The present invention relates to a MIMO receiving apparatus and receiving method for making multiple-input multiple-output (MIMO) radio communications using radio technologies, and more particularly, to a MIMO receiving apparatus and receiving method which converts single-carrier MIMO signals to signals in a frequency domain, and which separates the MIMO signals through processing in the frequency domain.
High-speed data transmissions are required for radio systems of the next-generation mobile communication. As a technology for realizing this high-speed data transmission, MIMO (Multiple Input Multiple Output) multiplexing has drawn attention, where data signals are transmitted from a plurality of transmission antennas using the same frequency, and where the data signals are demodulated using a plurality of reception antennas.
The MIMO transmission/reception system shown in
While a variety of schemes have been proposed for the signal separation of MIMO, the simplest scheme, for example involves restraining interference from transmission antennas other than the transmission antenna subjected to demodulation, using a linear filter based on minimum mean square error (MMSE). However, the reception characteristics cannot be said to be good when this scheme is used. To improve the reception characteristics using this scheme, a scheme has been proposed for combining an MMSE filter with transmission antenna interference removal.
Also, a maximum likelihood detection (MLD) has been devised for generating replicas of all transmission antenna signals and selecting the most likely transmission antenna signal. This detection method, though exhibiting good reception characteristics, causes an exponential increase in processing amount in accordance with an increase in the number of transmission antennas and the modulation multiple-value.
On the other hand, in an uplink radio scheme of the next-generation mobile communication, a terminal must realize a high transmission power efficiency in order to expand a communication area, and a single-carrier (SC) scheme having a low peak to average power ratio (PAPR) is regarded as predominant. Also, when MIMO multiplexing is performed in the SC scheme, multipath interference constitutes a problem. Also, when an MMSE filter is used to separate signals of MIMO, an MMSE filter (two-dimensional MMSE filter) is required to simultaneously separate the MIMO signals and to restrain the multipath interference, i.e., in a spatial direction and a time direction. Then, a scheme which combines the two-dimensional MMSE filter with transmission antenna interference removal excels in characteristics, and is regarded as predominant.
In the MIMO receiving apparatus shown in
The MIMO receiving apparatus shown in
The frame shown in
Also, reception antenna 101-1-101-N shown in
NDFT1(NDFT1 is an integer equal to or larger than two) [Expression 1]
to convert the reception signal to the frequency domain. Reception filter 104-1-104-N filters the reception signal in the frequency domain, and performs waveform shaping, noise suppression, user separation, and the like. Generally, raised cosine roll-off filters are used for reception filters 104-1-104-N. In the configuration shown in
Subtracting part 105 shown in
Assuming that reception signals at sub-carriers
k(1≦k≦NDFT1) [Expression 2]
after DFT are given by:
R(k)(R(k) is an N-row column vector) [Expression 3]
and i-th repetition interference replica of transmission antenna m is given by:
Îm(i)(k)(Îm(i)(k) is an N-row column vector), [Expression 4]
an equalizing signal for transmission antenna m after removal of the i-th repetition interference:
Rm(i)(k)(Rm(i)(k) is an N-row column vector), [Expression 5]
is given by the following equation:
Here, interference removal is not performed in the first reception processing (zero-th repetition), and a reception signal is used as it is. Specifically,
[Expression 7]
Rm(0)(k)=R(k) (Equation 2)
is established.
Channel estimating part 106 estimates a channel gain between a transmission antenna and a reception antenna in the frequency domain using a pilot signal inserted into each transmission antenna.
