JPH0738608B2 - Satellite receiver C / N display device - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a satellite receiver for satellite broadcasting or satellite communication, and in particular, it is useful for adjusting the installation angle of an antenna by displaying the quality of received radio waves. Can C / N
The present invention relates to a display device.

Conventional technology It has been a long time since the broadcasting of TV signals from broadcasting stations using geostationary satellites became practical. In the conventional TV signal transmission, as the antenna direction adjustment and image quality display, in the detection output after FM detection,
It was previously disclosed in Japanese Patent Application No. 62-334535 (Japanese Patent Application Laid-Open No. 1-1754) to pick up and use noise in a certain portion having no signal component.
No. 87). In recent years, a system for directly transmitting only data, not a TV signal, has been developed, and noise within a band cannot be used.

DISCLOSURE OF THE INVENTION Problems to be Solved by the Invention As described above, in a system in which only data is directly transmitted, the spectrum in the band is spread, and it is impossible to pick up only noise without a signal,
In addition, in the configuration for picking up noise as described above, since the analog signal processing is performed, there are large errors due to temperature changes and component variations.

Means for Solving the Problems In order to solve the above problems, the satellite receiver C / N display device of the present invention is a Viterbi decoding used when the digital convolutional code after demodulation detection is decoded on the receiver side. At
A means for detecting the number of errors is provided by a state metric (path metric) by a Hamming distance calculation when the convolutional code is processed in a hard-decision Viterbi decoding algorithm,
A microcomputer that is equipped with a means for counting the number of errors, measures C / N to BER (Bite Error Rate, bit error rate) in advance in the adjustment process, and inputs correction data corresponding to the measured value. Is provided and the count number of the error number is input to the microcomputer, and the C / N value corresponding to the count number of the error number is displayed on the display.

With the above-described configuration, the present invention can detect the number of errors by observing the state metric in the Viterbi decoding algorithm when Viterbi decoding the satellite digital data on the receiver side. In addition, if the C / N vs. BER (bit error rate) is measured in advance and the correction value is stored in the microcomputer, and the calculation is performed on the detected error count, the error count will be calculated. By the second
The C / N ratio can be obtained accurately from the curve of bit error rate against the C / N ratio as shown in the figure, and by displaying the C / N value on the display, it is used for direction adjustment when the antenna is installed. It is possible.

Embodiment One embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of a satellite receiver C / N display device according to an embodiment of the present invention. In FIG. 1, reference numeral 1 is an input terminal to which the convolutional code used in digital data transmission for satellite communication is demodulated and detected, and 2 is input terminal 1.
It is a Viterbi decoder that Viterbi-decodes the convolutional code input to. Reference numerals 3 and 4 denote a state metric and an error counter built in the Viterbi decoder 2. The state metric 3 accumulates the Hamming distance when the input convolutional code is processed in the hard-decision Viterbi decoding algorithm. The counted number is counted and the state is measured.
The output of is further cumulatively added for a certain fixed time, and the number of errors within this fixed time is held. Reference numeral 5 is a microprocessor for detecting the error number by reading the output of the error number counter 4, and reference numeral 6 is a level indicator for displaying the error number by the display output according to the error number output from the microprocessor 5.

Here, the convolutional code used in the digital data transmission for satellite communication and the Viterbi decoding (and the algorithm) used in the decoding on the receiver side will be briefly described.
FIG. 3 shows the constraint length K (= 3), the coding rate (or also called the information amount) r (= 1/2), and the sign vector G ₁ = (1,
0,1), G2 = (1,1,1) convolutional encoders are shown. In FIG. 3, 7 shift register, 8 mod2 adder, 9 denotes a _{switch, u = (u 1, u} 2, u 3, u 4, u 5, ...) = (1,0,1 , 1,0,1,0,0,0). Further, 10 is an encoded information sequence output obtained through the switch 9, and X (u) = (X ₁ , X ₂ , X ₃ , X ₄ , X ₅ , ...) X _i = (X ₁ ^I , X ₁ ^Q ). Here, the shift register 7 is all "0"
It has been reset to.

Therefore, the coded symbol corresponding to the first four information bits is X ₁ = u ₁ G ₁ = 11 = (x ₁ ^I x ₁ ^Q ) X ₂ = U ₂ G ₁ u ₁ G ₂ = 01 = (x _{2 ^{I x 2 Q) X 3 =}} u 3 G 1 u 2 G 2 u 1 G 3 = 00 = (x 3 I x 3 Q) X 4 = u 4 G 1 u 3 G 2 u 2 G 3 = 10 = ( X ₄ ^I x ₄ ^Q , so that the encoded output is X (u) = (11,01,00,10 ...) where the convolutional encoder of FIG. FIG. 4 shows the state transitions of the above, which is the basis of the trellis diagrams of FIG. 6 and subsequent figures.

