an adaptive receiver that removes waveform distortion caused by strong multipath fading propagation.

(Prior Art) FIG. 4 shows a conventional adaptive receiver that removes waveform distortion caused in a multipath fading line. In FIG. 4, 40
1 is a shift register that constitutes a tapped delay element, 402
Is a multiplier, 403 is an adder, 404 is a decision feedback equalizer (DF
E) and 405 are correlators. A matched filter (MF) is composed of components other than DFE404. Correlator 405 is DFE
The tap coefficient of the matched filter is obtained by correlating the determination signal _n of the output signal of 404 and the received signal on each tap of the shift register 401. These tap coefficients are complex conjugates due to the time reversal of the impulse response of the transmission system. That is, the received signal on the shift register 401 is subjected to matched filtering by being convolved with the tap coefficient obtained by the correlator 405 in the multiplier 402 and the adder 403. The output of this matched filter is passed through the DFE 404, and intersymbol interference is removed.

This MF / DFE receiver is based on the Institute of Electronics and Communication Engineers, Communication Method Research Group 19
It was proposed in February 1979 (CS78-203) as an "adaptive reception method in a multipath transmission line", and has already been put to practical use in non-line-of-sight communication which becomes a severe multipath fading line. This adaptive reception scheme improves the DFE characteristics for non-minimum phase transition fading in which the leading edge of the impulse response is larger than the main response. Since the asymmetrical impulse response is symmetrized by the matched filtering, a part of the power of the Precursor is converted to the trailing edge (Postcursor) of the impulse response. Therefore,
The load on the DFE linear equalization unit is reduced, and the DFE equalization characteristics are improved. On the other hand, in the minimum phase shift fading in which Postcursor dominates, the Po
Since a part of stcursor is converted to Precursor, DFE will perform linear equalization in addition to nonlinear equalization. Ie MF
In the case of a DFE without using, only equalization by decision feedback is performed, whereas in MF / DFE, the DFE performs linear equalization in addition to nonlinear equalization, so it is degraded from the equalization characteristics of the DFE alone. When this adaptive reception method is applied to terrestrial multilevel QAM microwave transmission, distortion cannot be ignored due to this matched filtering as the multilevel level increases.

(Problems to be Solved by the Invention) In the conventional MF / DFE receiver described above, the equalization characteristic of DFE is improved with respect to non-minimum phase shift fading,
There is a problem with the minimum phase shift fading that the equalization characteristics are worse than those of the DFE alone.

Therefore, an object of the present invention is to provide an adaptive receiver that solves the characteristic deterioration due to the minimum phase shift fading.

(Means for Solving the Problem) A first adaptive receiver according to the present invention comprises a matched filter which is composed of a transversal filter and which receives a received signal, and a decision feedback type which receives an output signal of the matched filter. An equalizer, a correlator that correlates a decision signal output from the decision feedback equalizer with a received signal on each tap of the matched filter, and a transmission system based on the correlation value output from the correlator And a control circuit for monitoring the impulse response of each of the impulse response states and multiplying the correlation value by a weighting coefficient as tap coefficients of the matched filter corresponding to the impulse response state obtained by the monitoring. .

A second adaptive receiver according to the present invention comprises a matched filter which is composed of a transversal filter and which receives a received signal, a decision feedback equalizer which receives an output signal of the matched filter, and the decision feedback. Transversal filter having the same configuration as the matched filter for inputting the decision signal output from the equalizer, a delay element for delaying the received signal, an output signal of the transversal filter and an output signal of the delay element And a tap for calculating a tap coefficient of the transversal filter that minimizes the error signal from the error signal output from the subtractor and the determination signal. A coefficient calculation circuit and a weighting factor for the tap coefficient corresponding to an impulse response state obtained from the tap coefficient generated by the tap coefficient calculation circuit. Characterized in that it comprises a control circuit for outputting a respective tap coefficients of the matched filter are multiplied by.

(Example) Next, this invention is demonstrated with reference to drawings. FIG. 1 is a block diagram showing the configuration of an embodiment of a first adaptive receiver according to the present invention. FIG. 2 is a block diagram showing the configuration of an embodiment of the second adaptive receiver according to the present invention. Third
The figure is an explanatory diagram for comparing the operations of the conventional method and the present invention.

In FIG. 1, 101 is a matched filter (MF) and 101a is 5
Stage shift register, 101b is a multiplier, 101c is an adder, 10
2 is a decision feedback equalizer (DFE), 103 is a correlator, 104 is a control circuit, 104a is a controller, and 104b is a multiplier.