Channel estimating part 106 shown in
Weight calculating part 107 calculates a weight for two-dimensional frequency domain equalization using the channel estimate between the transmission antenna and the reception antenna. Generally, an MMSE algorithm is used for weight calculating part 107. An i-th repetition MMSE weight for transmission antenna m:
Wm(i)(k)(Wm(i)(k) is an N-column row vector) [Expression 8]
is calculated using a channel estimate matrix:
Ĥ(k)(Ĥ(k) is an N-row, M-column matrix) [Expression 9]
and noise power:
σ2 [Equation 10]
by the following equation:
[Expression 11]
Wm(i)(k)=ĤmH(k)[Ĥ(k)Gm(i)ĤH(k)+σ2I]−1 (Equation 3)
where
[Expression 12]
Ĥ(k)=[Ĥ1(k), . . . ,Ĥm(k), . . . ,ĤM(k)] (Equation 4)
Also,
Ĥm(k)(Ĥm(k) is an N-row column vector) [Expression 13]
is a channel estimate between transmission antenna m and the reception antenna.
Gm(i) [Equation 14]
is an i-th repetition interference removal consideration matrix of transmission antenna m, and is given by the following equation:
[Expression 15]
Gm(i)=diag[gm,1(i), . . . ,gm,m′(i), . . . ,gm,M(i)] (Equation 5)
Here,
gm,m′(i) [Expression 16]
is calculated using, for example, the average power of an i-th repetition soft decision symbol replica in the time domain of transmission antenna m:
{circumflex over (d)}m(i)(t) [Equation 17]
by the following equation:
Here,
NSYMB [Expression 19]
is the number of symbols in a data block.
Two-dimensional frequency domain equalizing part 108 receives the two-dimensional equalization weight calculated in weight calculating part 107 and the output of subtracting part 105, and multiplies them on a sub-carrier by sub-carrier basis to simultaneously separate MIMO signals and suppress multipath interference in the frequency domain. Assuming that the weight calculated in weight calculating part 107 is given by:
Wm(i)(k) [Expression 20]
and the output of subtracting part 105 is given by:
Rm(i)(k) [Expression 21]
an equalization signal of transmission antenna m two-dimensionally equalized in two-dimensional frequency domain equalizing part 108:
{tilde over (R)}m(i)(k) [Expression 22]
is given by the following equation:
[Expression 23]
{tilde over (R)}m(i)(k)=Wm(i)(k)Rm(i)(k) (Equation 7)
IDFT part 109-1-109-M which receives an equalization signal of each transmission antenna which is the output of two-dimensional frequency domain equalizing part 108, performs IDFT at point:
NIDFT(NIDFT is an integer equal to or larger than two) [Expression 24]
to convert the equalization signal to the time domain. Outputs of an i-th repetition (i≧1) of IDFT parts 109-1-109-M constitute a final demodulated signal.
Bit likelihood calculating part 110-1-110-M calculates a likelihood for each bit transmitted from the equalization signal of each transmission antenna. Bit likelihood calculating parts 110-1-110-M also include hard decision of bits.
Symbol replica generating part 111-1-111-M generates a symbol replica of transmission antenna m from the likelihood of a bit transmitted from each transmission antenna. Used in symbol replica generating parts 111-1-111-M are a method of generating a hard decision symbol replica, a method of generating a hard decision symbol replica and multiplying it by a predetermined replica weighting coefficient (constant equal to or less than one), a method of generating a soft decision symbol replica, and the like. In particular, the method of generating a soft decision symbol replica exhibits good characteristics.
DFT part 112-1-112-M receives a symbol replica of each transmission antenna generated in symbol replica generating part 111-1-111-M, and performs DFT at point:
NDFT2(NDFT2 is an integer equal to or larger than two) [Expression 25]
to convert the symbol replica to the frequency domain.
Antenna interference replica generating part 113-1-113-M generates a transmission antenna interference replica using a symbol replica signal in the frequency domain of each transmission antenna and a channel estimate. Assuming that a symbol replica signal in the frequency domain of transmission antenna m is given by:
{circumflex over (D)}m(i)(k) [Expression 26]
and the channel estimate is given by:
Ĥm(k) [Expression 27]
an i-th repetition interference replica of transmission antenna m:
Îm(i)(k) [Expression 28]
is given by the following equation:
[Expression 29]
Îm(i)(k)=Ĥm(k){circumflex over (D)}m(i−1)(k) (Equation 8)
As described above, the conventional MIMO receiving apparatus is characterized in that the processing amount is largely reduced as compared with signal processing in the time domain, and it is not affected by a timing error when the two-dimensional frequency domain equalization and antenna interference removal in the frequency domain, including the channel estimation, are performed.
