JP2005328391A - Multipath interference canceler and method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a multipath interference canceler and a method thereof which can obtain a satisfactory error rate characteristic, even when a delay wave caused by a multipath interference exceeds a guard interval (GI) length and reaches approximately one half the length of an OFDM (orthogonal frequency division multiplexing) symbol. <P>SOLUTION: The canceler comprises means for determining received symbols 401 and 402 for decoding and determining the symbols which have received multipath interference, a means for generating replicas 403 for generating replicas of desired waves based on the result of the determination made by the above means, a means for replacement 404 for replacing interferring parts which have received multipath interference among the received symbols with interferring parts which have received the interference among the replicas of the desired waves generated by the means of generating replicas, and means of determining the replaced symbols 405 and 406 for decoding and determining the received symbols whose multipath interferring parts have been replaced with the replicas of the desired waves by the means of replacement. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a multipath interference canceller used for a receiver adopting the MC-CDM method or the MC-CDMA method, which has been developed as a candidate for a fourth generation mobile communication system, and a method thereof.

Fujii, Sakurai, Sato, Nagato, Study of channel configuration in MC-CDMA system, IEICE Technical Report, RCS2001-309, pp. 171-178, 2001 N. Suzuki, H .; Uehara, M .; Yokoyama, A New OFDM Demodulation Method to Reduce Influencing of ISI Due to Longer Delay Than Guard Interval, IEEE ICCS 02, Session 1P03. Suzuki, Uehara, Yokoyama, Improvement of characteristics of variable symbol length OFDM demodulation system by soft ICI canceller, 2000 IEICE Communication Society B-5-23 JP 2003-188847 A

  Conventionally, as candidates for this kind of fourth generation mobile communication system, MC-CDMA (Multi-carrier Code Division Multiple Access) system, MC-CDM (Multi-carrier Code Division Multiplex) system, or these systems Various schemes such as an OFDM (Orthogonal Frequency Division Multiplexing) scheme, which is a modulation scheme in which each frequency spectrum profile to be combined has an orthogonal relationship, have already been proposed.

  Among these methods, Non-Patent Document 1 proposes an MC-CDMA system that reduces co-channel interference by adopting a CDMA method and reduces multipath interference by adopting an OFDM method. Yes. In the MC-CDMA system according to this proposal, pilot signals can be continuously transmitted on all subcarriers, so that highly accurate channel estimation is possible.

  An MC-CDM system that combines an OFDM system that can reduce multipath interference and a CDM system that spreads in the time domain and applies time division for multiple access is also one of the candidates for the fourth generation mobile communication system. As one.

  As described above, the mobile communication system to which the OFDM system is applied can reduce multipath interference by adding a guard interval to the OFDM symbol. It is a promising candidate.

  However, in the case of the OFDM scheme, for multipath interference in which a delay wave exceeding the guard interval (GI) length exists, inter-symbol interference and inter-carrier interference occur simultaneously, as shown in FIG. It is known that the error rate characteristic is significantly deteriorated.

  For such a problem, it is possible to prevent deterioration of the error rate characteristics by setting the guard interval (GI) length in accordance with the slowest delay wave that can occur in the actual transmission path. In this case, since it is necessary to set a long guard interval (GI) length to be added to the OFDM symbol, a new problem arises that the frequency utilization efficiency is lowered.

  Therefore, even if there is a delay wave exceeding the guard interval (GI) length while avoiding the problem of lowering the frequency utilization efficiency due to setting the guard interval (GI) length long, the error rate As a technique for preventing the characteristics from deteriorating, for example, as disclosed in Non-Patent Document 2, Non-Patent Document 3, and Patent Document 1, there is a multipath interference canceller that compensates for a delayed wave exceeding the guard interval length. It has already been proposed.

  In the technique disclosed in Patent Document 1, the effective symbol length is T, the frequency interval between adjacent N subcarriers is 1 / T, and NL (L <N) subcarriers are null carriers. In a multi-carrier demodulation method that receives a multi-carrier modulation signal and separates and demodulates each sub-carrier, the delay time difference of the delayed wave is estimated, and based on the estimated delay time difference, the waveform distortion caused by the delayed wave is A portion to be a use symbol having a length TM / N (L ≦ M <N) is determined from the effective symbol length T so as not to be included, and the use symbol portion is determined from a complex digital signal orthogonally demodulated at a sampling interval T / N. In this configuration, L subcarriers are separated and demodulated using M points.

However, the conventional technique has the following problems. That is, in the case of the techniques disclosed in Non-Patent Document 1, Non-Patent Document 2 and Patent Document 1, as shown in FIG. 20B, the delay time of the delayed wave is about ½ of the OFDM symbol length. Then, an error floor of about 10 ⁻² to 10 ⁻³ occurs in the error rate characteristics, and the delay time of about 3/8 of the OFDM symbol length is the application limit as shown in FIG. . Therefore, in the case of the techniques disclosed in Non-Patent Document 1, Non-Patent Document 2 and Patent Document 1, when the delay time of the delayed wave becomes about 1/2 of the OFDM symbol length, the error rate characteristic reaches the limit. However, it has a problem that it is difficult to apply to actual communication.

  Therefore, the present invention has been made to solve the above-described problems of the prior art, and the object of the present invention is that the delayed wave due to multipath interference exceeds the guard interval (GI) length and the OFDM symbol length. It is an object of the present invention to provide a multipath interference canceller capable of obtaining a good error rate characteristic even when it reaches about 1/2 of the above and a method thereof.

