This application is based upon and claims the benefit of priority from prior Japanese Patent Applications No. 2004-071732, filed Mar. 12, 2004, No. 2004-176096, filed Jun. 14, 2004, No. 2004-235349, filed Aug. 12, 2004, No. 2004-256247, filed Sep. 2, 2004; and No. 2005-052949, filed Feb. 28, 2005, the entire contents of all of which are incorporated herein by reference.
1. Field of the Invention
The present invention relates to an orthogonal frequency division multiplexing (OFDM) signal transmission method and apparatus using a plurality of transmission antennas.
2. Description of the Related Art
Of OFDM signal transmission apparatuses, especially, a system which transmits different data simultaneously through a plurality of transmission antennas can transmit data at high transmission. On the other hand, the data error rate characteristic readily degrades. A method is known, in which on the transmitter side, a signal known to receiver, called a pilot symbol is superposed on one or a plurality of specific subcarriers to form pilot subcarriers. On the receiver side, channel equalization or frequency offset compensation is done for each subcarrier on the basis of the pilot subcarriers, thereby obtaining a decoded signal with an improved data error rate performance.
When a single known signal is transmitted from a plurality of transmission antennas by using pilot subcarriers of the same frequency, as described above, the transmission signals of the respective pilot subcarriers interfere with each other to form directional beams. In the IEEE802.11a standard, there are 52 subcarrier populated and four of them are assigned as pilot subcarrier. However, the directional beam of all the pilot subcarriers are directed in almost the same direction because the interval (about 4.4 MHz) between the pilot subcarriers is smaller than the carrier frequency (5 GHz). In this case, the null points at which the electric fields of the respective directional beams abruptly decrease are also directed in the same direction. For this reason, it is almost impossible to receive the pilot subcarrier in the direction of the null points, and the reception performance drastically degrades.
To cope with this problem, Jpn. Pat. Appln. KOKAI Publication No. 2003-304216 discloses a technique, in which a pilot subcarrier is transmitted from only one transmission antenna, and null signals in the frequency band of the pilot subcarrier are transmitted from the remaining transmission antennas. According to this technique, the problem of mutual interference between pilot subcarriers, which occurs when pilot subcarriers are transmitted from a plurality of transmission antennas, is avoided. Hence, any degradation in reception performance due to directional beam formation can be prevented.
In the method of transmitting a pilot subcarrier from a single transmission antenna, like Jpn. Pat. Appln. KOKAI Publication No. 2003-304216, the total transmission power of the pilot subcarrier is low as compared to a system which transmits pilot subcarriers from a plurality of transmission antennas. This degrades the reception performance of the receiver.
When the transmission power of the pilot subcarrier from the single transmission antenna is made higher than that of the data subcarrier from each transmission antenna, the total transmission power of the pilot subcarrier can be increased, and the reception performance improves. However, when the transmission power of the pilot subcarrier from the single transmission antenna is increased, a variation in transmission power occurs in the frequency band of the OFDM signal. This may result in composite triple beat (CTB) or increase the dynamic range of the transmission signal to make the specifications (especially input dynamic range) of a digital to analog (D/A) converter of the receiver strict.
It is an object of the present invention to provide an OFDM signal transmission apparatus which decreases the composite triple beat without decreasing the transmission power of a pilot subcarrier. It is another object of the present invention to increase the area where high-quality reception is possible.
The first aspect of the present invention provides an orthogonal frequency division multiplexing (OFDM) signal transmission apparatus which transmits OFDM signals by using a plurality of transmission antennas, comprising: a subcarrier setting device configured to set some of subcarriers of the OFDM signals to pilot subcarriers for transmitting pilot signals, and remaining subcarriers thereof to data subcarries for transmitting a data signal, polarities of the pilot subcarriers differing among the transmission antennas.
The embodiments of the present invention will be described below with reference to the accompanying drawing.
