KR20020050213A - Method of Correlation Flattening for the Transmitter of Binary CDMA Communication Systems - Google Patents

Abstract

PURPOSE: A correlation flattening method of binary CDMA communication mode transmitting unit is provided to prevent degradation of performance of overall system caused due to a performance degradation of specific one or two channels and improve a BER as bit transmission performance of every channel is almost the same by flattening correlation values making an unbalanced distribution in a binary CDMA mode. CONSTITUTION: An output of an adder includes binarized 65 chip codewords. A correlation predicting unit calculates correlation values(W1, W2, ... Wk) for each channel of a receiving unit, for which the 64 chip codewords, transmission data bits and code generator information are all necessary. It is checked whether there is values smaller than a reference value among the correlation values. If correlation values of every channel are above the reference value, 64 chip codewords are transmitted as it is. If there is a correlation value(W) of some channel is smaller than the reference value, it is checked whether the correlation value is increased while changing a code for each of 64 chips. If the correlation chip of the channel is increased, it is checked whether a correlation value(W) of a channel with the next smaller correlation value(W) is reduced below the reference value. While changing values of 'I' from 1 to 64 for the entire 64 chips, if correction is necessary for the Ith chip, a correction command(G) is transmitted to a chip correcting unit.

Description

Binary CDMA Communication Method Correlation Flattening Method of Transmitter {Method of Correlation Flattening for the Transmitter of Binary CDMA Communication Systems}

The CDMA communication method is rooted in the direct spread spectrum spread method originally used in military communication. In the spread spectrum method, the transmitter multiplies the input data by a specific code (pattern) and transmits the data by transforming it into a form of random noise. In the spread spectrum communication method, data is restored only when the codes of the receiver and the transmitter are the same, and thus are widely used in digital mobile communication because they are robust against radio wave interference. One or more channels are allocated to each user while multiple users simultaneously use one broadband channel among the CDMA communication methods, and DS / S is a communication method in which each user's signal is spread by a unique code. Called CDMA.

In the DS / CDMA scheme, the voltage level of each channel signal is a binary value of (+1) and (-1), but the voltage level of a transmission signal in which signals of several channels are summed simultaneously has various values. When the number of users is K, the voltage level of the transmission signal is one of {K, K-2, ....,-(K-2), -K}. Since they must be transmitted, the linearity of the high frequency output amplifier becomes very important. Therefore, in the DS / CDMA scheme, an output amplifier having excellent linearity must be used in the base station, which increases system cost and significantly reduces power efficiency.

In addition, in the multi-carrier CDMA method used in the next generation mobile communication, since the terminal uses multiple basic channels of voice speed simultaneously, the high frequency amplifier of the terminal also requires linearity, thereby degrading the power efficiency of the amplifier.

Binary CDMA is a system that simplifies the entire system by transforming a multicarrier CDMA transmission signal into a simple binary signal. The binary CDMA transmitter is a binary unit (Truncation unit) added to the conventional DS / CDMA transmitter, and the receiver has the same structure as the DS / CDMA system. Signals of various levels of {K, K-2, .....,-(K-2), -K} of the transmitter are binarized to have only values of (+1) / (-1). Even if only the code is transmitted in this manner, when the spreading code is properly selected, the DS / CDMA performance can be approximated. Since the transmitted signal is a binary signal, the linearity of the power amplifier is not required, which simplifies the system. Binary CDMA method is inexpensive, but the advantages of DS / CDMA method such as user separation by code, multiple access and interference cancellation are maintained.

The code rate of spreading codes used in the spread spectrum method is tens to hundreds of times higher than the original data rate. When referring to these small bits, the term chip is used instead of the bits.

As such, since one bit of data is transmitted by being divided into several chips, the original data bit can be determined even if a transmission error occurs in several chips. For example, when one bit is transmitted to 64 chips in the DS / CDMA scheme, multiplying the received signal by a unique spreading code results in all 64 chips having (+1) or all (-1) values. Transmission noise of the chip may occur due to channel noise during transmission. When E _C transmission error occurs among N chips, if the original data value is (+1), the number of (+1) chips in the correlator output signal is (N-E _C ) and (-1) Since the number of chips is E _C , the sum of all of them adds Z, and the size of Z is (N-2E _C ). Therefore, when the value of Z is positive, the data is determined to be (+1), and on the contrary, the judgment is made by (-1).