In order to solve the above problems, the invention described in claim 1 is
In the MC-CDM or MC-CDMA signal multipath interference canceller, received symbol determination means for demodulating and determining a received symbol including a received symbol that has received multipath interference;
Replica generating means for generating a replica of a desired wave based on the determination result of the received symbol determining means;
Replacement means for replacing an interference part that has received multipath interference in the received symbol with a part corresponding to an interference part that has received multipath interference among replicas of a desired wave generated by the replica generation means;
A replacement symbol determination unit that demodulates and determines a received symbol in which the interference part subjected to multipath interference by the replacement unit is replaced with a replica of a desired wave;
A multipath interference canceller.

In addition, the invention described in claim 2
The received symbol determination means includes:
A signal obtained by extracting and demodulating samples for the number of FFT points not including the guard interval in the desired wave;
A signal obtained by synthesizing a demodulated signal obtained by extracting samples for the number of FFT points including the guard interval in the interference wave,
The multipath interference canceller according to claim 1, wherein the determination is performed.

Furthermore, the invention described in claim 3
In the MC-CDM or MC-CDMA signal multipath interference canceller using phase modulation as the primary modulation,
After demodulating a received symbol that includes a received symbol that has undergone multipath interference, a phase error is estimated, and it is determined whether or not to make a determination based on a comparison result between the estimated phase error and a predetermined reference value. Received symbol determination means for performing determination;
Replica generating means for generating a replica of a desired wave based on the determination result of the received symbol determining means;
Replacement means for replacing an interference part that has received multipath interference in the received symbol with a part corresponding to an interference part that has received multipath interference among replicas of a desired wave generated by the replica generation means;
A replacement symbol determination unit that demodulates and determines a received symbol in which the interference part subjected to multipath interference by the replacement unit is replaced with a replica of a desired wave;
A multipath interference canceller.

  According to a fourth aspect of the present invention, the received symbol that has not been determined by the received symbol determining means feeds back the same amplitude value as the received symbol that has been determined. The described multipath interference canceller.

  Furthermore, the invention described in claim 5 reduces multipath interference by repeatedly performing the processing by the replica generation means, the replacement means, and the replacement symbol determination means. Is a multipath interference canceller.

In addition, the invention described in claim 6
In the MC-CDM or MC-CDMA signal multipath interference canceling method,
After generating a replica of the desired wave based on the result of demodulating and determining the received symbol containing the received symbol that has undergone multipath interference,
The multipath interference canceling method is characterized in that the process of demodulating and determining a received symbol obtained by replacing a part of the received symbol subjected to the multipath interference with a replica of a desired wave is repeated at least once.

  According to the present invention, the interference part that has received multipath interference in the received symbol is replaced with the part corresponding to the interference part that has received multipath interference in the replica of the desired wave generated by the replica generation means. By demodulating and determining the received symbol whose part has been replaced, the influence of multipath interference can be reduced to a sufficiently satisfactory level, and a good error rate characteristic can be obtained. For this reason, in the OFDM scheme, when the guard interval (GI) length is set to the same level as the conventional one, communication can be performed even when a delayed wave having a longer delay time exists. Also, in a communication environment where there is a delayed wave having a delay time comparable to that of the conventional case, the guard interval (GI) length can be set short, and the frequency utilization efficiency can be improved. According to the inventor's calculation simulation, as will be described later, the error rate characteristics can be sufficiently reduced even when the delay time of the delayed wave reaches about ½ of the OFDM symbol length as shown in FIG. all right.

  Embodiments of the present invention will be described below with reference to the drawings.

Embodiment 1
2 and 3 show schematic configurations of a transmitter and a receiver in an MC-CDM mobile communication system to which the multipath interference canceller according to Embodiment 1 of the present invention can be applied. More specifically, FIG. 2 is a block diagram showing a configuration of a transmitter in an MC-CDM mobile communication system spreading in the time domain. Here, the transmitter is assumed to be a transmitter deployed as a base station, and the receiver is assumed to be a receiver as a mobile station used by each user. The present invention mainly relates to the downlink (base station). To the mobile station). However, it is not limited to this.

  In FIG. 2, reference numeral 100 denotes a transmitter. The transmitter 100 mainly includes an S / P (Serial / Parallel) converter 101, a plurality of spreaders 102, and an inverse fast Fourier transform (IFFT: Inverse Fast Fourier Transform) 103 and a guard interval adder 104 are provided.

The S / P converter 101 receives a transmission symbol sequence 105 that has undergone primary modulation by a primary modulator (not shown). In the primary modulator, transmission data composed of digital information such as audio signals, characters, or image signals is subjected to primary modulation every 1 bit or every several bits, and a transmission symbol string 105 as a serial signal is generated. In the primary modulator, for example, QPSK (Quadrature Phase Shift Keying) modulation, which is a kind of phase modulation, is performed on transmission data. However, other types of modulation may be performed. good. The transmission symbol sequence 105 subjected to the primary modulation by the primary modulator is converted into subchannel signals 106 _{0 to} 106 _N-1 which are parallel signals by the S / P converter 102. The subchannel signals 106 _{0 to} 106 _N-1 converted into parallel signals by the serial / parallel converter 101 are spreaders 102 connected in correspondence with the subchannel signals 106 _{0 to} 106 _N-1 , respectively. After being spread in the time domain using a spreading code by _{0 to} 102 _N−1 , it is converted into a multicarrier modulation signal 107 by an inverse fast Fourier transformer 103. Furthermore, a guard interval (GI) signal is added to the multicarrier modulation signal 107 by the guard interval adder 104, whereby a transmission symbol 108 is generated. The transmission symbol 108 is transmitted from the transmission antenna through a filter (not shown).

