JP2578775B2 - Signal receiver - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a maximum likelihood receiver, and particularly to a maximum likelihood receiver having a correlator to which digital signal processing technology is applied.

[Conventional technology]

When receiving a signal indicating discrete information from the transmitting side via data transmission, etc. via a transmission line with noise, it is possible to use a maximum likelihood receiver to minimize the error probability at the time of reception. is there. In this type of maximum likelihood receiver, a digital type circuit is used for the correlator and the likelihood comparison / judgment unit in order to accumulate circuits and reduce signal processing time. Therefore,
When the received signal is in an analog format, the level of the received signal is quantized and coded by an analog-digital converter having an equal quantization step, converted into a digital signal, and then guided to a correlator.

[Problems to be solved by the invention]

In the conventional maximum likelihood receiver in which the analog-digital converter is provided at the reception input terminal as described above, if the number of output bits of the analog-digital converter is set small in order to reduce the circuit size, the likelihood of the likelihood is reduced. Since the accuracy of the calculation result decreases due to the increase in the quantization noise, the reception error probability increases,
Conversely, if the number of bits is increased to reduce the error probability,
There is a problem that the circuit size becomes large and the price rises. In particular, this problem is remarkable when a wide dynamic range is required because the received signal level greatly changes depending on the length of the transmission path.

An object of the present invention is to provide a maximum likelihood receiver capable of solving the above-mentioned problems and reducing the number of bits after reception signal conversion without increasing a reception error probability as in the case of a conventional receiver. It is in.

[Means for solving the problem]

The signal receiver according to the present invention includes: an analog-digital converter that receives one of N different signal waveforms transmitted at a transmission side at an interval of T seconds as an analog reception signal and converts the reception signal into a digital signal; A limiter having a defined amplitude limiting characteristic, limiting an amplitude change of the analog reception signal to a range of an input signal level allowed to the analog-digital converter, and sending the analog-signal to the analog-digital converter; and N weighting signals. N multiplying circuits for multiplying the output of the analog-to-digital converter with the outputs of the analog-to-digital converters; N integrating means for integrating the outputs of the multiplying circuits; And a comparison circuit for outputting a discrimination result indicating the amplitude pattern.

〔Example〕

 Next, the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention.
FIG. 2 is a characteristic diagram for explaining the operation.

In FIG. 1, a correlation circuit group including multipliers 3-1 to 3-n and integrators 4-1 to 4-n and a comparison determination circuit 5 are configured in a digit format.
In this embodiment, the received signal in the analog format is
After that, the signal is converted into a digital signal by an analog-digital converter (ADC) 2. FIG. 2 illustrates an input / output level characteristic of the limiter 1. Limiter 1 of the present embodiment
Is a soft limiter, and when a received signal within a range of the input level from the lower limit −vm to the upper limit + vm is input, the output level is within a range smaller than the input level range, that is, a range from the lower limit −vm to the upper limit + vm. Is limited to

The received signal whose level has been compressed as described above through the limiter 1 is converted into a digital signal by the ADC 2 and sent to the multipliers 3-1 to 3-n. The natural number n is
This is the number of code types to be transmitted on the transmission side, and is a multiplier 3-1.
The through 3-n, respectively, as described later, are given to no weighting signals w ₁ shows each amplitude pattern of the transmission signal corresponding to the n code w _n. Multipliers 3-1 through 3-n includes a to weighting signals w ₁ no w _n, performs multiplication of the reception signal sent from ADC2, gives to the 4-n no integrators 4-1 each multiplication result . Each of the integrators 4-1 to 4-n first resets to a predetermined initial value at each code period break, and then sequentially accumulates the multiplication results of the multipliers 3-1 to 3-n, The result of the accumulation is sent to the comparison and judgment circuit 5. The comparison / determination circuit 5 compares the final cumulative addition results of the integrators 4-1 to 4-n, that is, the likelihood of each code with each other at each code period break, and outputs a data signal indicating the code of the maximum likelihood. Is generated and output as an output signal.

The following describes the weighting signal w ₁ to w _n.

This weighting signal is determined by the method described in JP-A-61-199358. More specifically, in FIG. 1, a transmitter (not shown) transmits N different waveforms at intervals of T seconds. N is a positive integer, which is the case of M = 3 in the embodiment circuit shown in FIG. It is assumed that waveform distortion does not occur in the transmission path. this is,
This corresponds to a case where the distortion of the transmission path can be sufficiently removed by a receiving filter (not shown). In the integrator, it is assumed that white Gaussian noise average power sigma ² is applied from the outside.

ADC2 performs sampling M times during T seconds. Vector results of sampling _{X p = (X p1, X} p2, ......, X pM) represented by. Here, p is a received signal sample value vector for the p-th transmission. N multiplication circuits 3-1 to 3-n
, Weighting vectors W _q = (W _q1 , W _q2 ,..., W _qN ) (q = 1,..., N) corresponding to N transmission waveforms are defined. In each multiplier, the input sample value vector X _p
And the weight vector _Wq . That is, become. Here, the inner product of the vectors is shown.