However, the conventional MIMO receiving apparatus has problems in that repetitive reception processing for two-dimensional frequency domain equalization and antenna interference removal still continues linear processing in the time direction, a weight after convergence of the two-dimensional frequency domain equalization is an MMSE weight for a transmission antenna signal subjected to demodulation, and noise emphasis due to the equalization remains even if repetitions are made. Accordingly, the conventional MIMO receiving apparatus is inferior to the reception characteristics of MLD which has excellent characteristics.
In order to solve the problems as described above, it is an object of the present invention to provide a MIMO receiving apparatus and receiving method which are capable of realizing excellent MIMO reception characteristics because of the ability to reduce the amount of processing for calculating weights in the removal of antenna interference and the equalization of two-dimensional frequency domain and the ability to eliminate not only other antenna interference but also multipath interference of a transmission antenna subjected to demodulation, provided by employing a configuration for removing all antenna interference replicas before two-dimensional frequency domain equalization and for adding a distortion-free signal replica to each transmission antenna after two-dimensional frequency domain equalization in a MIMO receiving apparatus for converting single-carrier MIMO signals to signals in a frequency domain and separating the MIMO signals through processing in the frequency domain.
To achieve the above object, the present invention provides a MIMO receiving apparatus for receiving single-carrier signals transmitted from a plurality of transmission antennas provided on a transmitting apparatus at a plurality of reception antennas and for separating the signals by a frequency domain. The MIMO receiving apparatus comprises:
a DFT part for discrete Fourier transforming the signals received at the reception antennas at a first point;
a channel estimating part for estimating a channel gain between the transmission antenna and the reception antenna using a pilot signal inserted into a signal transmitted from the transmission antenna;
a replica reproducing part for performing two-dimensional frequency domain equalization based on a weight calculated using the channel gain, and generating an interference replica and a distortion-free signal replica of each transmission antenna;
a subtracting part for removing the interference replicas of all the transmission antennas from the signals discrete Fourier transformed in the DFT part; and
a demodulating part for performing two-dimensional frequency domain equalization with a weight calculated on the basis of the interference removal by the subtracting part using the channel gain for the signals from which the interference replicas have been eliminated, and for adding the distortion-free signal replicas to the equalized signal to demodulate the signals transmitted from the transmission antennas.
Also, the MIMO receiving apparatus is characterized by further comprising a plurality of subtracting parts and a plurality of replica reproducing parts after the replica reproducing part, for repeatedly performing antenna interference removal and replica reproduction processing.
The MIMO receiving apparatus is also characterized by simultaneously demodulating transmission antenna signals to eliminate antenna interference in parallel.
The MIMO receiving apparatus is also characterized by ordering transmission antenna signals based on a reception quality, demodulating the transmission antenna signals with higher qualities, and sequentially removing the antenna interference.
The MIMO receiving apparatus is also characterized in that the replica reproducing part comprises:
a weight calculating part for calculating a weight for two-dimensional frequency domain equalization using a channel estimate between the transmission antenna and the reception antenna;
a two-dimensional frequency domain equalizing part applied with the weight and a signal output from the subtracting part to multiply the same on a sub-carrier by sub-carrier basis to perform and output MIMO signal separation and multipath interference suppression in a frequency domain;
an IDFT part for discrete inverse Fourier transforming a demodulated signal of each transmission antenna which is a signal output from the two-dimensional frequency domain equalizing part at a second point;
a bit likelihood calculating part for calculating a bit likelihood for each transmitted bit based on the demodulated signal of each transmission antenna;
a symbol replica generating part for generating a symbol replica based on the bit likelihood;
a DFT part for discrete Fourier transforming the symbol replica at a third point;
an antenna interference replica generating part for generating a transmission antenna interference replica using a symbol replica signal in the frequency domain and a channel estimate; and
a distortion-free signal replica generating part for generating a distortion-free signal replica using the symbol replica signal in the frequency domain and a sub-carrier average value of the channel gain after two-dimensional frequency domain equalization.