  The guard interval adder 104 cuts out a part of the output of the inverse fast Fourier transformer 103, for example, 1/4 of the rear end, and performs the inverse fast Fourier transform on the guard interval (GI) signal thus cut out. The transmission symbol 108 is configured by adding it to the head of the output 107 of the device 103. Thus, for example, as shown in FIG. 4, the guard interval (GI) is set to 1/4 of the OFDM symbol length, and the total transmitted symbol length is 5/5 of the OFDM symbol length. 4 However, the guard interval (GI) length is not limited to ¼ of the OFDM symbol length, and may be set to other values.

  FIG. 3 is a block diagram showing a configuration of a receiver in an MC-CDM mobile communication system spreading in the time domain.

  The receiver 200 mainly includes a guard interval remover 201, a fast Fourier transform (FFT) unit 202, a 1 / H (t) multiplier 203, a plurality of despreaders 204, P / And an S (Parallel / Serial) converter 205.

In the receiver 200, a received signal 206 received by a receiving antenna (not shown) and inputted through a filter is input to the guard interval remover 201, and the guard interval (GI) is removed from the received signal 206 by the guard interval remover 201. . Thereafter, the received signal 206 from which the guard interval (GI) has been removed is obtained by the fast Fourier transformer 202 as received signals 207 _{0 to} 207 _N−1 for each subcarrier. The received signals 207 _{0 to} 207 _N−1 obtained for each subcarrier by the fast Fourier transformer 202 are multiplied by the inverse of the channel transfer function H (t) by the 1 / H (t) multiplier 203. Thus, channel compensation is performed. The obtained demodulated signal 208 for each subcarrier is despread in the time domain by the despreader 204 and then converted to a serial signal by the P / S converter 205 to obtain a received symbol sequence 209. The received symbol string 209 obtained in this way is determined to be 1/0 by a determiner, as will be described later.

  FIG. 5 schematically shows the arrangement of transmission symbols in the MC-CDM system that performs spreading in the time domain as described above.

  As shown in FIG. 5, the transmission symbol 300 in the MC-CDM system described above is multiplexed using the OFDM method, and is configured by a large number of subcarriers 301 that are orthogonal to each other. The number N of subcarriers 301 is set to 1024, for example, but is not limited to this. Here, the bandwidth occupied by the N subcarriers 301 is, for example, 40.96 MHz.

  Each subcarrier signal 301 is spread in the time domain by the spreader 102, and the spreading factor is set to 16, for example. As a spreading code used in the spreader 102, for example, by using an orthogonal code, it is possible to transmit signals multiplexed on the same subcarrier. In the MC-CDM system, the same spreading code can be used for each subcarrier.

  Further, pilot symbols 302 are spread in the time axis direction together with data symbols 303 and are code-multiplexed. Here, the pilot signal 302 is spread with, for example, a code of all “1”, and the time waveform of the pilot signal 302 is set equal in all OFDM symbols that are continuous in the time axis.

  Then, each carrier signal spread in the time axis direction is subjected to inverse fast Fourier transform by the inverse fast Fourier transformer 103 and superimposed to generate a baseband OFDM signal.

  By the way, the configuration of the transceiver of the MC-CDM system spreading in the time domain is basically the same as the configuration of the transceiver of the MC-CDMA system spreading in the time domain, but the multiple access scheme is MC-CDMA. Different from the system. That is, in the MC-CDMA system, users are multiple-accessed by code division, whereas in the MC-CDM system, users are multiple-accessed by time division. As described above, the present invention can be similarly applied not only to the MC-CDM system but also to the MC-CDMA system except that the multiple access method is different.

  FIG. 6 shows a multiple access method in the MC-CDM system.

  A plurality of users 1, 2, 3,... In the cell that is a range to be handled by one base station are time-divided and multiple-accessed as shown in FIG.

  FIG. 1 is a block diagram showing an MC-CDM signal multipath interference canceller according to Embodiment 1 of the present invention.

  By the way, in the present embodiment, in the MC-CDM signal multipath interference canceller, received symbol determination means for demodulating and determining received symbols including received symbols that have received multipath interference, and the received symbol determination means Replica generating means for generating a replica of a desired wave based on the determination result of the above, and interfering portions of the received symbol that have received multipath interference from the replica of the desired wave generated by the replica generating means. Replacement means for replacing with a portion corresponding to the received interference portion, and replacement symbol determination means for demodulating and determining a received symbol in which the interference portion subjected to multipath interference by the replacement means is replaced with a replica of the desired wave. It is configured as follows.

  The present invention provides a multipath interference canceller for delayed waves exceeding the guard interval (GI) length.

  FIG. 7 is a schematic diagram showing multipath interference due to a delayed wave exceeding the guard interval (GI) length. In FIG. 7, the received samples in the area indicated by the oblique line of the preceding wave have received interference from the adjacent OFDM symbols of the delayed wave. The multipath interference canceller according to the present invention reduces the multipath interference by substituting the reproduced signal by determining and returning the signal that has received the multipath interference.

