JP4606993B2 - Demodulation method in multi-carrier code division multiplex transmission and receiver using the demodulation method - Google Patents

Description

  The present invention relates to a method and a receiver for reducing the amount of calculation at the time of demodulation in a multicarrier code division multiplex transmission system.

  As a communication method adopted in the new generation mobile communication system, a multi-carrier transmission method has attracted attention. Representative examples of the multicarrier transmission scheme include an OFDM (Orthogonal Frequency Division Multiplexing) scheme and an MC-CDM (Multi-Carrier-Code Division Multiplexing) scheme. According to MC-CDM, modulation symbols are spread over a plurality of subcarriers, multiplexed, and transmitted, whereby frequency diversity can be obtained and inter-cell interference can be made uniform.

  However, since intersymbol interference occurs in the frequency selective transmission path, it has been reported that the signal-to-noise power ratio and the interference energy ratio after despreading deteriorate (see Non-Patent Document 1). As a countermeasure, multi-dimensional demodulator that demodulates a signal without despreading has been studied (see Non-Patent Document 2).

  FIG. 6 shows a block diagram of a conventional MC-CDM communication system that employs a multidimensional demodulation method. As shown in the figure, this communication system includes a transmitter 5100 and a receiver 5400. The transmitter 5100 includes a transmission line encoder 5101, a puncture processing unit 5102, a bit interleaver 5103, a modulator 5104, a serial / parallel converter 5105, a spreader 5106, an inverse Fourier transformer 5107, and a parallel processing unit for information bits. / Serial converter 5108, cyclic prefix inserter 5109, and as a pilot symbol processing system, serial / parallel converter 5110, inverse Fourier transformer 5111, parallel / serial converter 5112, cyclic prefix insertion And a time multiplexer 5114 for time-multiplexing the output signal of the information bit processing system and the output signal of the pilot symbol processing system. The output signal of the transmitter 5100 is received by the receiver 5400 through the frequency selective transmission line 5200 and the additive white Gaussian noise transmission line 5300.

  The receiver 5400 includes a time demultiplexing unit 5401 that separates a received signal into a data signal and a pilot signal. As a data signal processing system, a cyclic prefix removing unit 5402, a serial / parallel converter 5403, a Fourier A converter 5404, an equalizer 5405, a parallel / serial converter 5406, a maximum likelihood estimation symbol generator 5407, a multidimensional demodulator 5408, and a decoder 5409, and a cyclic prefix as a pilot signal processing system A remover 5410, a serial / parallel converter 5411, a Fourier transformer 5412, and a transmission path estimator 5413 are provided.

Next, the operation of the above-described MC-CDM transmission / reception system will be described.
Here, for simplicity of explanation, it is assumed that all information bits can be transmitted by one MC-CDM symbol.
First, the transmitter 5100 performs transmission path encoding processing, puncturing processing, and interleaving processing on information bits to be transmitted, and then modulates the information bits by the modulation unit 5104, and the n _m -th modulation symbol M _t. (N _m ) is obtained. Thereafter, for diffusion, modulation symbols are allocated to i _d th frequency band having a bandwidth equal to the band occupied by the spread code, serial / parallel conversion by a serial / parallel converter 5105. assigned to the i _d th frequency band

Is spread by a spreading unit 5106 using a different spreading code for each modulation symbol and multiplexed to form a data subcarrier.
Here, if the spreading factor is N _w , the number of code multiplexes is M _w , and the number of data subcarriers is N _d ,

The following relationship holds between

  In general, in MC-CDM, orthogonal codes such as Walsh codes are used as spreading codes, and the orthogonal codes have the following characteristics.

  Therefore, the correlation value between the same codes is the sequence length, and the correlation value between different codes is “0”. Thereafter, the data subcarrier is subjected to inverse Fourier transform by an inverse Fourier transformer 5107 to be converted into a time domain signal, parallel / serial converted by a parallel / serial converter 5108, and then cyclically inserted by a cyclic prefix inserter 5109. A click prefix is inserted and provided to the time multiplexer 5114.