The MIMO receiving apparatus is also characterized in that the demodulating part comprises:
a weight calculating part for calculating a weight for two-dimensional frequency domain equalization using a channel estimation value between the transmission antenna and the reception antenna based on interference removal by the subtracting part;
a two-dimensional frequency domain equalizing part applied with the weight and a signal output from the subtracting part to multiply the same on a sub-carrier by sub-carrier basis to perform and output MIMO signal separation and multipath interference suppression in a frequency domain;
an adding part for adding and outputting a distortion-free signal replica to a signal output from the two-dimensional frequency domain equalizing part; and
an IDFT part for discrete inverse Fourier transforming a demodulated signal of each transmission antenna which is a signal output from the adding part at a second point.
The MIMO receiving apparatus is also characterized in that the plurality of replica reproducing parts provided after the replica reproducing part comprises:
a weight calculating part for calculating a weight for two-dimensional frequency domain equalization using a channel estimation value between the transmission antenna and the reception antenna based on interference removal by the subtracting part;
a two-dimensional frequency domain equalizing part applied with the weight and a signal output from the subtracting part to multiply the same on a sub-carrier by sub-carrier basis to perform and output MIMO signal separation and multipath interference suppression in a frequency domain;
an adding part for adding and outputting a distortion-free signal replica to a signal output from the two-dimensional frequency domain equalizing part; and
an IDFT part for discrete inverse Fourier transforming a demodulated signal of each transmission antenna which is a signal output from the adding part at a second point;
a bit likelihood calculating part for calculating a bit likelihood for each transmitted bit based on the demodulated signal of each transmission antenna;
a symbol replica generating part for generating a symbol replica based on the bit likelihood;
a DFT part for discrete Fourier transforming the symbol replica at a third point;
an antenna interference replica generating part for generating a transmission antenna interference replica using a symbol replica signal in the frequency domain and a channel estimate; and
a distortion-free signal replica generating part for generating a distortion-free signal replica using the symbol replica signal in the frequency domain and a sub-carrier average value of the channel gain after two-dimensional frequency domain equalization.
The MIMO receiving apparatus is also characterized in that the replica reproducing part further comprises a decoding part after the bit likelihood calculating part, and generates a symbol replica using error correction decoded bits.
The MIMO receiving apparatus is also characterized in that the plurality of replica reproducing parts provided after the replica reproducing part further comprise a decoding part after the bit likelihood calculating part, and generate a symbol replica using error correction decoded bits.
The present invention also provides a MIMO receiving method for receiving single-carrier signals transmitted from a plurality of transmission antennas at a plurality of reception antennas and for separating the signals through processing in a frequency domain. The MIMO receiving method comprises:
processing for discrete Fourier transforming the signals received at the reception antennas at a first point;
processing for estimating a channel gain between the transmission antenna and the reception antenna using a pilot signal inserted into a signal transmitted from the transmission antenna;
processing for performing two-dimensional frequency domain equalization based on a weight calculated using the channel gain, and for generating an interference replica and a distortion-free signal replica of each transmission antenna;
processing for removing the interference replicas of all the transmission antennas from the signals that have been discrete Fourier transformed;
processing for performing two-dimensional frequency domain equalization with a weight calculated on the basis of interference removal by the subtracting part using the channel gain for the signals from which the interference replicas have been eliminated; and
processing for adding the distortion-free signal replicas to the equalized signal to demodulate the signals transmitted from the transmission antennas.