  FIG. 1 is a block diagram showing an MC-CDM signal multipath interference canceller according to Embodiment 1 of the present invention.

  The multipath interference canceller 400 is roughly divided into an initial determination MC-CDM demodulator 401 and initial determination unit 402 as received symbol determination means, a replica generator 403 as replica generation means, and a replica as replacement means. And a MC-CDM demodulator 405 and a determination unit 406 for iterative calculation as replacement symbol determination means. Reference numeral 407 denotes a secretor that switches the determination result of the initial determination unit 402 and the determination result of the determination unit 406.

  The initial determination MC-CDM demodulator 401 and initial determination unit 402 are basically configured in the same manner as the receiver shown in FIG. As shown in FIG. 8, the MC-CDM demodulator 401 for initial determination includes a guard interval remover 201 that removes a guard interval (GI) from the received signal 206, a fast Fourier transformer 202, and 1 / H ( t) It is composed of a multiplier 203 and a despreader 204. In FIG. 8, the P / S converter 205 is omitted.

  The received symbol sequence 209 demodulated by the MC-CDM demodulator 401 for initial determination is determined by the initial determiner 402. The determination result of the initial determination unit 402 is input to the replica generator 403 via the selector 407, and the replica generator 403 receives the reception signal itself that provides the determination result based on the determination result, that is, the initial determination unit 402. If the determination result determined by the above is a true value, a replica of the received signal (desired wave) that is not subjected to multipath interference for transmitting the determination result is generated.

  The replica generator 403 is configured in the same manner as the transmitter when the determination result is a transmission symbol string. In the replica generator 403, as shown in FIG. 8, after the initial determination result is spread by the spreader 102, the H (f) multiplier 109 multiplies the channel transfer function H (f), and the inverse high-speed Inverse Fourier transform is performed by the Fourier transformer 103 to generate a replica of the desired wave. In FIG. 8, the S / P converter 101 is omitted.

  Further, in the interfered sample replacement unit 404, the interference part that has received multipath interference in the received symbol 206 is changed to the interference part that has received multipath interference in the replica of the desired wave generated by the replica generator 403. Replaced by the corresponding part. The received signal 206 replaced with the replica of the desired wave is input to the MC-CDM demodulator 405 for iterative calculation and demodulated by the MC-CDM demodulator 405 for iterative calculation as shown in FIG. . Similar to the MC-CDM demodulator 401 for initial determination, the MC-CDM demodulator 405 for iterative calculation is basically configured similarly to the receiver shown in FIG.

  As shown in FIG. 8, the MC-CDM demodulator 405 for iterative calculation includes a fast Fourier transformer 202 that Fourier transforms the received signal 206 replaced with a replica of the desired wave, and a 1 / H (t) multiplier. 203, a despreader 204, and a P / S converter 205.

  The received signal demodulated by the MC-CDM demodulator 405 for iterative calculation is determined by the determiner 406, and the determination result by the determiner 406 is input to the replica generator 403 again via the selector 407, The same processing as described above is repeated a predetermined number of times (about 1 to 4 times), and then output as an output signal. Note that the number of times the above process is repeated can be set by the determiner 406.

  When the determination result initially determined by the determination unit 406 is output as it is, the number of repetitions is 0, and the determination result first determined by the determination unit 406 is again transmitted to the replica generator 403 via the secretor 407. Is output once, the number of repetitions is one.

  Of the received symbols 206 that are replaced by the replacer 404, the interference portion that has received multipath interference is determined by using a circuit as shown in FIG.

  That is, as shown in FIG. 9, after the guard interval (GI) is removed based on the timing synchronization by the guard interval remover 201, the received signal sequence and the time of the transmission pilot signal are obtained by the cross-correlation calculator 408. By taking the cross-correlation with the waveform replica, a delay profile is obtained, and the interfering part that is receiving multipath interference is determined by the replacer 404. The replacer 404 also has a function of storing the received signal from which the guard interval is removed for a predetermined time.

  In the above configuration, in the multipath interference canceller according to Embodiment 1, the delayed wave due to multipath interference exceeds the guard interval (GI) length and is about ½ of the OFDM symbol length as follows. Even in such a case, good error rate characteristics can be obtained.

Multipath interference removal by the multipath interference canceller 400 is realized by the following procedures 1) to 5).
1) Demodulation of MC-CDM signal 2) Initial decision 3) Reproduction of received signal replica by decision feedback 4) Replacement of interfered sample with replica 5) Demodulation of signal with reduced interference 3) to 5) Thus, the error rate characteristic can be improved. Then, after reducing interference by a plurality of repetitive calculations, a final determination result is obtained.

  That is, in the multipath interference canceller 300 according to the first embodiment, as shown in FIGS. 2 and 3, a transmission signal sent from the transmitter 100 via a transmission path (not shown) is transmitted to the receiver 200. Is received by a receiving antenna that is not received, and is extracted as a received signal 206 by a filter (not shown). The received signal 206 received by the receiver 200 has an MC-CDM demodulator for initial determination as shown in FIG. 1 after the guard interval is removed by the guard interval remover 201 as shown in FIG. 401 is demodulated. In the MC-CDM demodulator 401 for initial determination, as shown in FIG. 8, the received signal from which the guard interval is removed is Fourier-transformed by the fast Fourier transformer 202 and converted into a signal for each subcarrier. The inverse of the channel transfer function H (t) is multiplied by the 1 / H (t) multiplier, despread in the time domain by the despreader 204, and further converted into a serial signal by the P / S converter 205. A reception signal 209 is obtained.