On the other hand, the pilot symbol is subjected to various processes by a serial / parallel converter 5110, an inverse Fourier transformer 5111, a parallel / serial converter 5112, and a cyclic prefix inserter 5113, and is provided to a time multiplexer 5114.
Finally, the data subcarrier is time-multiplexed with the pilot signal by the time multiplexing unit 5114 and transmitted. A signal transmitted from the transmitter 5100 is transmitted through a frequency selective transmission line 5200.

In addition, noise having a noise power density N ₀ is added to the additive white Gaussian noise transmission line 5300 and received by the receiver 5400.

  The signal received by the receiver 5400 is first divided into a pilot signal and a data signal by the time demultiplexing unit 5401, and the cyclic prefix is removed from each signal by the cyclic prefix removers 5402 and 5410. The signal from which the cyclic prefix has been removed is serial / parallel converted by serial / parallel converters 5403 and 5411 and then converted into pilot subcarriers and data subcarriers by Fourier transform in Fourier transformers 5404 and 5412. . Among these, the channel fluctuation is estimated from the pilot subcarrier by the channel estimator 5413, and the data subcarrier is equalized (phase rotation compensation) by the equalizer 5405 based on the estimation result. The data subcarrier after equalization is obtained from the following equation.

  Thereafter, the equalized data subcarrier is parallel / serial converted by a parallel / serial converter 5406 and converted to a maximum likelihood estimation symbol by a maximum likelihood estimation symbol generator 5407.

Finally, the received modulation symbol M _r (n _m ) is obtained by the processing of equation (7).

FIG. 7 shows the principle of maximum likelihood estimation symbol generation processing when the spreading factor is 2 in the above-described maximum likelihood estimation symbol generator 5407.
As described above, the channel value is obtained by performing multidimensional demodulation without despreading.
N. Miyazaki and T. Suzuki, “A Study on Forward Link Capacity in MC-CDMA Cellular System with MMSEC Receiver”, IEICE Trans. Commun., Vol. E88-B, No. 2, pp. 585-593, Feb. 2005. 3GPP TSG RAN WG1 # 42 bis, R1-051261, “Enhancement of Distributed Mode for Maximizing Frequency Diversity”, Oct. 2005

However, the amount of computation required for multi-dimensional demodulation increases as the number of reference signal points for measuring the square distance from the received signal point for maximum likelihood estimation increases with the power of the number of code multiplexes and the modulation order as indices. Have the problem of doing.
An object of the present invention is to provide a demodulation method capable of reducing the amount of calculation while maintaining good characteristics of multidimensional demodulation, and a receiver using the method.

  In the demodulation method according to the present invention, in a multicarrier code division multiplex transmission system, a channel value is determined by despreading for a subcarrier that does not cause characteristic deterioration due to orthogonality loss even if the channel estimation value is observed and despread. The channel value is obtained by multi-dimensional demodulation for subcarriers whose characteristics are deteriorated due to orthogonality destruction. This reduces the amount of computation while maintaining good characteristics of multidimensional demodulation.

  The receiver according to the present invention includes a separating unit (time demultiplexer 5401) for separating a received signal into a data signal and a pilot signal, and a converting unit (cyclic prefix removal) for converting the data signal into a frequency domain signal. 5402 to Fourier transformer 5404), first demodulating means (equalizer 5405 to multidimensional demodulator 5408) for demodulating the data signal converted by the converting means, and converting by the converting means A second demodulating means (equalizer 1200 to demodulator 1500) for despreading and demodulating the received data signal, connected between the converting means and the first and second demodulating means; Switch means (switch 1100) for selectively transferring the data signal converted by the means to either the first demodulation means or the second demodulation means; An estimation means (cyclic prefix remover 5410 to transmission path estimator 5413) for converting the pilot signal into a frequency domain signal and estimating the transmission path from the signal, and evaluating the transmission path based on the estimation result of the estimation means And an evaluation control means (transmission path evaluator 1800) for controlling the switch means based on the result of the evaluation, wherein the evaluation control means sets the number of subcarriers corresponding to the spreading factor (SF) as one set. The difference in received power of each subcarrier in the set is obtained, and it is determined whether or not the difference in received power is less than or equal to a predetermined threshold value. The switch means is controlled so as to transfer the processed data signal to the second demodulating means, and if it is larger than the predetermined threshold value, the data signal converted by the converting means is The switch means is controlled to be transferred to the first demodulating means.

