KR0171012B1 - Method of measuring symbol error rate of multiple modulation transceiving system - Google Patents

Abstract

The present invention is a method for generating a transmission pattern using a DSP board of a 16 quadrature amplituded modulation (QAM) system and receiving a waveform-formed signal thereof from another DSP board to obtain a symbol error rate of a selected symbol at an appropriate sampling point in real time. A signal processing means of any of a plurality of signal processing means provided in a quadrature amplitude modulation system of a predetermined bit generates a transmission pattern, and determines which symbol to transmit at which time among 16 symbols of 16 QAM, and transmits the transmission pattern. A first method of applying a pattern of PN to symbols and having all frequency characteristics within a predetermined band to store data formed in a table declared as Unsigned Long, and converting the stored data into an analog signal for transmission. Process and extract the appropriate sampling point of the waveform-generated signal being transmitted. And a second step of receiving symbols sampled by an algorithm in another arbitrary signal processing means, and a third step of calculating symbol error rates of the symbols selected at the set point of the received waveform signal in real time to match symbol synchronization. Symbol error rate measuring method in a multi-stage modulation and demodulation system.

Description

Symbol Error Rate Measurement in Multistage Modulation Demodulation Transceiver System

1 is an illustration of the configuration and hard decision boundary of the 16QAM symbol.

2 is a diagram illustrating a table structure and a symbol extraction method of a 32-bit word unsigned long type of a 16QAM symbol.

3 is a flowchart of symbol error rate measurement.

4 is a diagram illustrating a structure of a pseudo noise generator using a 15 TAB shift register.

5 is a diagram illustrating a configuration of a transmitter.

6 is an exemplary configuration of a receiver.

The present invention generates a transmission pattern using a digital signal processing (DSP) board of a 16 quadrature amplituded modulation (QAM) system and receives a waveform-formed signal thereof from another DSP board to receive the symbol error rate of a symbol selected at an appropriate sampling point in real time. How to get it.

In general, the ultimate performance of the collected communication system can be evaluated as the bit error rate or symbol error rate for the signal-to-noise ratio of power over a constant frequency bandwidth. In order to measure the symbol error rate in real time, the transmission pattern at the transmitting end must be known at the receiving end. The number of error symbols with respect to the total number of transmission / reception symbols should be measured by comparing them.

At this time, in order to maintain objectivity, there is a difficulty in that the transmission pattern should be measured using a signal generated with a maximum length of pseudo noise PN.

Exemplary conventional approaches to address this difficulty are well illustrated in the following references.

[1] R. L. Cupo and R. D. Gitlin, Adaptive carrier recovery systems for digital data communications receivers, IEE J-SAC Vol 7 No 9 Dec. 1989 pp 1328-1339

[2] W. C. Lindsay and M. K. Simon, Carrier synchronization and detection of polyphase signals, IEEE Trans. Comm. June 1972 pp441-454

However, the conventional methods introduced in the above reference have a pseudo noise generator (PN Generator) at the transmitting end and the receiving end, synchronize them, and measure the symbol error rate, thereby making it difficult to measure the troublesomeness and reliability of the measurement. Problems have arisen that the configuration of the system is complicated by using a problem noise method or by using a pseudo noise generator after presymbols are placed in front of the transmission symbols to adjust symbol synchronization.

An object of the present invention for solving the above problems is the generation of a transmission pattern using an arbitrary DSP board of a 16-bit quadrature amplitude modulation system and the waveform-formed signal thereof received from another DSP board to select a symbol selected at an appropriate sampling point. The present invention provides a symbol error rate measuring method in a multi-stage demodulation transceiver system for obtaining symbol error rates in real time.