  At that time, as shown in FIG. 7, the transmission symbol transmitted from the transmitter 100 is reflected or diffused by surrounding buildings and becomes a delayed wave in addition to being directly received by the receiver 200 as a preceding wave. And received by the receiver 200.

  If the delay time of the delay wave is within the guard interval (GI) length, the influence of the delay wave can be eliminated and the error rate can be prevented from decreasing, but the delay time of the delay wave is guarded. When the interval (GI) length is exceeded, as shown in FIG. 7, the leading portion of the preceding wave is affected by interference due to the delayed wave over a predetermined symbol length, and the error rate is significantly reduced as it is.

  Therefore, in multipath interference canceller 300 according to Embodiment 1, the received signal demodulated by MC-CDM demodulator 401 for initial determination is determined by initial determiner 402 as shown in FIG. Is not directly adopted as a reception result, but a replica of the desired wave is generated by the replica generator 403 based on the reception result determined by the initial determination unit 402. Of the replica signals generated by the replica generator 403, as shown in FIG. 7, only the signal of the portion receiving multipath interference is replaced with the received signal 206 by the replacer 404. .

  These processes will be further described as follows.

Now, the nth sample of the mth chip of the received signal is set to a _mn , the nth sample of the mth chip of the replica in the i−1th iteration calculation is set to b ⁱ⁻¹ _mn, and demodulated at the ith time. _Assuming that the n th sample of the m th chip of the interference-reduced signal is c ⁱ _mn , c ⁱ _mn can be expressed by the following equation (1). Here, L corresponds to the delay time of the delayed wave, and samples a _{m0 to} a _{m (L−1)} are subjected to multipath interference.

Incidentally, the m-th chip C ⁱ _mn of the n-th subcarrier is given by the following equation (2). Here, N is the number of subcarriers.

When Hn is the channel transfer function of the transmission line, the m-th chip D ⁱ _mn of the n-th subcarrier after channel compensation is the reciprocal of the channel transfer function Hn to the m-th chip C ⁱ _mn of the n-th subcarrier. Is given by equation (3).

From this, the reception symbol E ⁱ _jn of the j-th spreading code of the n-th subcarrier after despreading is _{expressed by} equation (4), where S _jmn is the m-th chip of the j-th spreading code of the n-th subcarrier. Given. Here, M is a spreading code length.

When the determination operation by the determiner 402 is represented by De (), the determination result F ⁱ _mn of the m-th spreading code of the n-th subcarrier is given by the following equation (5).

If the obtained determination results of all symbols are respread, multiplied by the channel transfer function Hn, and subjected to multicarrier modulation by IFFT, replicas b ⁱ _mn in the i-th iterative calculation can be obtained. These correspond to the reverse operations of equations (2) to (4).

Of the replica b ⁱ _mn of the desired wave in the i-th iterative calculation obtained in this way, the part corresponding to the interference part that has received multipath interference is replaced by interference 404 that has received multipath interference among received symbols by the replacer 404. It is replaced with the part a _mn .

  Then, as described above, the interference part that has received multipath interference in the received symbol is replaced with the part corresponding to the interference part that has received multipath interference in the replica of the desired wave generated by the replica generator 403. As shown in FIG. 1, the symbol is demodulated by the MC-CDM demodulator 405 for iterative calculation, and then determined by the determiner 406.

  The above repeated calculation is repeated a predetermined number of times, and finally the determination result by the determiner 406 is output as a reception result.

Thus, in multipath interference canceller 400 according to Embodiment 1 described above, it is possible to improve the error rate characteristics by iterative calculation, as shown in FIG. In particular, when Eb / N0 is 30 dB, the number of iterations is 2, and the error rate characteristic can be reduced to 10 ⁻³ or less, and the error rate characteristic can be greatly improved.

  The reason why the error rate characteristic can be significantly improved in the multipath interference canceller 400 according to Embodiment 1 will be described as follows based on FIG.

  In A in FIG. 11, multipath interference that affects the preceding wave is indicated by an arrow in a state before multipath interference is removed. In FIG. 11, the horizontal axis represents time, and the vertical axis represents the average interference level per sample. Multipath interference affecting the preceding wave is concentrated and distributed in the shaded area in the figure. In B in FIG. 10, the residual interference included in the replica of the received signal generated based on the initial determination result is indicated by an arrow. The residual interference due to the initial determination error is not concentrated on a part but is distributed uniformly on the OFDM symbol, and the average spreading level per sample is reduced. In C of FIG. 11, the residual interference in the replica sample that replaces the interfered sample of the preceding wave is indicated by an arrow. Since the sample to be replaced is a part of all the samples of the replica, the residual interference level is reduced according to the ratio.

  By repeating the above operation (ABC), the residual interference level can be reduced, and the error rate characteristic can be improved as shown in FIG.

