JPH10173717A - Multi-decision digital demodulation system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the scale of hardware, without damaging characteristics due to phasing by narrowing demodulation candidates per symbol and determining demodulation symbol candidate, based on the reliability of symbols in the past when demodulating an MPSK modulated signal. SOLUTION: A phase signal θk for four symbols is extracted in every symbol cycle from a signal, which is inputted to an input terminal 10 and stored in a buffer memory 100. While using this phase signal, the signal is converted to three kinds of differential signals θi by performing prescribed operation through a differential signal generator 200. Then, these differential signals are stored in a differential signal memory 250, together with differential signals similarly converted in the past. A phase decider 300 calculates and outputs an optimal symbol through evaluation out of the differential signals θi and the demodulation candidates stored in a demodulation phase candidate buffer 400 in the past, and at the same time, this symbol is stored in the buffer 400 as a demodulation symbol candidate. Thus, the number of demodulation candidates per symbol is narrowed so that arithmetic processing can be reduced markedly.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention demodulates a differentially coded MPSK modulated signal and selects from among a plurality of demodulated symbol candidates respectively output at the time of demodulation at each symbol time of the modulated signal. The present invention relates to a multi-decision digital demodulation method for selecting a desired demodulation symbol candidate.

[0002]

2. Description of the Related Art When demodulating a digitally modulated signal, there are basically synchronous detection and delay detection. Considering the noise due to regular white noise, it is known that the basic error characteristic of synchronous detection is superior to the error characteristic of differential detection by about 2 dB in Eb / No (for example, Reference 1, digital radio communication). Modulation and demodulation, Yoichi Saito, IEICE (Heisei 3
8 years)).

[0003] However, when the transmission path characteristics greatly change due to fading as in the case of mobile communications such as automobile telephones and mobile telephones, there is a problem that the error characteristics due to synchronous detection are degraded compared to the error characteristics due to delay detection. is there. Also,
In order to achieve optimum characteristics by synchronous detection, it is necessary to completely perform carrier synchronization and symbol synchronization.However, in a fading environment such as mobile communication, a clock for carrier recovery and symbol synchronization for carrier synchronization is used. It is difficult to perform reproduction with high accuracy. Therefore, in mobile communication, demodulation by differential detection has been considered.

The most basic method for improving the basic error characteristic of differential detection is to use multi-symbols. A demodulation method using multi-symbols of differentially encoded MPSK is described in, for example, Reference 2 (D.
K. Simon, "Multiple-Symbol
Differential Detection of
MPSK, "IEEE Com., Vol. 38,
No. 3, pp. 300-308, 1990). This method detects a plurality of symbols at the same time while adopting the structure of differential detection, and it has been confirmed that by increasing the number of simultaneously demodulated symbols, characteristics equivalent to those of synchronous detection can be obtained.

[0005]

However, the demodulation method using multi-symbols described in the above reference 2 has a drawback that the amount of calculation and the hardware scale are extremely large. Now, the phase phi _k of the transmitted signal at the symbol point k is assumed to be _{φ k = φ k-1 +} δφ k (1). In this case, δφk is a transmission symbol.

Here, the received signal at the symbol point is represented by r
_k and the modulation scheme in
Assuming that K is 3, and the number of simultaneously demodulated symbols is 3, the evaluation criterion for finding the optimal set of transmission symbols {δφ _k−1 , δφ _k } is to maximize the value η obtained by equation (2). is there. That is,

[0007]

(Equation 1)

In the equation (2), Re indicates a real number, and * indicates a conjugate relationship. Here, when the modulation scheme is QPSK, the value that δφ _k can take is:
{± π / 4, ± 3π / 4}, the above η
In order to determine the maximum value of, it is necessary to calculate the product of 4 ² × 6 complex numbers. Therefore, in the case of M symbols, 4 ^M
And there is a disadvantage that the amount of computation increases exponentially. Further, in the multi-symbol, since a received signal of a longer section is used for detecting one symbol than that of the differential detection, a signal which has deteriorated due to fading or the like is included in the section. There is also a disadvantage that is deteriorated.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to reduce the amount of arithmetic processing and suppress deterioration in characteristics due to fading when delay detection is performed using multi-symbols, and to realize the invention with smaller and less expensive hardware.