In the present invention configured as described above, in the MIMO receiving apparatus in which single-carrier signals transmitted from a plurality of transmission antennas are received at a plurality of reception antennas, and in which the signals are separated by a frequency domain, interference replica of the antennas is eliminated from the received signals before two-dimensional frequency domain equalization is performed, and distortion-free signal replicas are added to the signals from which interference replicas of the antennas have been eliminated, after two dimensional frequency domain equalization was performed.
Accordingly, it is possible to reduce the amount of processing for weight calculation for antenna interference removal and two-dimensional frequency domain equalization, and to eliminate not only interference caused by other antennas but also multipath interference of a transmission antenna subjected to demodulation.
As described above, in the MIMO receiving apparatus in which single-carrier signals transmitted from a plurality of transmission antennas are received at a plurality of reception antennas, and in which the signals are separated by a frequency domain, the present invention is configured to eliminate interference replica of the antennas from the received signals before the two-dimensional frequency domain equalization is performed, and to add distortion-free signal replicas to the signals from which 15 the interference replicas of the antennas have been eliminated, after two dimensional frequency domain equalization was performed, so that the present invention can realize excellent MIMO reception characteristics.
A diagram showing an exemplary configuration of a conventional MIMO transmission/reception system which employs MIMO multiplexing.
A diagram showing an exemplary configuration of a conventional MIMO receiving apparatus.
A diagram showing an example of a radio frame format when frequency domain equalization is employed.
A diagram showing an exemplary configuration of a subtracting part for a DFT signal of reception antenna n.
A diagram showing an exemplary configuration of a channel estimating part for calculating a channel gain of transmission antenna m in reception antenna n.
A diagram showing a first embodiment of a MIMO receiving apparatus of the present invention.
A diagram showing a second embodiment of a MIMO receiving apparatus of the present invention.
A diagram showing a mode which is a combination of the first embodiment shown in
A diagram showing the configuration of a subtracting part for a DFT signal of reception antenna n.
In the following, embodiments of the present invention will be described with reference to the drawings.
As shown in
DFT part 51 discrete Fourier transforms signals received at reception antennas 1-1-1-N at point:
NDFT1(NDFT is an integer equal to or larger than two) [Expression 30]
which is a first point. Channel estimating part 52 estimates a channel gain between a transmission antenna and a reception antenna using a pilot signal inserted in a signal transmitted from each transmission antenna. Replica reproducing part 53 performs two-dimensional frequency domain equalization based on a weight calculated using the channel gain, and generates an interference replica which imitates an interference signal of each transmission antenna, and a distortion-free signal replica which imitates a distortion-free signal. Subtracting part 54 eliminates interference replicas of all transmission antennas from a signal output from DFT part 51. Demodulating part 55 performs two-dimensional frequency domain equalization with a weight calculated on the basis of interference removal of subtracting part 54 using the channel gain for signals from which the interference replicas have been eliminated, and adds the distortion-free signal replicas to equalized signals to demodulate each transmission antenna signal.
The MIMO receiving apparatus of this embodiment is configured such that subtracting part 54 eliminates the interference replicas of all transmission antennas, and the distortion-free signal replica is added to each transmission antenna signal after two-dimensional frequency domain equalization by demodulating part 55, thereby making it possible to eliminate not only other antenna interference but also multipath interference of a transmission antenna subjected to demodulation. Accordingly, excellent MIMO reception characteristics can be realized without increasing the amount of processing as compared with before.