As shown in
In this embodiment, as schematically shown in
Let Haa be the transfer function of the channel (the transfer function of the channel will be referred to as a channel response value hereinafter) from the transmission antenna 101a to the reception antenna 201a, Hab be the channel response value from the transmission antenna 101a to the reception antenna 201b, Hba be the channel response value from the transmission antenna 101b to the reception antenna 201a, and Hbb be the channel response value from the transmission antenna 101b to the reception antenna 201b. A reception signal RXa of the reception antenna 201a and a reception signal RXb of the reception antenna 201b are expressed by
where TXa and TXb are transmission signals from the transmission antennas 101a and 101b, respectively. When the reception signals RXa and RXb are multiplied by the inverse matrix of the matrix formed by the channel response values Haa, Hab, Hba, and Hbb, the transmission signals TXa and TXb can be demodulated.
In the first embodiment, a pilot subcarrier to transmit a known signal to be used to compensate for the residual phase error of a frequency offset or clock offset is used independently of a data subcarrier to transmit data. More specifically, in the reception mode, the residual phase error is detected and compensated by using the known signal transmitted by the pilot subcarrier.
A description will be made here for comparison. Jpn. Pat. Appln. KOKAI Publication No. 2003-304216, an OFDM signal shown in
According to the first embodiment, a satisfactory reception characteristic can be obtained while sufficiently ensuring the total transmission power of pilot subcarriers by transmitting them from the two transmission antennas 101a and 101b.
The OFDM signal transmission apparatus 100 shown in
Input transmission data is a wireless packet having a structure to be described later. The transmission data is encoded by the encoder 102. The encoded data is subjected to serial-parallel conversion by the serial-parallel converter 103 and divided into first transmission data corresponding to the transmission antenna 101a and second transmission data corresponding to the transmission antenna 101b. The first and second transmission data are subcarrier-modulated by the modulators 104a and 104b, respectively. As the modulation method of the modulators 104a and 104b, for example, binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), 16 quadrature amplitude modulation (QAM), or 64QAM is used, although the present invention is not limited to these methods.
The modulated data output from the modulator 104a is divided into a plurality of first data subcarriers by the serial-parallel converter 105a. The modulated data output from the modulator 104b is divided into a plurality of second data subcarriers by the serial-parallel converter 105b.
The modulated data (to be referred to as first data subcarriers and second data subcarriers hereinafter) allocated to the first and second data subcarriers are input to the pilot subcarrier insertion unit 106. In the pilot subcarrier insertion unit 106, some of the subcarriers of OFDM signals are allocated to pilot subcarriers to transmit pilot signals, and the remaining subcarriers are allocated to data subcarriers to transmit data signals.
More specifically, the pilot subcarrier insertion unit 106 inserts pilot signals (to be referred to as first pilot subcarriers hereinafter in this specification) to be transmitted by at least one first pilot subcarrier between the first data subcarriers. In addition, the pilot subcarrier insertion unit 106 inserts pilot signals (to be referred to as second pilot subcarriers hereinafter in this specification) to be transmitted by at least one second pilot subcarrier between the second data subcarriers. A set of first data subcarriers and first pilot subcarriers will be referred to as a first subcarrier signal. A set of second data subcarriers and second pilot subcarriers will be referred to as a second subcarrier signal.
The first and second subcarrier signals output from the pilot subcarrier insertion unit 106 are subjected to inverse fast Fourier transform by the IFFT units 107a and 107b, respectively. As a result of inverse fast Fourier transform, the first and second subcarrier signals are multiplexed as they are converted from the signals on the frequency axis to signals on the time axis. Hence, a first OFDM signal a and second OFDM signal b as shown in
The pilot subcarrier insertion unit 106 will be described next with reference to
In the pilot subcarrier insertion unit 106, the first data subcarriers from the serial-parallel converter 105a and the second data subcarriers from the serial-parallel converter 105b are directly output to the IFFT units 107a and 107b. At this time, first pilot subcarriers are inserted between the first data subcarriers, and second pilot subcarriers are inserted between the second data subcarriers.
In this embodiment, four first pilot subcarriers and four second pilot subcarriers are present.