In order to achieve the above object, the present invention generates a transmission pattern by any one of a plurality of signal processing means provided in a quadrature amplitude modulation system of a predetermined bit, and at any time of 16 symbols of 16 QAM. It decides which symbol to transmit and stores the data formed by applying the pattern of PN to all the frequency characteristics in a predetermined band in a table declared as Unsigned Long. Converts the signal into an analog signal, transmits the signal, and receives the sampled symbols from other arbitrary signal processing means through an algorithm that extracts the proper sampling point of the generated waveform signal. The symbol synchronization rate of the symbol is calculated in real time to match the symbol synchronization. It is done.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a diagram illustrating the configuration of 16 QAM symbols and hard decision boundaries. FIG. 2 illustrates a table configuration and a symbol extraction method of a 32-bit word unsigned long type of 16 QAM symbols.

3 is a flowchart of symbol error rate measurement.

4 is a diagram illustrating a structure of a pseudo noise generator using a 15 TAB shift register, FIG. 5 is a diagram illustrating a configuration of a transmitter, and FIG. 6 is a diagram illustrating a configuration of a receiver.

Since the following description should be divided into a transmitter and a receiver, the transmitter is first described, and the transmitter should determine which symbol to transmit at which time among 16 symbols of 16 QAM.

At this time, the format of the transmitted pattern should use a pseudo noise pattern to ensure that the symbols have all frequency characteristics within a predetermined band to ensure objectivity.

In this case, the method may be transmitted using a PN generator, or may be transmitted by leaving the pattern of the PN generator as a table.

In this system, a binary pseudonoise generator with a maximum length of 32767 with 15 shift registers is used to store the information on the symbol string of symbol period 8192 in a table declared as unsigned long to construct a transmission system. See also attached figure 4).

At this time, the table is designated in the form of an array of 1024 words having a size of 32 bits.

Each storage format of 1024 words is shown below.

Table [0] = 0xd19fa342 === {13,1,9,15,10,3,4,2}

It consists of.

That is, it has a format of storing eight signal point indexes of the signal constellation of FIG. 1 in the table [0] and [15], and the 32-bit word unsigned of 16 QAM symbols according to the format. An example of a table structure of a long type (word unsigned long type) and a symbol extraction method is shown in FIG.

A mapping signal of 16 QAMs of in-phase and quadrature during one symbol period is referred to as in-phase signal sig_i and quadrature-phase signal sig_q, respectively {sig_i [1], sig_q [One]},… … Generate signals of {sig_i [15], sig_q [15]}.

It is sent to a processor that performs waveform shaping, up-sampling, and then by a time interrupt to a digital / analog (D / A) converter.

As shown in FIG. 5 attached to a system in which the operation of the above-described series of transmitters is configured using a plurality of DSPs.

In this case, for the transmitting end operating as described above, the receiving end uses an algorithm that extracts the proper sampling point from the analog / digital (A / D) converter, so that the sampled symbols are perfectly synchronized and are not affected by noise. It must be equal to the value.

If the sampling synchronization is perfectly matched and there are no symbols added or lost, then the received symbol in Gaussian sampling is prepared as a symbol that can be compared with the pattern of the transmitter in the manner of hard decision as shown in FIG.

It should be considered here that the currently accepted symbol does not know which part of the 8192 symbols it is.

Therefore, the function of measuring the symbol error rate is to wait at the beginning of the 8192 symbol pattern, and if there is no error between the symbols to be compared, the next symbol pattern is compared with the next received symbol, and the point of the pattern is increased. You can find information about that starting point.

When the point of this increasing pattern reaches 8192, the point is set to 0 again and the symbol error rate is measured.

The problem with this is that if the symbol pattern is different from the received symbol at the first point of finding, the point goes back to the point of zero. This is caused by the influence of noise and reset again due to an error caused by hard decision. The problem is that it can be.

Therefore, the operation of returning the point of the symbol pattern back to 0 should be performed until the error does not occur to some extent in the comparison of the symbol pattern.

The configuration of the receiver for performing the above-described operation is shown in FIG. 6, and the detailed operation of the reception process will be described with reference to FIG. 3. In FIG. The total signal Total_S variable, which is the number of symbols, is increased, and the count variable value is checked in step 303.