As shown in
DFT part 51 discrete Fourier transforms signals received at reception antennas 1-1-1-N at point:
NDFT1(NDFT1 is an integer equal to or larger than two) [Expression 31]
which is a first point. Channel estimating part 52 estimates a channel gain between a transmission antenna and a reception antenna using a pilot signal inserted in a signal transmitted from each transmission antenna. Replica reproducing part 53 performs two-dimensional frequency domain equalization based on a weight calculated using the channel gain, and generates an interference replica and a distortion-free signal replica of each transmission antenna. Subtracting parts 56-1-56-K eliminate interference replicas of all transmission antennas from a signal output from DFT part 51, respectively. Replica reproducing parts 57-1-57-K respectively perform two-dimensional frequency domain equalization with a weight which is calculated taking in to consideration interference removal by subtracting parts 56-1-56-K from the channel gain. Then, replica reproducing parts 57-1-57-K add distortion-free signal replicas to their outputs to demodulate each transmission antenna signal, and generate interference replica and a distortion-free signal replica of each transmission antenna. Subtracting part 54 eliminates interference replicas of all transmission antennas from the signal output from DFT part 51. Demodulating part 55 performs two-dimensional frequency domain equalization with a weight calculated on the basis of interference removal by subtracting part 54 using the channel gain for signals from which the interference replicas have been eliminated, and adds the distortion-free signal replicas to equalized signals to demodulate each transmission antenna signal.
The MIMO receiving apparatus of this embodiment is configured such that subtracting parts 56-1-56-K and subtracting part 54 eliminate the interference replicas of all transmission antennas, and the distortion-free signal replica is added to each transmission antenna signal after two-dimensional frequency domain equalization by replica reproducing parts 57-1-57-K and demodulating part 55. In this way, since the MIMO receiving apparatus can eliminate not only other antenna interference but also multipath interference of a transmission antenna subjected to demodulation, it can realize excellent MIMO reception characteristics without increasing the amount of processing as compared with before. In addition, the MIMO reception characteristics can be further improved by generating highly accurate replicas through repetitions of antenna interference removal and replica reproduction processing.
Next, detailed operations of the MIMO receiving apparatus of the present invention will be described with reference to
As shown in
Reception antennas 1-1-1-N receive single-carrier MIMO signals. CP removing part 2-1-2-N is applied with the signal received at each reception antenna 1-1-1-N, and removes part of the signal corresponding to CP at a common timing. DFT part 3-1-3-N receives each reception antenna signal, from which the CP has been removed, and performs DFT at point:
NDFT1(NDFT1 is an integer equal to or larger than two) [Expression 32]
which is a first point, to convert the reception signal to the frequency domain. Reception filters 4-1-4-N filter the reception signals in the frequency domain and perform waveform shaping, noise suppression, user separation, and the like. Generally, raised cosine roll-off filters are used for reception filters 4-1-4-N. In the configuration shown in
Subtracting part 5 shown in
k(1≦k≦NDFT1) [Expression 33]
after DFT are given by:
R(k)(R(k) is an N-row column vector), [Expression 34]
and i-th repetition interference replica of transmission antenna m is given by:
Îm(i)(k)(Îm(i)(k) is an N-row column vector), [Expression 35]
an equalizing signal for each transmission antenna after removal of the i-th repetition interference:
R(i)(k)(R(i)(k) is an N-row column vector), [Expression 36]
is common to all transmission antennas, and given by the following equation:
Here, interference removal is not performed in the first reception processing (zero-th repetition), and a reception signal is used as it is. Specifically,
[Expression 38]
R(0)(k)=R(k) (Equation 10)
is established. Since the calculation of antenna interference removal in Equation (9) is common to each transmission antenna, the amount of processing can be reduced as compared with a conventional processing.
Channel estimating part 6 estimates a channel gain between a transmission antenna and a reception antenna in the frequency domain using a pilot signal inserted in a signal transmitted from each transmission antenna. Channel estimating part 6 in this embodiment performs signal processing in the frequency domain, but can also perform signal processing in the time domain prior to DFT parts 3-1-3-N.