Embodiment 2
FIG. 12 shows a second embodiment of the present invention. The same reference numerals are given to the same parts as those of the above-described embodiment. In the second embodiment, the received symbol determination means is a desired symbol. A signal obtained by extracting and demodulating samples corresponding to the number of FFT points not including the guard interval in the wave and a signal obtained by combining the signals obtained by extracting and demodulating the samples corresponding to the number of FFT points including the guard interval in the interference wave are determined. Has been.

  That is, in the multipath interference canceller 400 according to the present invention, since the first received signal replica is generated based on the initial determination result, the reliability of the initial determination result can be improved by performing repeated calculation. This is very important in improving the error rate performance. This is because if a highly reliable determination result is obtained, the accuracy of the replica can be improved, and a high interference reduction capability can be obtained. Therefore, it is effective to apply maximum ratio combining to the received signal at the time of initial determination.

  FIG. 12 shows the configuration of the maximum ratio combiner used in the multipath interference canceller 400 according to the second embodiment.

In the second embodiment, a double spike model is assumed as multipath interference. As shown in FIG. 12, the maximum ratio synthesizer 500 includes a sample extractor 501 corresponding to pass 1 and a sample extractor 504 corresponding to pass 2, and each sample extractor 501 and 504 performs extraction. The obtained samples are fast Fourier transformed by FFT units 502 and 505, respectively, and then multiplied by reciprocals 503 and 504 of the transfer functions H (f) corresponding to the respective paths, and the reciprocals of the respective transfer functions H (f). Are multiplied by coefficients A ₁ and A ₂ by multipliers 507 and 508, and synthesized and output by an adder 509.

  Next, the operation of the maximum ratio synthesizer 500 will be described.

  FIG. 13 shows the FFT calculation range at the time of maximum ratio synthesis. In FIG. 13, the received samples in the shaded area receive interference from adjacent OFDM symbols. Extraction of the received sample for the preceding wave (path 1) is performed from the head of the data sample, and extraction of the received sample for the delayed wave (path 2) is performed from the head of the sample of the guard interval (GI).

The reason for extracting the received sample for the delayed wave (path 2) from the head of the guard interval (GI) sample is to reduce the intersymbol interference level from the adjacent symbol. Delayed wave guard interval (GI) samples are moved backward in the symbol and used for demodulation. The resulting advance wave and delay the m-th against waves of chips n th sample a _1mn, a _2mn is fast Fourier transform, n-th sub-carrier m-th chip C _1mn, it is C _2mn obtained. Further, channel compensation is performed by multiplying the inverse of the channel transfer function by 1 / H _1n and 1 / H _2n . H _1n and H _2n are channel transfer functions when the preceding wave and the delayed wave are the desired waves, respectively. The obtained m-th chip of the n-th subcarrier after channel compensation is weighted and added with the reception levels A ₁ and A ₂ of the preceding wave and the delayed wave to obtain a signal with the maximum ratio synthesis. The m-th chip D ⁰ _mn of the n-th subcarrier after synthesis is shown in Equation (6). D ⁰ _mn is used as an initial value for the iterative calculation.

  If the delay time is longer than the guard interval length, the received samples of the preceding wave and the delayed wave are overlapped but do not completely coincide with each other. Therefore, the error rate characteristic can be improved by the maximum ratio combining. Further, the longer the delay time, the less the duplication of received samples, so that the effect of synthesis increases.

Embodiment 3
FIG. 14 shows a third embodiment of the present invention. The same reference numerals are given to the same parts as those in the previous embodiment. In the third embodiment, phase modulation is performed as primary modulation. In the MC-CDM signal multipath interference canceller used, after demodulating a received symbol including a received symbol that has received multipath interference, a phase error is estimated, and the estimated phase error and a predetermined reference value are estimated. Received symbol determination means for determining whether or not to perform determination based on a comparison result with the above, a replica generation means for generating a replica of a desired wave based on the determination result of the received symbol determination means, and the received symbol The interference part that has received multipath interference is placed in the part corresponding to the interference part that has received multipath interference in the replica of the desired wave generated by the replica generation means. Means replacement changing, the interference portion which has received multipath interference is configured to have a symbol decision means replaces determining demodulate received symbols which are replaced by replicas of the desired signal by said replacement means.

  That is, in the present invention, since there is an error in the determination result of the symbol to be fed back, residual interference occurs. Therefore, in the third embodiment, the multipath interference canceller employs a method of selecting a symbol to be determined and fed back based on reliability information in order to reduce the influence of a determination error. In the third embodiment, a method for selecting a symbol for decision feedback when the QPSK scheme is employed as the primary modulation in the transmitter 100 will be described.

  FIG. 14 shows a decision feedback symbol selection method used in the third embodiment. The primary modulation is QPSK. In this embodiment, the phase error of the received symbol is estimated, the estimated error is compared with a predetermined reference value, and it is determined whether or not to make a determination based on the result. That is, when the phase error is larger than a predetermined reference value, the determination is not performed and the determination result is not fed back. Therefore, the following equation (7) shows the selection logic when selecting the decision feedback symbol. Δφ is an estimated phase error of the received symbol when the determination result is correct. θth is a predetermined reference value. For example, π / 8 may be considered as θth, but may be a value in the vicinity thereof, and is not limited to this as long as the effect of the present invention is exhibited.