In the binary CDMA scheme, due to the influence of the binarization process, a value obtained by multiplying a received signal by a unique spreading code does not have a value of (+1) or (-1) in N chips. However, since most of the N chips have the same value, the same data recovery method is used as in the DS / CDMA method. The performance indicator of the binary CDMA method simply uses the value of Z multiplied by the sign of the original data rather than the value of Z itself. When the Z value is positive at the receiver, the data is restored to (+1), and when the Z value is negative, the data is restored to (-1), so W is defined as follows.

W = (value of data) × (value of Z)

In the binary CDMA scheme, a large difference occurs in reception performance for each channel. The main reason is that the orthogonality between channels is broken by the binarization process. Loss of orthogonality makes signal separation difficult, reducing the processing gain, which is an advantage of CDMA. The correlation value and the magnitude of the W value in the CDMA system indicate the robustness to the channel noise. When the W value is large, the processing gain increases and the transmission performance is excellent. As the number of channels used, K, increases, the influence of the binarization process becomes more severe, and the value of W becomes smaller. An extreme example of decreasing the correlation value in the binary CDMA scheme is when the value of W has a negative value. If W has a negative value, the sign of the Z value and the sign of the original input data are reversed, and thus a recovery error of the data occurs at the receiver. This error is not eliminated no matter how large the SNR is.

Increasing the signal output increases the value of W as a whole, but in the case of binary CDMA, the value of W shows a severe imbalance for each channel, unlike DS / CDMA. In the DS / CDMA method, the correlation values of all channels are evenly distributed, and thus the reception performance of all channels is equally affected by the size of the SNR. Therefore, the reception performance is not good or bad only for a certain channel.

In contrast, in binary CDMA, some channels have a high correlation value, W, which is more robust than necessary for a very large level of noise, and some channels have very low correlation values, resulting in a drop in transmission performance. Occurs, the overall performance of the system depends on some low performance channels. Therefore, in order to improve the performance of binary CDMA, it is essential to design the distribution of correlation values to be uneven and even.

Accordingly, the present invention smoothes the correlation value in the binary CDMA receiver to have an even distribution, prevents the W value from decreasing, and improves the BER performance of the binary CDMA communication method, thereby achieving a level similar to that of the DS / CDMA method. It is an object of the present invention to provide a design method of a correlation smoothing module of a binary CDMA transmitter.

In the CDMA scheme, the length of a spreading code to multiply one bit of data, and the value of N, is usually a power of 2, such as 32, 64, 128, ..., but for convenience, the value of N is set to 64 for convenience. . Table 1 shows the values of W in the DS / CDMA scheme when data is simultaneously transmitted on 13 channels of one broadband channel. If there is no transmission error, W has a value of 64 on all channels. The value of W becomes (64 -2E _C ) if E _C chip transmission error occurs among 64 chips, and if the number of chip errors does not exceed 32, W becomes positive and data is correctly restored.

Table 1

In the case of binary CDMA, as shown in Table 2, the value of W decreases due to the binarization process, and generally has a value smaller than 64. The degree to which the value of W decreases is random, and as the number of channels increases, the degree to which W decreases becomes more severe. If the value of W decreases to 0, even if there is only one chip error, W can easily become negative and cause a bit error.

TABLE 2

Table 3 shows an example of including a self error in the same binary CDMA scheme as shown above. Due to the binarization, a bit error occurs in the twelfth channel where the value of W is negative. In this case, the BER is not improved even when the SNR is large. If such sign inversion occurs at a rate of, for example, 1%, the BER is less likely to have a value lower than 10 ⁻² .

TABLE 3

Table 4 shows, as the probability of occurrence, that the W value decreases due to the influence of binarization when the number of channels varies from 3 to 17 in a binary CDMA system. As the number of channels increases, the mean value of W decreases, and a phenomenon occurs that W becomes negative from the number of channels 13.