  According to the present invention, in the multicarrier code division multiplexing transmission system, the despreading, which has a small amount of calculation but causes characteristic deterioration due to orthogonality loss, is limitedly used to subcarriers with little orthogonality loss. It can be expected to reduce the amount of calculation while suppressing deterioration. Furthermore, since wasteful processing is reduced, the processing delay time can be reduced by reducing the power consumption of the system at the time of mounting and reducing the number of clock cycles until the processing is completed.

FIG. 1 shows a configuration of a communication system of a multicarrier code division multiplex transmission system according to an embodiment of the present invention.
As shown in the figure, this communication system is composed of a transmitter 5100 and a receiver 1000. Among these, the transmitter 5100 is the same as the transmitter 5100 shown in FIG. The description regarding is omitted.

  The receiver 1000 according to the present embodiment has a switch 1100, an equalizer 1200, a despreader 1300, a parallel / serial converter 1400, a demodulator 1500, and a multiplexer as compared with the receiver 5400 shown in FIG. 1600, an MMSE equalization weight calculator 1700, and a transmission path evaluator 1800. Here, the switch 1100 is connected between the Fourier transformer 5404 and the equalizers 5405 and 1200, and the frequency domain data signal subjected to the Fourier transform is equalizer under the control of the transmission path evaluator 1800. 5405 and the equalizer 1200 are selectively transferred. The transmission path evaluator 1800 is for evaluating the transmission path based on the estimation result of the transmission path, and controls the switch 1100 based on the evaluation result to switch the transfer destination of the data signal.

The flow (demodulation method) of demodulation processing of the receiver 1000 in multicarrier code division multiplexing transmission is as follows.
A received signal received from transmitter 5100 is separated into a data signal and a pilot signal by time demultiplexer 5401. Among them, the data signal is processed by a cyclic prefix remover 5402, a serial / parallel converter 5403, and a Fourier transformer 5404, and the pilot signal is processed by a cyclic prefix remover 5410, a serial / parallel converter 5411, and a Fourier. Processed by the converter 5412. Thereafter, the transmission path estimator 5413 obtains a transmission path estimation value and a noise power density estimation value using the Fourier-transformed pilot signal. Among these, the channel estimation value is output to the channel evaluation unit 1800, the equalizer 5405, and the MMSE equalization weight calculator 1700, and the noise output density estimation value is output to the MMSE equalization weight calculator 1700.

  Transmission path evaluator 1800 determines whether or not characteristic degradation due to orthogonality loss will occur even if despreading each subcarrier set with SF (spreading factor) subcarriers as one set. If it is determined that there is no characteristic deterioration due to the loss of orthogonality, the switch 110 is controlled, and the subcarrier is transferred to the equalizer 1200 via the switch 1100, and is transferred to the equalizer 5405 otherwise. The data transferred to the equalizer 5405 is processed by the parallel / serial converter 5406 and the maximum likelihood estimation symbol generator 5407 in the same procedure as in FIG. Demodulated and the resulting channel value is input to multiplexer 1600.

  On the other hand, the data sent to the equalizer 1200 is equalized by the output of the MMSE equalization weight calculator 1700, then despreading by the despreader 1300, and parallel / serial by the parallel / serial converter 1400. After the conversion process, the channel value is obtained by the demodulator 1500, and this channel value is input to the multiplexer 1600. The multiplexer 1600 mixes the channel value obtained by the despreading and the channel value obtained by the multidimensional demodulation described above based on the evaluation information of the transmission channel evaluator 1800, and the data is decoded by the decoder 5409. To decode the original information bits.