[0010]

SUMMARY OF THE INVENTION In order to solve such a problem, the present invention demodulates a differentially coded MPSK modulated signal, and simultaneously demodulates the modulated signal at each symbol time. A digital demodulation method having a function of outputting a demodulated symbol candidate of a symbol, when demodulating a modulated signal at each symbol time, using a plurality of demodulated symbol candidates output in a demodulation process at a past symbol time, the symbol This is to determine a demodulation symbol candidate that is determined to best represent the modulation signal at the time. In addition, when a plurality of demodulated symbol candidates are output, a function of outputting the reliability of each demodulated symbol candidate is provided, and when a modulated signal is demodulated at each symbol time, it is output in a demodulation process at a past symbol time. Based on the plurality of demodulated symbol candidates and the reliability of the demodulated symbols, a demodulated symbol candidate at the symbol time is determined.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In the present invention, the modulation method is Q
Considering that the k-th symbol δφ _k is determined as PSK, the demodulation result of the past three symbols is used to demodulate the symbol point. The evaluation criterion for demodulating the k-th symbol at this time is to maximize the value η obtained by Expression (3). That is,

[0012]

(Equation 2)

[0013]

(Equation 3)

In equation (3), Re indicates a real number, and * indicates a conjugate relationship. Also, the symbol δ
A symbol with a triangle added to φ indicates a previously determined symbol in the past, and a symbol without a symbol Δ indicates a symbol to be determined from now on. Here, in the past demodulated symbol points {k-3, k-2, k-1},
It is assumed that a plurality of past demodulated symbol candidates A indicated by the codes A and B, respectively, and the reliability B at that time have been determined. Therefore, the present invention does not use all combinations of the values {δφ _ki , i = 0,..., 3} in calculating the value η of the equation (3), and uses ｛j for δφ _ki (0 <i). = 1,
.., J _{(ki) w} } using the above-mentioned past demodulated symbol candidates C, their values, and δφ _k ∈ ｛± π / 4, ± 3
The above value η is calculated in the combination of π / 4｝, and the value of δφ _k having the maximum value as the maximum value is output as a demodulation result of the symbol point. In the present invention, at the same time,
Among the combinations of δφ _k {± π / 4, ± 3π / 4}, the top J _k values that increase the value of η and the value of η at that time are demodulated symbol candidates and the reliability of the demodulated symbol candidates. Output as

The reliability of each demodulated symbol candidate is calculated by J _{(ki) w} used for calculating η from the selected candidate J _ki.
Used to select. Here, if the reliability is smaller than a predetermined threshold value, it is considered that the waveform of the symbol section is significantly distorted due to fading or the like. Therefore, in such a case, by setting J _{(ki) w} = 0, it is possible to prevent an indeterminate determination result from affecting the demodulation result of a future symbol.

As described above, in the present invention, the above-described D-type calculation is performed in calculating η. On the other hand, since in the conventional manner so that the calculation of quadruplicate ^4, if J _{(ki) w} <
If it is 4, demodulated symbols can be obtained with a smaller amount of calculation than in the past. Here, 4 to J
Deterioration of the error rate by narrowing down to _{(ki) w} is 2 ≦ J _{(ki ) w}
If so, it is considered that there is almost no deterioration. This is QPS
Considering the case where normal white noise is superimposed on K, δφ _k
This is because the probability that an error occurs when the hamming distance is equal to or greater than 2 with the true value of is considered to be 0.1% or less. And
If the signal is received as a signal that is very far from the true value of δφ _k already obtained, the transmission error is quite large, and the reception results of all adjacent symbols are also incorrect. This is because it is considered that the receiving environment is not in the receiving environment.

Hereinafter, the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a multi-decision digital demodulator to which the present invention is applied. In the figure, 100 is a buffer memory, 200 is a difference signal generator, 250 is a difference signal memory, 300 is a phase determiner, 4
00 is a demodulation phase candidate buffer. Here, in the present embodiment, as described above, QPSK is considered as the modulation / demodulation method, and a signal including only phase information is considered as a signal to be demodulated. It is also assumed that the phase signal has been quantized by quantization means (not shown). Furthermore, the number of multi-symbols is 4
And

In FIG. 1, a reception phase signal φ _k extracted from a reception signal for each symbol period is supplied to an input terminal 10, and the reception phase signal φ _k is stored in a buffer memory 100. Here, it is assumed that the buffer memory 100 stores the reception phase signals φ _k , φ _k−1 , φ _k−2 and φ _k−3 for four symbols. The difference signal generator 200 uses the stored received phase signals for the four symbols to obtain the equation (4).
(6) to perform three types of difference signals θ ₁
To convert to ~θ _3. That is, θ ₁ = φ _k −φ _k−1 (4) θ ₂ = φ _k −φ _k−2 (5) θ ₃ = φ _k −φ _k−3 (6)

The thus converted difference signals θ _{1 to} θ ₃
Are stored in the difference signal memory 250 together with the difference signals θ _{4 to} θ ₆ received at the past symbol points and similarly converted by the difference signal generator 200 using the equations (7) to (9). θ ₄ = φ _k-1 −φ _k-2 (7) θ ₅ = φ _k-2 −φ _k-3 (8) θ ₆ = φ _k-1 −φ _k-3 (9)