Table 4

As shown in Table 4, for a given number of channels, the average of the W values of all channels is constant. Thus, if the W value of one channel becomes too small, the W value of the other channels increases by that smaller value. In the example of Table 2 and Table 3, the average of W in all 13 channels is equal to 14.4. In the case of Table 2, the value of W is evenly distributed around the mean value, but in Table 3, the value of W in the 12th channel is negative, and instead, the value of W in the other channel is larger than that. To achieve. Since the performance of the system as a whole depends on the channels with small values of W, the BER performance of a binary CDMA system can be improved if the unbalanced distribution of W as shown in Table 3 can be changed to a uniform distribution as shown in Table 2.

Accordingly, the present invention provides a method of designing a correlation flattening module of correlation values that can improve BER performance in addition to the existing binary CDMA system. The correlation smoothing module replaces the binarization part of the conventional binary CDMA, and the transmission signal passing through the module has a binary value of (+1 / (-1) as in the case of the conventional binary CDMA. Although the structure of the receiver does not need to be changed even when used in the conventional binary CDMA scheme, the correlation of the channels and distribution of W may be very unbalanced, but when the smoothing module is used, the distribution of W of each channel is uniformly distributed at the receiver. It has a performance similar to that of DS / CDMA.

As described above, an object of the present invention is to provide a design method in which the value of BER becomes similar to that of DS / CDMA by improving the performance of the binary CDMA method which is inexpensive and retains the advantages of the existing DS / CDMA method.

1 is a transmitter of a conventional DS / CDMA communication system

2 is a transmitter of a conventional binary CDMA communication system

3 is a transmitter of a conventional binary CDMA communication system

4 is a transmitter of a binary CDMA communication system using a correlation smoothing module

5 is a receiver of a binary CDMA communication system using a correlation smoothing module.

6 is a system diagram of a correlation smoothing module

7 is a detailed diagram of an input signal and an output signal of the correlation predictor;

8 is a chip code inversion logic of the correlation predictor.

The present invention improves the unbalanced distribution of the correlation value W of the receiver of the conventional binary CDMA, smoothes the correlation so that the value of W has an even distribution, so that the performance of the binary CDMA is similar to that of the DS / CDMA. It is to. The overall principle of the present invention is described below.

The correlation smoothing technique is performed by changing the code of the binary transmission signal of the transmitter in units of chips. Each time the inversion of the code of any of the N binarized chips of the transmitter is inverted (a chip with a value of +1 is -1 and a chip with a value of -1 is +1), each channel at the receiver is W values increase or decrease by two. At this time, whether the W value of the channel increases or decreases can be calculated in advance before signal transmission. Therefore, by inverting the sign, the values of the channels that have a larger W value than necessary are reduced, and only the chips that increase the values of the channels of the W value that are too small are inverted and the signs of the chips are reversed. It can be kept above a certain value.

For example, it demonstrates in more detail. In case of the 12th channel of Table 3, the original data is (+1), but the value of W at the receiver becomes negative of (-2), which is the smallest value among the 13 channels. Now change the sign for every 64 chips to see if the 12th channel's W increases. When the sign of the first chip is inverted, the W value of each channel is improved in Table 5. The W value of the 12th channel is increased by 2 from (-2), and the size becomes 0. . The W values of the other channels also change, but if the original W value is large enough, changing W by a magnitude of 2 does not affect performance. In the case of Table 6, the first chip is not changed and the sign of the second chip is inverted, but the W value of the 12th channel is rather decreased by 2, which is worsened by (-4).

Table 5

Table 6

Now it is clear which chips to choose. That is, chips should be selected that increase too small W values of the channels and decrease W values that are larger than necessary. At this time, it is checked whether the W values of the other channels do not decrease below a certain reference value. By changing the sign of the chips that meet this condition among the 64 chips, a correlation smoothed transmission signal is obtained. Of the 64 chips in total, six of them met the {1 6 8 9 14 22} th chip. By inverting the codes of these chips, the following results were obtained.

TABLE 7

In the above result, it can be seen that the W value of the 12th channel having its own error has increased to 10, and the W values of the remaining channels all have a value greater than 10. Compared to before applying the correlation smoothing, it can be seen that the size of the W values larger than necessary (16 or more) was reduced to near the average value, and the W values of the small size were larger. As a result, the self-error of which the value of W becomes negative disappears completely, and the occurrence of bit error can be suppressed as a whole. By using the correlation smoothing technique, BER performance is improved.