Next, the operation of the transmission path evaluator 1800 will be described in detail with reference to FIG.
FIG. 2 shows the relationship between subcarrier frequency and power. The transmission path evaluator 1800 makes SF subcarriers as one set, and determines whether characteristic deterioration due to orthogonality loss will not occur even if each set is despread based on the difference in received power of the subcarriers. To do. The term “difference” in received power of subcarriers is a broad concept that includes “difference” and “ratio” in received power of subcarriers, and can express mathematically the difference in received power of each subcarrier. In this embodiment, “difference” will be described as an example.

The criteria for determining whether or not characteristic deterioration due to the above-described orthogonality breakdown will occur are as follows.
First, there is a determination criterion that focuses on the “difference” in received power between subcarriers. That is, as shown in FIG. 2, in one set of subcarriers, there is no disruption of orthogonality if the difference in received signal power is smaller than a predetermined threshold value a, and orthogonality disruption occurs if it is greater than the threshold value a. To do. Therefore, the first criterion is whether or not the difference in received signal power is smaller than a predetermined threshold a in one set of subcarriers.

  Second, there is a determination criterion that focuses on the value (amplitude) of the received power of each subcarrier. A set that includes SF subcarriers as one set, and the received signal power of the SF portion of the portion corresponding to the subcarrier includes a value whose received power value itself is lower than the predetermined threshold value b. If the channel value is obtained by despreading or the channel value is obtained by multidimensional demodulation, the orthogonality collapse occurs. Therefore, the second determination criterion includes a set of SF subcarriers, and one portion of the received signal power at the SF locations corresponding to the subcarriers is less than the predetermined threshold value b. Whether there is a pair.

The transmission path evaluator 1800 determines whether or not characteristic degradation due to the orthogonality breakdown occurs using either or both of the first and second determination criteria (conditions) described above.
Each of the above threshold values is appropriately set based on the difference or ratio with the noise power per subcarrier. That is, it is set appropriately according to the required communication quality. Although not particularly illustrated, the receiver 1000 includes measurement means for measuring the noise power.

Next, an example of the effect by this embodiment on the following measurement conditions is shown.
<Measurement conditions>
Diffusion rate: 2
Modulation method: QPSK
Transmission path model; 16-path quasi-static multipath Rayleigh model measurement conditions; threshold a = 0.1, 0.5, 1 (threshold b ignored)

  FIG. 3 shows the measurement conditions and the characteristics of “multidimensional demodulation” and “despread”. In the figure, the horizontal axis represents the signal-to-noise ratio (Es / No), and the vertical axis represents the packet error rate (PER). FIG. 4 shows how much of the subcarriers are demodulated by despreading under the above measurement conditions. In the figure, the horizontal axis represents the signal-to-noise ratio (Es / No), and the vertical axis represents the ratio of the despread subcarriers among all the subcarriers. Further, FIG. 5 shows the measurement conditions and the number of floating-point operations for “multidimensional demodulation” and “despread”. In the figure, the horizontal axis represents the signal-to-noise ratio (Es / No), and the vertical axis represents the number of floating point operations.

FIG. 5 shows that when the measurement condition is the threshold value a = 0.5, approximately 50% of the subcarriers are demodulated by despreading. Further, it can be seen from FIG. 6 that the amount of increase in the amount of calculation based on “despreading” is reduced by half compared to “multi-dimensional demodulation”. Furthermore, it can be seen from FIG. 4 that the characteristic deterioration is suppressed to about 0.05 dB on the basis of “multidimensional demodulation”.
Therefore, according to the present embodiment, it is possible to effectively reduce the amount of calculation while maintaining good characteristics of multidimensional demodulation.