The digital data comparison circuit 5 calculates Y _{p, q} (q = 1,...)
N), the one having the maximum value is selected and output as the received signal identification result. When the discrimination result as _{q 0, q op = MAX (} y p1, y p2, ......, y pN) a ... (2). In addition, the transmission signal is represented by a sample value vector as _{Sp, q} = ( _{Sp, a, 1} , _{Sp, q, 2} , ..., _{Sp, q, M} ) (3). Performing a pay judgment (maximum likelihood test) on the duplicated sample value vector {X _p }
The p-th determination result is Is determined by

Next, calculation of p ( _Xp / _Sq ) is performed. Assuming that there is no interference between each sample point of the received signal sample value vector and that they are independent, for the transmitted signal vector _Sq to be calculated here, the probability that the current received vector _Xp occurs corresponds to It is the product of the posterior probabilities calculated for each sample point. Assuming white Gaussian noise of average power σ ² as noise, Becomes Here, since the value of the error function erf (·) is smaller than 1, it is assumed that higher-order terms can be ignored. Becomes It can be seen that it is possible to estimate the transmit vector from the received vector X _p by evaluating the equation (6) as described above. The value of equation (6) is the second value in {}
The second term is the inner product of the vector X _p and the vector {erf (Sq, i / 2σ)}. Is determined.

Note that the sample value S _{p, i} of the transmission signal is calculated by the equation (7).
Can be determined from the transmission waveform. Σ can be determined by assuming noise power σxσ by actual measurement or the like.
If S _{p, 1} and σ are determined, the error function erf (·) can be calculated, and the value of the weight vector is determined.

Next, the circuit operation of FIG. 1 will be described with reference to FIGS.

Here, a case where the number of weighting signals is two (w ₁ , w ₂ ) is illustrated for simplification of description.

First, FIG. 3 shows the relationship between the input and output of the limiter,
Input signals above a certain value are limited by the limiter,
Supplied to C. As shown in FIG. 4, the input signal limited by the limiter is multiplied by a weighting signal in multipliers 3-1 and 3-2, and then integrators 4-1 and 4-2 as shown in the figure. Is integrated.

The comparison determination circuit 5 outputs the one with the largest value of the integrator as the one with the highest likelihood (determined as a received signal). In FIG. 4, since the output of integrator 1 is larger than the output of integrator 2,
The data (data corresponding to the input signal) indicated by the integrator 1 is output.

On the other hand, a signal smaller than the limiter amplitude (Fig. 5)
In this case, as shown in FIG. 6, the result of the multiplication is integrated by the integrator, and the comparator 5 determines that the larger value is the received signal and outputs it.

Thus, by providing the limiter 1, the level range of the received signal is compressed and given to the ADC 2 as described above. In addition, the level compression ratio is small in a low input level range where the signal-to-noise (S / N) tends to be low, and large in a high input level range where the S / N ratio is high.
Input / output level characteristics are set. As a result, comparing with the conventional receiver that directly supplies the received signal to ADC2,
In the present embodiment, the number of output bits of the ADC 2 required to make the accuracy of the likelihood obtained by the correlation circuit the same level is smaller in the present embodiment.
Further, as the number of bits of the ADC 2 is reduced, the number of input bits of the multipliers 3-1 to 3-n is also reduced, and the circuit scale is reduced without increasing the reception error probability as in the conventional receiver. , A reduction in price can be achieved.

Even if a hard limiter is used as the limiter 1,
Similar effects can be obtained.

〔The invention's effect〕

As described above, the present invention has an effect of realizing a maximum likelihood receiver capable of reducing the number of bits after reception signal conversion without increasing the reception error probability as in the case of the conventional receiver.

[Brief description of the drawings]

1 and 2 are block diagrams showing an embodiment of the present invention and characteristic diagrams for explaining the operation thereof, respectively, and FIGS. 3 to 6 are diagrams for explaining the operation of the circuit of FIG. It is. 1 Limiter 2 Analog-to-Digital Converter (ADC) 3-1 to 3-n Multiplier 4-1 to 4-n
…… Integrator, 5 …… Comparative judgment circuit.

 ──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-60-6546 (JP, A) JP-A-60-9060 (JP, A) JP-A-60-128735 (JP, A) JP-A-56 61010 (JP, A) JP-A-61-199358 (JP, A)

Claims (1)

(57) [Claims] 1. An analog-to-digital converter for receiving one of N different signal waveforms transmitted at a transmission side at an interval of T seconds as an analog reception signal and converting the reception signal into a digital signal; A limiter having a limiting characteristic and limiting an amplitude change of the analog reception signal to a range of an input signal level allowable to the analog-digital converter and sending the analog-digital signal to the analog-digital converter; N multiplying circuits for multiplying the outputs of the digital converters, N integrating means for integrating the outputs of the multiplying circuits, and comparing the outputs of the integrating means to obtain the maximum likelihood amplitude pattern. A comparison circuit that outputs an identification result indicating the following.