In the phase determiner 300, the difference signal memory 2
Using the difference signal {θ _i } supplied from 50 and the past demodulation symbol candidate C stored in the demodulation phase candidate buffer 400, the value ε is calculated by Expression (10). That is,

[0021]

(Equation 4)

[0022]

(Equation 5)

Here, δφ _k is δφ _k ∈ ｛± π / 4,
There is a relationship of ± 3π / 4}, and A indicates a demodulated symbol candidate determined at a past symbol point. Here, among all the combinations for calculating the value ε, the phase determiner 300 outputs δφ _k in the combination that minimizes ε from the output terminal 20 to the outside as a determination result of the symbol point. At the same time, among all the combinations for calculating the value of ε, the upper Jk values for decreasing ε are output to the demodulation phase candidate buffer 400 and stored as the above-mentioned respective demodulation symbol candidates E.

FIG. 2 is a block diagram showing a second configuration of the multi-decision digital demodulator. The basic configuration of this embodiment is the same as that of the first example shown in FIG. The difference from the example of FIG. 1 is that a phase determiner 500 is provided instead of the phase determiner 300, and in the calculation of ε performed by the phase determiner 500, the reliability of the demodulated symbol candidates F is It has a function not to calculate θ _i including a value larger than a predetermined threshold value. However, when the number of terms for which θ _i is not calculated is N, the value ε at that time is multiplied by N / 6 and evaluated as a new ε.

In the present embodiment, ε thus calculated
Among the combinations, δφ _k in the combination that minimizes ε is output from the output terminal 20 to the outside as a determination result of the symbol point. Among all the combinations for calculating the value of ε, the upper Jk values for decreasing ε are output and stored in the demodulation phase candidate buffer 400 as the demodulation symbol candidates E, and each demodulation symbol candidate E is The value of ε at the time of selection is output to and stored in the candidate reliability buffer 600 as reliability. Note that when the phase determiner 500 calculates the value ε, the demodulation phase candidate buffer 400 selects a demodulation symbol candidate based on the reliability supplied from the candidate reliability buffer 600 and outputs it to the phase determiner 500. I do.

As described above, when delay detection is performed using multi-symbols, the amount of arithmetic processing can be significantly reduced by reducing the number of demodulation candidates per symbol. In addition, it has a function of calculating the reliability of the demodulation determination result and also has a function of automatically restricting a symbol section to be selected according to the calculated reliability. Characteristics can be secured.

[0027]

As described above, according to the present invention, the differentially encoded MPSK modulated signal is demodulated,
When a demodulation signal is demodulated by a digital demodulator having a function of outputting a plurality of demodulation symbol candidates at the time of demodulation at each symbol time of the modulation signal, a plurality of demodulations output in a demodulation process at a past symbol time are performed. Since the demodulation symbol candidates that are determined to best represent the modulation signal at the symbol time point are determined using the symbol candidates, it is possible to narrow down the demodulation candidates for each symbol. Can be greatly reduced, and the amount of hardware can be reduced. In addition, when a plurality of demodulated symbol candidates are output, a function of outputting the reliability of each demodulated symbol candidate is provided, and when a modulated signal is demodulated at each symbol time, it is output in a demodulation process at a past symbol time. Since the demodulated symbol candidate at the time of the symbol is determined based on the plurality of demodulated symbol candidates and the reliability related to the demodulated symbol, it is possible to secure at least the characteristic of the conventional delay detection even in a fading environment.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a multi-decision digital demodulator to which the present invention is applied.

FIG. 2 is a block diagram showing a second configuration example of the demodulation device.

[Explanation of symbols]

100: buffer memory, 200: differential signal generator, 2
50: difference signal memory, 300, 500: phase determiner,
400: demodulation phase candidate buffer; 600: candidate reliability buffer.

Claims (2)

[Claims] 1. A digital demodulation method having a function of demodulating a differentially encoded MPSK modulated signal and outputting a plurality of demodulated symbol candidates at the time of demodulation at each symbol time of the modulated signal. When demodulating a modulated signal at each symbol time, a demodulated symbol determined to best represent the modulated signal at the symbol time using a plurality of demodulated symbol candidates output in the demodulation process at the past symbol time. A multi-decision digital demodulation method comprising means for determining a candidate. 2. The method according to claim 1, further comprising: means for outputting the reliability of each demodulated symbol candidate when a plurality of demodulated symbol candidates are output; A multi-decision digital demodulation method, wherein a demodulation symbol candidate at a symbol time is determined based on a plurality of demodulation symbol candidates output in a demodulation process at the time and reliability of the demodulation symbol.