In FIG. 2, 201 is a matched filter (MF) and 201a is 5
Stage shift register, 201b is a multiplier, 201c is an adder, 20
2 is a decision feedback equalizer (DFE), 203 is a delay element, 204 is a control circuit, 204a is a controller, 204b is a multiplier, 205 is a 5-stage shift register, 206 is a multiplier, 207 is an adder, Reference numeral 208 is a subtractor, and 209 is a tap coefficient calculation circuit.

In FIG. 1, the shift register 101a is normally set to the symbol interval T or T / 2, and particularly when it is set to T / 2, the MF 101 has a gastric iming control function in addition to the matched filtering. Where the transmission symbol string is a
_n (n = -∞ ... + ∞), where h _n is the discrete value of the impulse response of the transmission system until it is input to the MF101, the discrete value r _n of the received signal is Indicated by. When the shift register 101a has T intervals, the reception signal r _{n + 2} represented by the equation (1) is provided at each stage (tap W _-2 , W _-1 , W ₀ , W ₊₁ , W ₊₂ ) of the shift register 101a. , r _{n + 1} , r _n , r _n-1 , r _n-2
Are distributed respectively. When the decision output of the DFE 102 is _n , the received signal on each tap of the shift register 101a is shown as follows.

W _-2 tap → r _{n + 2} =… h ₊₁ a _{n + 1} ＋ h ₊₂ a _n ＋ h ₊₃ a _n-1 ＋… W _-1 tap → r _{n + 1} ＝… h ₀ a _{n + 1} ＋ h ₊₁ a _n + h ₊₂ a _n-1 +… W ₀ tap → r _n =… h _-1 a _{n + 1} + h ₀ a _n + h ₊₁ a _n-1 +… W ₊₁ tap → r _n-1 =… H _-2 a _{n + 1} + h _-1 a _n + h ₀ a _n-1 +… W ₊₂ taps → r _n-2 =… h _-3 a _{n + 1} + h _-2 a _n + h _-1 a _{n −1} + ... Therefore, the impulse response discrete value h ₁ can be obtained by correlating the determination output _n of the DFE 102 and the content of each tap of the shift register 101a with the correlator 103. These tap coefficients are related to the impulse response of the transmission system by complex conjugate and time inversion. Each output of the correlator 103 is input to the multiplier 104b in the control circuit 104, and is multiplied by the coefficients g _-2 , g _-1 , g ₊₁ and g ₊₂ from the controller 104a. However, W ₀ corresponding to the center tap of MF 101 is not multiplied. When the coefficients g ₁ are all 1, the output of the control circuit 104 is the tap coefficients W _-2 , W _-1 , W ₀ , W of the MF101.
₊₁ and W ₊₂ are supplied to the multiplier 101b, and the MF 101 performs matching filtering by convolving the received signal and the tap coefficients W _-2 , W _-1 , W ₀ , W ₊₁ and W ₊₂ . The controller 104a is the correlator 10
Each output of 3 is input, and the impulse response state of the transmission system is monitored by this signal. In the case of the non-minimum phase transition fading of the main wave + leading wave as shown in (a) of FIG. 3, all the tap coefficients g ₁ of the multiplier 104b of the controller 104a are set to 1
Output as. That is, the adaptive receiver of FIG. 1 performs exactly the same operation as the MF / DFE shown in FIG. 3 (c) to make the impulse response of (a) symmetrical as shown in FIG. 3 (e).
That is, the leading edge (Precursor) of the impulse response is reduced equivalently, and the equalization ability of the DFE 102 with respect to the Precursor is improved. On the other hand, in the case of the minimum phase transition fading of the main wave + delayed wave, the controller 104a gradually reduces the tap coefficient g ₁ to zero. Here, the decreasing change speed is slower than the tracking speed of the DFE 102. At this time, since matched filtering is not performed on the impulse response as shown in FIG. 3 (b), the impulse response state (f) at the input of the DFE 102 is obtained. That is, most of the multipath distortion is due to the trailing edge of the impulse response (Postcursor), and all of these are removed by equalization by the decision feedback of the DFE 102. Since the MF / DFE also makes the impulse response symmetrical with respect to FIG. 3 (b) as shown in FIG. In contrast, since matched filtering is not performed, high equalization performance can be obtained as in the case of a DFE alone.

Next, the embodiment shown in FIG. 2 will be described. In FIG. 2, the received signal represented by the equation (1) is the MF201 and the delay element 203.
To enter. The judgment output _n of the DFE202 is the shift register 20.
5, the multiplier 206 multiplies the tap coefficients W ₋₂ , W ₋₁ , W ₀ , W ₊₁ and W ₊₂ of the output of the control circuit 204, and the adder 207 combines them. The convolution value of this judgment signal _n and the estimated impulse response value W ₁ is the reproduced waveform (replica) of the received signal.
Has become. The difference between the received signal delayed by the delay element 203 and the output of the adder 207 is subtracted by the subtractor 208, which becomes the error signal e of the reproduced waveform for the received signal. The tap coefficient calculation circuit 209 uses the error signal e and the determination signal _n.
And the impulse response estimation value W ₁ is sequentially calculated by the LMS algorithm shown below.