The above is the description of the convolutional code. Next, the principle of the Viterbi decoding and the actual Viterbi decoding of the convolutional code having the coding rate 1/2 and the constraint length K = 3 will be described.

A code sequence of q (q is an integer) original convolutional code C having a coding rate k / n (kn is an integer) is sent through a memoryless stationary communication channel, and a received sequence Y = y ₀ y ₁ y ₂ ... y _N-1 Is received, N is an arbitrarily large integer and the time-series block of the received sequence is Y
_{t 1} = (y _1t , y _2t , ... y _nt ), where y _1t is an element of the communication path, and the output of the communication path is a finite set or a set of all real or complex numbers. At this time, in order to perform the maximum likelihood decoding, the code sequence W = w ₀ w ₁ w ₂ ... w that maximizes the likelihood function (conditional probability or conditional probability density) P (Y | W) for the reception sequence Y. _It may be determined that _N-1 has been sent. Therefore, since the communication path is a memoryless steady communication path,
This likelihood function is And then take this logarithm and multiply by -1 Becomes Here, the reason for multiplying −1 is 0 ≦ P (Y _t
This is because | W _t ) ≦.

Therefore, for each block y _{t of the} reception sequence, −logP (y _t
Maximum likelihood decoding can be performed by obtaining | w _t ), making this sum for all possible code sequences, comparing the results, and finding the code sequence with the smallest sum.

Here, the amount is not limited to −logP (y _t | w _t ), but may be an amount proportional to this. For example, when the communication path is a binary symmetrical communication path, the Hamming distance dH of y _t and w _t
(Y _t , w _t ) can be used. Such an amount is called a branch metric and is represented by λ (y _t , w _t ). The branch metric is called because w _t is associated with the branch of the trellis diagram (FIGS. 5 and 6). Hereinafter, the branch and the code block associated with the branch may be equated with each other.

In the trellis diagram, the code sequence corresponds to the path starting from the initial state at time 0. Therefore, the reception sequence Y = y ₀
y ₁ y ₂ ... y _N−1 and a code sequence W = w ₀ w ₁ corresponding to a certain path
w ₂ ... w _N−1 , sum of branch metrics λ (y ₀ , w ₀ ) +
... + λ (y _N−1 , w _N−1 ) is called a path metric or metric (state metric) of this path with respect to the reception sequence y, and is represented by λ (Y, W). Especially when the branch metric is the basic branch metric, it is called the basic path metric. Ultimately, the maximum likelihood decoding can be performed by obtaining the path (maximum likelihood path) having the smallest metric in the trellis diagram.

Here, in the trellis diagram as shown in FIG.
Considering a path that passes through the state Si and reaches the state Sf at the time point N-1, it is assumed that the path indicated by the solid line has the largest metric among the paths satisfying such conditions. The solid line path at this time starts from the initial state and is in the state at time t.
Of all the paths to Si, the metric up to time t must be the smallest.

If the metric up to the time point t of the broken line path is smaller than that of the solid line path, the metric of the path that becomes a solid line after that until the time point t becomes smaller than the metric of the solid line path only. Because. Therefore, when the path reaching the state Sf at the time point N-1 is determined to have the smallest metric, only the path having the smallest metric up to the time point t is selected among the paths passing through the state Si at the time point t, and the others are You can throw it away. That is, at each time point, among the paths reaching the respective states, only the one having the smallest metric up to that time point may be retained and the others may be discarded. In this way, the path with the smallest metric among all paths can be obtained.

Next, with a convolutional code with a coding rate of 1/2 and a constraint length K = 3, the transmission data is 101101000, and the transmission sequence is 11 01 00 10 1
0 00 01 11 00, one error occurs in the reception sequence, and the reception sequence becomes 11 01 00 100 0 00 01 11 00
The Viterbi decoding in the case of is described (○ indicates the location of the error).

In FIG. 6, consider the case where the above reception sequence is input from the all-zero state.

In FIG. 7, () indicates transmitted data (information), and y _1t y _2t (00,01,10,11) next to it indicates the output of the convolutional code. Further, ~ indicates a state metric (path metric), and indicates the sum of Hamming distances up to time tn.

First, at time t ₁ , the reception sequence is 11, and on the trellis line,
Since (0) is 00 and (1) is 11, the Hamming distance is 2 in the state S ₀ , and the Hamming distance is ₀ in the state S ₁ .