  If the received signal point after the spread demodulation is in the shaded area in FIG. 14, the received signal is determined and fed back. If the received signal point is in another area, the corresponding symbol is fed back without being determined. At this time, for the received symbols that have not been determined by the received symbol determining means, the same amplitude value as that of the received symbol for which the determination has been made is fed back. By doing so, since the reception levels of all symbols are the same and QPSK is assumed, the noise included in symbols that are not determined with low reliability is only phase noise.

Experimental example 1
In order to evaluate the effect of the multipath interference canceller according to the present embodiment, the present inventor performed the following computer simulation.

  FIG. 15 shows simulation conditions. The modulation method is QPSK, the number of subcarriers is 64, the guard interval length is 1/4 of the OFDM symbol length, and the spreading ratio is 16. The delay profile was a double spike model, and the channel condition was quasi-static Rayleigh fading. Synchronization and channel estimation were ideal.

  FIG. 16 shows the result of the computer simulation. Here, the error rate characteristic is shown when the delay time of the multipath interference wave is half the OFDM symbol length (corresponding to a delay time of 12.5 μs).

  In FIG. 16, the horizontal axis represents average Eb / N0, and the vertical axis represents average error rate (BER). In FIG. 16, the theoretical error rate characteristic during Rayleigh fading is indicated by a dotted line as a reference. Further, in FIG. 16, the one-dot chain line and the two-dot chain line indicate the characteristics when the maximum ratio synthesis is performed and without the repeated calculation, respectively.

As can be seen from FIG. 16, the error floor of 8.times.10.sup.- ² can be improved to 3.times.10.sup.- ² by combining the signals of each path with the maximum ratio. Further, the solid line represents the error rate characteristic when the iterative calculation of the present embodiment is applied after performing the maximum ratio combining process. The number of repetitions is 1 to 4 times, and is indicated by a suffix in the figure. In the decision feedback at the time of iterative calculation, the feedback symbol selection method shown in FIG. 14 described above is employed.

  FIG. 16 shows that the error rate characteristics can be improved by increasing the number of repetition calculations. This is because the residual interference level is reduced by repeated calculation as described above. When the number of repetitions is 2 or less, an error rate characteristic with a floor is obtained. In addition, it was found that the error rate characteristic converges by repeating three times, and an error characteristic without a floor can be obtained.

Experimental example 2
FIG. 17 shows the error rate characteristics when the delay time of the multipath interference wave is 5/8 of the OFDM symbol length (corresponding to a delay time of 15.6 μs). As is clear from FIG. 17, a result showing the same tendency as in FIG. 16 was obtained. It can be seen that the error rate characteristics can be improved as the number of repeated calculations increases. An error rate characteristic with no floor was obtained by four iterations.

Experimental example 3
Next, the present inventor conducted a computer simulation for examining the dependency of the error rate on the short code length.

  FIG. 18 shows the short code length dependence of the error rate when noise is free. The short code length for spreading OFDM symbols was set from 1 to 16. The horizontal axis in FIG. 18 indicates the number of repeated calculations, and the vertical axis indicates the error rate. Here, symbol selection at the time of decision feedback is not performed, and all symbols are determined to be hard. The delay time of the multipath interference wave is 9/16 of the OFDM symbol length.

  As is apparent from FIG. 18, the error rate characteristics are improved with an increase in the number of repetition calculations. However, when the code length is 1, 2, 4, the error rate is almost constant when the number of repetitions is 3 or more. I understand that In the region where the error rate is constant, the residual interference is concentrated in the region where the judgment is fed back and it is considered that the residual interference level is not reduced. This corresponds to the fact that, when viewed in time series, if the same result as the previous determination result is obtained in the repeated calculation, the same result is obtained even if the calculation is repeated.

  When the number of repetitions is 1, as is apparent from FIG. 18, there is no dependency of the error rate on the short code length, and the same error rate is obtained for all code lengths. It can be seen that when the number of repetitions is 2 or more, the error rate characteristics are improved as the short code length increases. Since all short codes are used and are of equal power, the effective processing gain is unity. Since the effective processing gain is always 1 even if the short code length is changed, it can be seen that the reason for the improvement of the error rate does not depend on the processing gain. The error rate is considered to be improved by the whitening effect of the colored noise of the CDM signal.

  Next, the reason why the error rate depends on the code length will be described. 1) When the number of repetitions is 1, the residual noise due to the initial determination error is uniformly distributed on the OFDM symbol, so that the interference noise is white. 2) When the number of repetitions is 2 or more, the residual interference is not distributed uniformly, and the probability of distribution in the decision feedback area increases, so the interference noise is colored. From 1), the residual interference at the time of initial determination is originally white and the whitening effect of CDM does not work. Therefore, increasing the short code length does not improve the error rate. It is considered that the error rate characteristic is improved as the short code length increases due to the whitening effect of the CDM colored noise.

  As described above in detail, it is possible to provide a multipath interference canceller applicable to the downlink of the MC-CDM system. Such a multipath interference canceller can improve error rate characteristics even when a delayed wave longer than the guard interval (GI) length exists. The removal of multipath interference is effectively realized by the following procedure. 1) Demodulation of MC-CDM signal using maximum ratio combining, 2) Initial decision, 3) Generation of received signal replica by decision feedback, 4) Replacing interfered sample with replica, 5) Demodulation of signal with reduced interference 6) Error rate characteristics can be improved by selecting a decision feedback symbol based on reliability information and by repeating the calculation from steps 3) to 6).