Note that error correction codes applicable in the present invention are not particularly limited.
However, in the case of applying an error correction code capable of iterative decoding, the present invention further increases the effect of reducing the amount of calculation. Examples of error correction codes that can be iteratively decoded include turbo codes and low-density parity-check codes (LDPC codes). In addition, when a twin turbo decoder is used as a turbo code decoder, the turbo code decoding characteristics can be improved and the amount of calculation can be reduced. Regarding the twin-turbo decoder, for example, “T. Suzuki, N. Miyazaki, Y. Hatakawa,“ A Proposal of Twin Turbo Detector and Its Evaluation for M-QAM OFDM ”, Shingakutsuso, B-5-6, Sep. 2005 ".

1 is a block diagram of an MC-CDM communication system according to an embodiment of the present invention. It is a characteristic view which shows the reception power-frequency characteristic when the orthogonality collapse according to the embodiment of the present invention occurs and when it does not occur. It is a characteristic view which shows the change of the packet error rate by the calculation amount reduction by embodiment of this invention. It is a figure which shows the ratio of the subcarrier despread among all the subcarriers in each measurement condition which concerns on embodiment of this invention. It is a characteristic view which shows the difference in the amount of calculations in each measurement condition which concerns on embodiment of this invention. It is a block diagram of the conventional MC-CDM communication system using multidimensional demodulation. FIG. 5 is a principle diagram of maximum likelihood estimation symbol generation processing when the spreading factor is 2.

Explanation of symbols

1000 receiver 1100 switch 1200 equalizer 1300 despreader 1400 parallel / serial converter 1500 demodulator 1600 multiplexer 1700 MMSE equalization weight calculator 1800 transmission line evaluator 5100 transmitter 5200 frequency selective transmission line 5300 additive white Gauss Noise transmission path 5401 Time demultiplexer 5402, 5409 Cyclic prefix remover 5403, 5410 Serial / parallel converter 5404, 5411 Fourier transformer 5405 Equalizer 5406 Parallel / serial converter 5407 Maximum likelihood estimation symbol generator 5408 Multidimensional demodulator 5409 Decoder 5412 Transmission path estimator

Claims (3)

A demodulation method in multicarrier code division multiplexing transmission,
A set of SF subcarriers is set, a difference in received power of the subcarriers is obtained, and it is determined whether the difference in received power is equal to or less than a predetermined threshold. And a channel value is obtained by multi-dimensional demodulation if the channel value is greater than the predetermined threshold value.   A demodulation method in multicarrier code division multiplexing transmission, in which SF subcarriers are set as one set, and one of the received powers of SF locations corresponding to the subcarriers is less than a predetermined threshold. A demodulation method characterized in that a channel value is obtained by despreading for a set, and a channel value is obtained by multidimensional demodulation for the other sets. Separating means for separating the received signal into a data signal and a pilot signal;
Conversion means for converting the data signal into a frequency domain data signal;
First demodulation means for multidimensionally demodulating the data signal converted by the conversion means;
Second demodulation means for despreading and demodulating the data signal converted by the conversion means;
A switch connected between the converting means and the first and second demodulating means for selectively transferring a data signal converted by the converting means to either the first demodulating means or the second demodulating means. Means,
An estimation means for converting the pilot signal into a frequency domain signal and estimating a transmission path from the signal;
An evaluation control means for evaluating a transmission path based on the estimation result of the estimation means, and controlling the switch means based on the evaluation result;
The evaluation control means sets a number of subcarriers corresponding to the spreading factor (SF) as one set, obtains a difference in received power of each subcarrier in the set, and the difference in received power is less than a predetermined threshold value. If it is less than or equal to the predetermined threshold value, the switch means is controlled to transfer the data signal converted by the converting means to the second demodulating means, and is greater than the predetermined threshold value. For example, a receiver that controls the switch means so that the data signal converted by the conversion means is transferred to the first demodulation means.

Images