The following is a brief summary.

The module constituting the present invention is composed of a correlation look-ahead and a chip correction unit. The correlation predictor receives the binarized signal as an input and calculates the correlation value to be generated at the receiver. The chip correction unit determines whether the distribution of the correlation value is improved when the sign of the chip is inverted for each chip, and changes the sign of the chip when the condition is satisfied.

Now, with reference to the accompanying drawings in detail as follows.

1 is a structure of a transmitter of a conventional DS / CDMA communication method, in which an input signal {d ₁ , d ₂ ..., d _k } is a multiplier with a spread code {c ₁ , c ₂ , ...., c _k }, respectively. Are multiplied by and summed in summer {1}. The value of the multiplier input signal {d ₁ , d ₂ ..., d _k } and the spreading code {c ₁ , c ₂ , ..., c _k } are each (+1) or (-1 Since the product also has a value of (+1) or (-1), the DS / CDMA signal sum of the K spread signals is {K, K-2, ...,-(K- 2),-K} has a multilevel value with one of

FIG. 2 is a structure in which a binary CDMA transmitter is added to the structure of the DS / CDMA transmitter shown in FIG. The binarization unit {2} outputs (+1) when the value of the input signal has a positive value, and outputs (-1) when the input signal has a negative value. The DS / CDMA signal, which is the sum of the K spread signals, has a {1} multi-level value. After passing through the binarization part, only the sign remains, and when the sum signal has a positive value, it is (+1). Is transmitted in binary form as (-1). The waveform of the binarization value is shown in {3}

3 shows a receiver of a conventional binary CDMA communication system and has the same structure as the receiver of a DS / CDMA communication system. It has a binary value of (+1) / (-1) after demodulation of the RF signal coming through the antenna. Since this binary signal has the sum of the signals of several channels, it is multiplied by the spreading codes {c ₁ , c ₂ , ..., c _k } and chip by multiplier in order to separate them, and then the value for all N chips. Is summed in summer {2 _a , 2 _b , ..., 2 _k }. When the output of the summer has a positive value, the original data is determined to be (+1), and vice versa, (-1).

4 is a transmitter of a binary CDMA communication system using a correlation smoothing module. The transmitter structure replaces the conventional binary CDMA transmitter binarization unit shown in FIG. 2 with a correlation smoothing module. The input of the correlation module includes an output of a summer {1} and a transmission data bit {d ₁ , d ₂ ... , d _k } are all required. Just as the conventional binary CDMA transmission signal {X} has a binary value of (+1) / (-1), the output of the correlation smoothing module {X} has a binary value.

5 is a structure of a receiver of a binary CDMA communication system using a correlation smoothing module. It has the same structure as the conventional binary CDMA receiver shown in FIG. Therefore, even when the correlation smoothing module is used, the structure of the binary CDMA receiver is not changed.

Figure 6 shows the correlation smoothing module of the present invention. The output of the summer is binarized to form a codeword of 64 chips. The correlation predictor calculates a correlation value {W ₁ , W ₂ , ..., W _k } for each channel of the receiver. A 64 chip codeword and a transmit data bit {d ₁ , d ₂ .. You need both., d _k } and code generator information. Check whether any of the correlation values {W ₁ , W ₂ , ..., W _k } has a value smaller than the reference value. If the W value of all channels is above the reference value, 64 chip codeword is transmitted as it is.If there is a value of W of some channel smaller than the reference value, check whether the W of the channel increases by changing the sign for every 64 chips. do.

If the W of that channel increases, then verify that the W of the next smaller W channel does not decrease below the threshold. For all 64 chips, a correction command {G} is sent to the chip correction unit if correction is necessary for the I-th chip while changing the value of I from 1 to 64.