In this case, when the count variable is '0', the value of the table is compared with the value of the input in step 305, and if it is the same, the index i of the table variable is increased, and in step 308, the index value is a duration variable (for example, , 8192).

If the index value is larger than the persistent variable, the index value is determined as '0' in step 309.

In step 310, the count variable value is increased, which means that the input value coincides with the table value.

Thereafter, when the steps 301 to 303 are performed and the step is reached to step 304, when the starting point of the table is synchronized with the input, the control unit proceeds to step 310 to increase the count variable value, and proceeds to step 301.

On the other hand, when the starting point of the table and the input is not synchronized in step 304, that is, the input value and the value of the table do not match, in step 311 the number of times that the count variable value, that is, the received symbol and the table value match If the value is larger than the threshold value, which is a threshold value, the operation returns to step 301 without any operation on the counter value.

In this case, it is appropriate to select about 10 as the arbitrary number of symbols.

At this time, since the value of the count variable does not increase but the total signal value increases, the value of the error is relatively reported to increase.

On the other hand, if the value of the count variable is smaller than the threshold value, the threshold value, rather than an error caused by noise, it is determined that the synchronization between the pseudo noise on the table and the received symbol is not synchronized to initialize all the variables to the next pseudo noise noise I will wait.

On the other hand, if the table value is different in step 305, the value of the count variable and the index value is set to '0' and starts from step 301.

Here, signal [2] [6] (sig [2] [16]) is a 16 QAM mapping value, and table [1024] is a symbol pattern table of a 32-bit word size shared with the above-mentioned transmitter. Value.

In the multi-stage demodulation transceiver system according to the present invention operating as described above to solve the problem that it is very difficult to know the information of an infinitely large transmission symbol due to the limitations of the memory when configuring the transceiver using digital signal processing. If the symbol error rate measurement method is provided, in case of 16 QAM, an index of one symbol information is included in 4 bits to store information in word units, and many symbols can be held in a comparison table.

In addition, since the maximum length of the pseudo noise generator is stored in the table, objectivity can be given to measuring the symbol error rate.

In addition, since the pseudo-noise code can be turned indefinitely, there is an advantage that even a small symbol error rate can be measured.

Claims (7)

A symbol error rate measuring method in a multi-stage modulation / modulation transmission / reception system, wherein any signal processing means of a plurality of signal processing means included in a quadrature amplitude modulation system of a predetermined bit generates a transmission pattern, and among 16 symbols of 16 QAM. Determining which symbol to transmit at which time, and applying the PN pattern to the transmitted symbols to have all frequency characteristics in a predetermined band, and store the data formed in an unsigned long table. A first step of converting and storing the stored data into an analog signal; A second step of receiving, by other arbitrary signal processing means, the sampled symbols through an algorithm for leaking an appropriate sampling point of the waveform-generated signal to be transmitted; And calculating a symbol error rate of a selected symbol at a set point in real time to match symbol synchronization with a received waveform signal. 2. The method of claim 1, wherein a data table storing a pattern of a PN generator is used to apply the PN pattern to the transmission symbol in the first step. The method of claim 1, wherein the storing of the data in the table declared as unsigned long in the first step comprises: using a binary pseudonoise generator having a predetermined number of shift registers for a symbol string of a predetermined symbol period. A method for measuring symbol error rate in a multi-stage demodulation transceiver system, comprising storing information in a table declared as an unsigned long. 4. The method of claim 3, wherein the number of shift registers is arbitrarily set to form a symbol sequence. 4. The method of claim 3, wherein the maximum length of the binary pseudo noise generator is continuously circulated. 4. The method of claim 3, wherein the symbol period of the symbol string is set to be continuously circulated. 2. The method of claim 1, wherein the third process comprises: a first step of accessing data for a synchronization point from a data table storing which of the 16 symbols of 16QAM at which time should be synchronized; A second step of comparing the data accessed in the step with symbols received and synchronously detected in the second step; A third step of searching for an error degree of synchronous matching of the data compared in said step; And a fourth step of matching synchronization of all received data symbols with reference to the data retrieved in the third step.