Next, at time t _2, a received sequence is 01, the state S at ₀ (0) 00 Hamming distance 1 for a is and adds its previous Hamming distance, the point in time t ₂ S _{It is 0} . Next, in the state S _{1 at} t _2, the hamming distance is 1 from (1) 11 and is added to the hamming distance 2 at the time t ₁ immediately before the path. Also at time t ₂
In the state S _{2 of} , the Hamming distance is 0 due to (0) 01, and the Hamming distance at the time of the previous pass is 0. In the state of S ₃ at time t _2, the Hamming distance is 2 because of (1) 10 and the Hamming distance at the time of the previous pass is 0. Therefore, at time t ₂ , state S ₀
At time t ₃ because that could pass each of S ₁ S ₂ S _3, in comparison Hamming distance has accumulated as shown in FIG survive small path, a large path of Hamming distance is discarded. The solid line is the surviving path, and the stale path is shown by the dashed line.

After proceeds to point t ₄ ~t _9, showing the state of the at each time point in FIG. 8 to 13 FIG.

Here, since the accumulated Hamming distances are the same in the state S ₃ at the time point t ₃ , and it is sufficient to select them at random, the paths coming from the upper side are left in this embodiment.

Further, referring to the state S ₀ at t ₉ in FIG. 13, the accumulated Hamming distance is the minimum, and the state S ₀ at t ₉ is. This indicates that only one error has occurred. The above is the description of Viterbi decoding.

At this time, the path on the trellis line that has entered the S ₀ state at t ₉ indicates that one error has occurred, so in the process of performing Viterbi decoding, the accumulated Hamming distance on this selected path is set for a certain period of time. By accumulating again during the period, the number of errors from the start of Viterbi decoding to a certain time t _k can be known.

Next, the operation will be described with reference to FIG. The received convolutional code passes through the input terminal 1 and the Viterbi decoder 2
The state metric 3 counts the accumulated convolutional code and the accumulated Hamming distance shown in the description of the Viterbi decoding described above, and performs the state metric. The output is further accumulated (accumulated addition) by the error number counter 4. At this time, within a certain fixed time (if the number of transmission bits is n _bits / second, nt is given at a certain fixed time t).
The error number counter 4 holds the number of errors in each piece of data). The microprocessor 5 reads the output of the error number counter 4 by the microprocessor 5.
Detects the number of errors, and the microprocessor 5 further outputs a display output corresponding to the number of errors to the level indicator 6 so that the number of errors can be displayed.

Furthermore, if the C / N value is measured in advance and the correlation with the number of errors at that time as shown in FIG. 2 is input to the microprocessor 5, the level indicator 6 will show it as shown in FIG. The C / N value can be displayed by using the relationship described above.

As described above, according to the present invention, the number of errors is detected by holding (cumulative addition) the value of the state metric used in the Viterbi decoding process for a certain period of time, and the number of errors is counted in advance. Also, the error count number is measured in advance by changing the actual C / N, correction data is created based on this measurement data, and if it is input to the microcomputer, the C / N value will be based on the above error count number. Can be converted, an accurate C / N value can be obtained, the C / N value of the receiver can be displayed, and it can be used as the direction adjustment of the satellite receiving antenna and as an image quality display. Also, since most of the circuits are digital processing, they can be built into the LSI, and C / N display is possible without increasing the cost, which has a great practical effect.

[Brief description of drawings]

FIG. 1 is a block diagram of a satellite receiver C / N display device according to an embodiment of the present invention, FIG. 2 is a characteristic diagram showing BER (bit error number) to C / N ratio, and FIG. 3 is a convolutional code. FIG. 4 is a block diagram showing an example of an encoder, FIG. 4 is a diagram showing a state transition of the convolutional encoder,
FIG. 5 is a diagram explaining the maximum likelihood path of the trellis diagram, and FIGS. 6 to 13 are explanatory diagrams of the trellis diagram of the convolutional encoder of FIG. 1 ... Reception convolutional code input terminal of reception sequence, 2 ... Viterbi decoder, 3 ... State metric (path metric) device, 4 ...
… Error number counter, 5 …… Microprocessor, 6 ……
Level indicator.

Claims (1)

[Claims] 1. Viterbi decoding used when decoding a convolutional code used in satellite communication digital data transmission at a receiver side, when the convolutional code is processed in a hard-decision Viterbi decoding algorithm. A means for detecting the number of errors by a state metric (path metric) based on the Hamming distance calculation, a means for counting the number of detected errors, and a C / N (carrier level /
Input noise level) vs. BER (bit error rate) curve is stored, a microcomputer that calculates and outputs a C / N value for the count number of the error number based on this curve, and a C output from this microcomputer A satellite receiver C / N display device equipped with a display for displaying the / N value.