7 is a detailed diagram of an input signal and an output signal of the correlation predictor. From the K-bit transmission data, the code generator information of each channel, and the 64 chip codeword information, the correlation value of each channel, W, of the receiver is calculated, and chips suitable for the chip correction conditions described in FIGS. 6 and 8 are determined. When inverting the sign of, the value of the correction command {G} is set to (-1), otherwise the value of the correction command is set to (+1). The chip codeword is composed of 64 stages of shift registers, and the end value is fed back to the start stage while being shifted by the clock signal. When the value of the correction command {G} is (-1), the sign of the feedback terminal is input with the sign reversed. This process is repeated 64 times.

8 illustrates chip code inversion logic of a correlation predictor. Using K transmission input data and 64 chip codewords as inputs, predict the W value of the receiver when the I-th chip is code inverted, and then check whether the W value of the smallest channel increases / decreases. do. At this time, if the value of W increases, the sign of the I-th chip is reversed. If the value of W decreases, no correction is made. If this process is confirmed for all chips, the process is terminated. Otherwise, the above process is repeated.

Conventional Binary / CDMA communication can generally maintain similar performance to DS / CDMA, but due to the process of binarizing multilevel DS / CDMA signals, the correlation at the receiver and W values are unbalanced. It has an enemy distribution. In particular, when the value of W becomes negative, the BER representing the communication performance does not get better than a certain value (usually 10 ⁻² ).

When the correlation smoothing technique provided by the present invention is used, the correlation values that form an unbalanced distribution in the binary CDMA scheme are smoothed so that the correlation values of all channels have an even distribution.

As a result, the bit transmission performance of all channels is almost the same, and the performance of the entire system is prevented from being degraded by the performance degradation of one or two channels, thereby improving the BER.

Table 8 shows the improvement effect by the correlation smoothing technique. In the case of 13 channels, the conventional binary CDMA method does not lower the BER value to 10 ^-2 even if the SNR is increased, but improves to 10 ^-4 or less using the correlation smoothing technique. When 15 or 17 channels are used, negative values of W do not occur, but BER is remarkably improved even in this case. This is because the W values of the channels of the receiver have an even distribution. The improvement in BER value is remarkable when the SNR is above -6dB. In the case of correlation smoothing, the BER improves more than 10 times from 0.0048 to 10 times when the SNR is -6dB for 15 channels. In addition, when the SNR is -3dB in the conventional binary CDMA, the BER value of 0.0003 has no bit error when the correlation smoothing is applied. Similar effects can be achieved with 17 channels.

In conclusion, the present invention is designed to maintain the performance of the binary CDMA scheme as the overall performance of the DS / CDMA by binarizing the transmission signal of the DS / CDMA, while evenly distributing the correlation values of all channels in the receiver. . Correlation facilitation module is a structure that replaces the binarization unit of the existing Binary CDMA transmitter, and even using this module, the receiver of the Binary CDMA can be used as it is without changing the structure of the receiver.

Binary CDMA communication has been proposed recently, but research and technology development are actively underway. The advantages of DS / CDMA, such as multiple access and interference elimination, are inexpensive, but Binary CDMA is rapidly expanding its application field. However, the correlation smoothing method provided by the present invention is Binary CDMA. It provides stable BER performance regardless of the form of spreading codes in a communication system.

Table 8

Claims (1)

In the conventional binary CDMA communication method, the correlation value of the receiver is unbalanced to prevent the performance of the entire system from being degraded due to the performance degradation of one or two channels. In order to improve the BER of the entire system by smoothing the values so that the correlation values are evenly distributed, as shown in FIG. 4, the binarization part of the transmitter of the Binary CDMA communication method is used as an alternative. As shown in FIG. It can be used without changing the receiver of the communication method, and as shown in Fig. 6, the output of the adder, the K-bit transmission input data and the code generator information are received as inputs, and the correlation predictor correlates each channel of the receiver. Calculating a value and checking whether any of the calculated correlation values have a value smaller than the reference value. Checking whether the correlation value of the channel is increased by changing the codes one by one for all chips of a spreading code when the correlation value of the system and some channels is smaller than the reference value, and the correlation value of the channel. If this increases, confirming that the correlation value of the channel having the next smallest correlation value remains above the reference value, and selecting chips suitable for the chip sign inversion condition as shown in FIG. 8 to invert the codes of the chips. Correlation Smoothing Module Design Method for Transmitter of Binary CDMA Communication System