_{^{▲ W n + 1 1 ▼ =}} ▲ W n 1 ▼ -βe n · n-1 ... (2) where, β is a correction factor. In the equation (2), the parameter n indicates the time for each symbol. Shift register 20
5 and the shift register 201a of MF201 both have a symbol length T interval, shift register 205, multiplier 206, adder
Delay time of replica filter consisting of 207 and MF201 is 2T
Becomes In addition, the forward equalizer (linear filter) of the DFE202 is set to T
When the interval is 3 taps, the delay time of DFE202 is 2T. In this case, if the delay amount of the delay element 203 is set to 6T, the correct timing relationship is established. Through the above operation, the root mean square value of the error signal e is controlled to be the minimum, and the estimated value W ₁ of the impulse response is obtained.

The output of the tap coefficient calculation circuit 209 is input to the control circuit 204,
Multiplier 2 excluding the coefficient corresponding to the center tap of MF201
At 04b, the coefficients g _-2 , g _-1 , g ₊₁ and g ₊₂ from the controller 204a are multiplied respectively. The controller 204a monitors the impulse response by the tap coefficient from the tap coefficient calculation circuit 209, outputs ₁ as g ₁ for the non-minimum phase transition fading, and outputs the tap coefficient as in the embodiment of FIG. W ₁ from the arithmetic circuit 209 is supplied as it is to the multiplier 201a to cause the MF 201 to perform matched filtering. MF201 components 201a, 201b, 2
01c performs the same operation as 101a, 101b, 101c in FIG. On the other hand, for minimum phase shift fading, g ₁ is set to DFE202
Gradually decreases to zero at a speed that can sufficiently follow. Therefore, matched filtering is not performed for the minimum phase shift fading, and a high equalization capability is obtained as in the case of the DFE alone.

(Effects of the Invention) As described above, the present invention improves the equalization characteristics of the DFE alone by performing matched filtering for non-minimum phase transition fading, and for minimum phase transition fading. By not performing matched filtering, the equalization characteristic of MF / DFE is not deteriorated as compared with the equalization characteristic of DFE alone.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of a first adaptive receiver according to the present invention. FIG. 2 shows the second of the present invention.
2 is a block diagram showing the configuration of an example of the adaptive receiver of FIG. FIG. 3 is an explanatory diagram for comparing the operation of the adaptive receiver according to the present invention and the conventional adaptive receiver, and FIG. 4 is a block diagram showing the configuration of the conventional adaptive receiver. 101,201 …… Matched filter (MF), 101a, 201a, 205,401…
... 5-stage shift register, 101b, 104b, 201b, 204b, 206, 40
2 ... Multiplier, 101c, 201c, 207,403 ... Adder, 102,202,
404 ... Decision feedback equalizer (DFE), 103,405 ... Correlator, 104,204 ... Control circuit, 104a, 204a ... Controller, 203
... delay element, 209 ... tap coefficient calculation circuit.

Claims (2)

[Claims] 1. A matched filter configured by a transversal filter for inputting a received signal, a decision feedback equalizer for inputting an output signal of the matched filter, and an output from the decision feedback equalizer. A correlator that takes the correlation between the judgment signal and the received signal on each tap of the matched filter, and the impulse response of the transmission system is monitored from the correlation value output from the correlator, and the impulse response state obtained by this monitoring is obtained. An adaptive receiver comprising: a control circuit that outputs the correlation value multiplied by a weighting coefficient as each tap coefficient of the matched filter. 2. A matched filter configured by a transversal filter for inputting a received signal, a decision feedback equalizer for inputting an output signal of the matched filter, and an output from the decision feedback equalizer. A transversal filter having the same configuration as the matched filter that inputs a determination signal, a delay element that delays the received signal, and an error signal by taking the difference between the output signal of the transversal filter and the output signal of the delay element And a tap coefficient calculation circuit for calculating a tap coefficient of the transversal filter that minimizes the error signal from the error signal output from the subtractor and the determination signal, and the tap coefficient. The matching is obtained by multiplying the tap coefficient by a weight coefficient corresponding to the impulse response state obtained from the tap coefficient generated by the arithmetic circuit. Adaptive receiver, characterized in that it comprises a control circuit for outputting a respective tap coefficients of the filter.