Weight calculating part 7 calculates a weight for two-dimensional frequency domain equalization using the channel estimate which is a channel gain between the transmission antenna and the reception antenna in a channel estimating part 6. Generally, an MMSE algorithm is used for weight calculating part 7. An i-th repetition MMSE weight for transmission antenna m:
Wm(i)(k)(Wm(i)(k) is an N-column row vector) [Expression 39]
is calculated using a channel estimate matrix:
Ĥ(k)(Ĥ(k) is an N-row, M-column matrix) [Expression 40]
and noise power:
σ2 [Equation 41]
by the following equation:
[Expression 42]
Wm(i)(k)=ĤmH(k)[Ĥ(k)G(i)ĤH(k)+σ2I]−1 (Equation 11)
where
[Expression 43]
Ĥ(k)=[Ĥ1(k), . . . ,Ĥm(k), . . . ,ĤM(k)] (Equation 12)
Also,
Ĥm(k)(Ĥm(k) is an N-row column vector) [Expression 44]
is a channel estimate between transmission antenna m and reception antenna.
G(i) [Equation 45]
is an i-th repetition interference removal consideration matrix, and is common to all transmission antennas, and is given by the following equation:
[Expression 46]
G(i)−diag[g1(i), . . . ,gm′(i), . . . ,gM(i)] (Equation 13)
Here,
gm′(i) [Expression 47]
is calculated using, for example, the average power of an i-th repetition soft decision symbol replica in the time domain of transmission antenna m:
{circumflex over (d)}m(i)(t) [Equation 48]
by the following equation:
Here,
NSYMB [Expression 50]
is the number of symbols in a data block. In the weight calculation by Equation (11), the calculation of inverse matrix [ ]−1 is common to transmission antenna m and has to be performed only once, so that the amount of processing can be reduced as compared with conventional processing.
Two-dimensional frequency domain equalizing part 8 receives the two-dimensional equalization weight calculated in weight calculating part 7 and the output of subtracting part 5, and multiplies them on a sub-carrier by sub-carrier basis to simultaneously perform and output MIMO signal separation and multipath interference suppression in the frequency domain. Assuming that the weight calculated in weight calculating part 7 is given by:
Wm(i)(k) [Expression 51]
and the output of subtracting part 5 is given by:
R(i)(k) [Expression 52]
an equalization signal of transmission antenna m two-dimensionally equalized in two-dimensional frequency domain equalizing part 8:
{tilde over (R)}m(i)(k) [Expression 53]
is given by the following equation:
[Expression 54]
{tilde over (R)}m(i)(k)=Wm(i)(k)R(i)(k) (Equation 15)
Adding part 9 adds a distortion-free signal replica to an equalization signal of each transmission antenna and outputs the resulting sum. Assuming that the distortion-free signal replica of transmission antenna m is given by:
Ŝm(i)(k) [Expression 55]
an i-th repetition demodulation signal of transmission antenna m:
{tilde over (R)}′m(i)(k) [Expression 56]
is given by the following equation:
[Expression 57]
{tilde over (R)}′m(i)(k)={tilde over (R)}m(i)(k)+Ŝm(i)(k) (Equation 16)
The processing of Equation (16) is not involved in a conventional one and causes an increase in the amount of processing, because this is additional processing for each transmission antenna and requires a smaller amount of processing, even in combination with the subtraction processing of Equation (9), as compared with the conventional subtraction processing.
IDFT part 10-1-10-M receives a demodulated signal of each transmission antenna which is the output of adding part 9, and performs IDFT at point:
[Expression 58]
NIDFT(NIDFT is an integer equal to or larger than two), which is a second point, to convert the modulated signal to the time domain. Outputs of an i-th repetition (i≧1) of IDFT parts 10-1-10-M constitute a final demodulated signal.
Bit likelihood calculating part 11-1-11-M calculates a likelihood for each bit transmitted on the basis of the demodulated signal of each transmission antenna. Bit likelihood calculating parts 11-1-11-M may also include hard decision of bits.