  Then, as a result of computer simulation to evaluate the characteristics of the multipath interference canceller according to the present invention, it was found that the present invention is applicable to a propagation environment in which a delayed wave up to 5/8 of the OFDM symbol length exists.

  In the second embodiment, the case where the maximum ratio combining is performed has been described. However, the present invention is not limited to this, and the received symbol determination means can perform sampling for the number of FFT points not including the guard interval in the desired wave. The signal is synthesized by simply synthesizing the demodulated signal and the signal obtained by extracting and demodulating the number of FFT points including the guard interval in the interference wave without multiplying the coefficients, and determining the synthesized signal. Any means may be used.

FIG. 1 is a block diagram showing an MC-CDM signal multipath interference canceller according to Embodiment 1 of the present invention. FIG. 2 is a block diagram showing a transmitter in the MC-CDM mobile communication system to which the multipath interference canceller according to Embodiment 1 of the present invention is applied. FIG. 3 is a block diagram showing a receiver in the MC-CDM mobile communication system to which the multipath interference canceller according to Embodiment 1 of the present invention is applied. FIG. 4 is an explanatory diagram showing an OFDM symbol with a guard interval added. FIG. 5 is an explanatory diagram showing symbol arrangement. FIG. 6 is an explanatory diagram showing a state of multiple access by the user. FIG. 7 is an explanatory diagram showing the influence of multipath interference. FIG. 8 is a block diagram showing a multipath interference canceller for MC-CDM signals according to Embodiment 1 of the present invention. FIG. 9 is a block diagram showing a circuit configuration for obtaining a delay profile. FIG. 10 is a graph showing the results of computer simulation. FIG. 11 is an explanatory diagram showing the operation of the multipath interference canceller for MC-CDM signals according to Embodiment 1 of the present invention. FIG. 12 is a block diagram showing a main part of a multipath interference canceller for MC-CDM signals according to Embodiment 2 of the present invention. FIG. 13 is an explanatory diagram showing the operation of the multipath interference canceller for MC-CDM signals according to Embodiment 2 of the present invention. FIG. 14 is an explanatory diagram showing a multipath interference canceller for MC-CDM signals according to Embodiment 3 of the present invention. FIG. 15 is a chart showing conditions for computer simulation. FIG. 16 is a graph showing the results of computer simulation. FIG. 17 is a graph showing the results of computer simulation. FIG. 18 is a graph showing the results of computer simulation. FIG. 19 is a graph showing the error rate characteristics when the delay time is changed with respect to the guard interval length in the conventional example. FIG. 20 is a graph showing an average error rate characteristic when the delay time is changed with respect to the guard interval length in another conventional example.

Explanation of symbols

  400: Multipath interference canceller, 401: MC-CDM demodulator for initial determination, 402: Initial determination unit, 403: Replica generator, 404: Replacer, 405: MC-CDM demodulator for iterative calculation, 406: Determinator, 407: Secreta.

Claims (6)

In the MC-CDM or MC-CDMA signal multipath interference canceller,
Received symbol determination means for demodulating and determining received symbols including received symbols that have undergone multipath interference;
Replica generating means for generating a replica of a desired wave based on the determination result of the received symbol determining means;
Replacement means for replacing an interference part that has received multipath interference in the received symbol with a part corresponding to an interference part that has received multipath interference among replicas of a desired wave generated by the replica generation means;
A replacement symbol determination unit that demodulates and determines a received symbol in which the interference part subjected to multipath interference by the replacement unit is replaced with a replica of a desired wave;
A multipath interference canceller comprising: The received symbol determination means includes:
A signal obtained by extracting and demodulating samples for the number of FFT points not including the guard interval in the desired wave;
A signal obtained by synthesizing a signal obtained by taking out and demodulating samples corresponding to the number of FFT points including the guard interval in the interference wave,
The multipath interference canceller according to claim 1, wherein the determination is made. In the MC-CDM or MC-CDMA signal multipath interference canceller using phase modulation as the primary modulation,
After demodulating a received symbol that includes a received symbol that has undergone multipath interference, a phase error is estimated, and it is determined whether or not to make a determination based on a comparison result between the estimated phase error and a predetermined reference value. Received symbol determination means for performing determination;
Replica generating means for generating a replica of a desired wave based on the determination result of the received symbol determining means;
Replacement means for replacing an interference part that has received multipath interference in the received symbol with a part corresponding to an interference part that has received multipath interference among replicas of a desired wave generated by the replica generation means;
A replacement symbol determination unit that demodulates and determines a received symbol in which the interference part subjected to multipath interference by the replacement unit is replaced with a replica of a desired wave;
A multipath interference canceller comprising: 4. The multipath interference canceller according to claim 3, wherein the same amplitude value as that of the received symbol for which a determination has been made is fed back for a received symbol that has not been determined by the received symbol determination means. 5. The multipath interference canceller according to claim 1, wherein multipath interference is reduced by repeatedly performing processing by the replica generation means, replacement means, and replacement symbol determination means. In the MC-CDM or MC-CDMA signal multipath interference canceling method,
After generating a replica of the desired wave based on the result of demodulating and determining the received symbol containing the received symbol that has undergone multipath interference,
A multipath interference canceling method characterized by repeating the process of demodulating and determining a received symbol obtained by replacing a part of the received symbol subjected to the multipath interference with a replica of a desired wave at least once.

Images