Symbol replica generating part 12-1-12-M generates a symbol replica of transmission antenna m based on the bit likelihood of a bit transmitted from each transmission antenna. Used in symbol replica generating parts 12-1-12-M are a method of generating a hard decision symbol replica, a method of generating a hard decision symbol replica and multiplying it by a predetermined replica weighting coefficient (a constant equal to or less than one), a method of generating a soft decision replica, and the like. Also, in the configuration shown in
DFT part 13-1-13-M receives a symbol replica of each transmission antenna generated in symbol replica generating part 12-1-12-M, and performs DFT at point:
[Expression 59]
NDFT2(NDFT2 is an integer equal to or larger than two), which is a third point, to convert the symbol replica to the frequency domain.
Antenna interference replica generating part 14-1-14-M generates a transmission antenna interference replica using a symbol replica signal in the frequency domain of each transmission antenna and a channel estimate. Assuming that a symbol replica signal in the frequency domain of transmission antenna m is given by:
{circumflex over (D)}m(i)(k) [Expression 60]
and the channel estimate is given by:
Ĥm(k) [Expression 61]
an i-th repetition interference replica of transmission antenna m:
Îm(i)(k) [Expression 62]
is given by the following equation:
[Expression 63]
Îm(i)(k)=Ĥm(k){circumflex over (D)}m(i−1)(k) (Equation 17)
Distortion-free signal replica generating part 15-1-15-M generates a distortion-free signal replica using a symbol replica signal in the frequency domain of each transmission antenna and a sub-carrier average value of the channel gain after two-dimensional frequency domain equalization. Assuming that the symbol replica signal in the frequency domain of transmission antenna m is given by:
{circumflex over (D)}m(i)(k) [Expression 64]
and the sub-carrier average value of the channel gain after equalization is given by:
Ĥ′m(i) [Expression 65]
an i-th repetition distortion-free signal replica of transmission antenna m:
Ŝm(i)(k) [Expression 66]
is given by the following equation:
In the first embodiment and second embodiment described above, in each repetitive processing operation for antenna interference removal and replica reproduction, each transmission antenna signal is simultaneously demodulated, and antenna interference replicas are eliminated in parallel. A method is also contemplated to order transmission antenna signals in each repetitive processing operation based on a reception quality to, demodulate transmission antenna signals with higher qualities, and sequentially to eliminate antenna interference. Specifically, in first (zero-th repetition) reception processing, transmission antenna signals with higher reception qualities are demodulated in order and interference replica is eliminated from DFT signals. In reception processing in an i-th repetition (i≧1), the difference between a current interference replica and the preceding interference replica is eliminated in the sequence of transmission antenna signal that have higher reception qualities from reception signals in which transmission antenna interference in all previous repetitions has been eliminated from the DFT signal. In this event, while the reception characteristics for the number of repetitions are improved by the effect of the sequential interference removal of transmission antenna signals, the amount of processing increases due to the requirements for antenna interference removal and the weight calculation for two-dimensional frequency domain equalization for each transmission antenna.
In this embodiment, conversion from a time domain signal to a frequency domain signal is performed by DFT, while conversion from a frequency domain signal to a temporal domain signal is performed by IDFT, but fast Fourier transform (FFT), inverse fast Fourier transform (IFFT), or another algorithm may be used.
Number | Date | Country | Kind |
---|---|---|---|
2006-110871 | Apr 2006 | JP | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/JP2007/051754 | 2/2/2007 | WO | 00 | 10/14/2008 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2007/119280 | 10/25/2007 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
20030189999 | Kadous | Oct 2003 | A1 |
20040165675 | Ito et al. | Aug 2004 | A1 |
20080219341 | Kim | Sep 2008 | A1 |
Number | Date | Country |
---|---|---|
2003-348045 | Dec 2003 | JP |
2004-235916 | Aug 2004 | JP |
2005-269383 | Sep 2005 | JP |
2005-328311 | Nov 2005 | JP |
Number | Date | Country | |
---|---|---|---|
20090262853 A1 | Oct 2009 | US |