A sequence generator 110 generates a pseudorandom noise (PN) sequence like an M sequence. The first pilot subcarriers are generated by causing multiplying units 111a to 111d to obtain the products between a PN sequence PN(i) and polarity data Sa(j) of the first pilot subcarriers, which are stored in a ROM 121a. The second pilot subcarriers are generated by causing multiplying units 112a to 112d to obtain the products between the PN sequence PN(i) and polarity data Sb(j) of the second pilot subcarriers, which are stored in a ROM 121b. A baseband signal Pa(i,j) of the first pilot subcarriers transmitted from the transmission antenna 101a is expressed by the product of PN(i) and Sa(j), which is given by
Pa(i, j)=PN(i)×Sa(j) (2)
where i is the symbol number in the time domain, and j is the pilot subcarrier number in the frequency domain. Similarly, a baseband signal Pb(i,j) of the pilot subcarriers from the transmission antenna 101b is expressed by the product of PN(i) and Sb(j), which is given by
Pb(i, j)=PN(i)×Sb(j) (3)
In the first embodiment, the number of pilot subcarriers transmitted from each of the transmission antennas 101a and 101b is four (j=1 to 4). The polarity data Sa(j) and Sb(j) (j=1, 2, 3, 4) of the first and second pilot subcarriers transmitted from the transmission antennas 101a and 101b are set as follows.
Sa(1)=1, Sa(2)=1, Sa(3)=1, Sa(4)=−1 (4)
Sb(1)=1, Sb(2)=−1, Sb(3)=1, Sa(4)=1 (5)
That is, the polarity data Sa(j) to be multiplied by the PN sequence in the first pilot subcarriers output from the transmission antenna 101a are different from the polarity data Sb(j) to be multiplied by the PN sequence in the second pilot subcarriers output from the transmission antenna 101b. Accordingly, the polarity pattern of the first pilot subcarriers is different from that of the second pilot subcarriers. The polarity pattern of the first pilot subcarriers indicates the pattern of combination of the polarities of the first pilot subcarriers. The polarity pattern of the second pilot subcarriers indicates the pattern of combination of the polarities of the second pilot subcarriers. The effect obtained by using different polarity patterns for the first and second pilot subcarriers will be described later in detail. The polarity data of each pilot subcarrier and the PN sequence are expressed by real numbers.
Instead, polarity data represented by a complex number or a PN sequence represented by a complex number can also be used.
The OFDM signal reception apparatus 200 shown in
The OFDM signal received by the reception antenna 201a is input to the FFT unit 202a through a wireless reception unit (not shown). The signal is divided into subcarrier signals by Fourier transform. The OFDM signal received by the reception antenna 201b is also subjected to Fourier transform by the FFT unit 202b and divided into subcarrier signals.
As shown in
When the inverse matrix represented by equation (6) is multiplied by received signal vectors generated from the received signals output from the reception antennas 201a and 201b, the OFDM signals from the transmission antennas 101a and 101b are demultiplexed. In a multipath environment, the channel response value changes between the subcarriers. As a result, derivation of the coefficient of the inverse matrix and multiplication of the inverse matrix are executed for each subcarrier. The signals demultiplexed by the interference suppression circuit 203 are sent to the residual phase error detector 204.
The residual phase error detector 204 detects a residual component such as a frequency offset or clock offset compensated by using the preamble of a wireless packet (not shown). The residual phase error detector 204 also detects the residual phase error of the received signals RXa and RXb by using the known signal transmitted by the subcarrier and sends the residual phase error to the phase compensation units 205a and 205b.
As in the first embodiment, when the first and second pilot subcarriers generated from the common PN sequence are transmitted from the transmission antennas 101a and 101b, the combination of reception signal points is (1,1) and (−1,−1), or (1,−1) and (−1,1). This combination does not change during wireless packet reception. For example, when the combination of reception signal points is (1,−1) and (−1,1), it looks for the OFDM signal reception apparatus as if a BPSK signal were transmitted from a single transmission antenna.
A case in which a residual phase error is detected by using pilot subcarriers Pa(1) and Pb(1) transmitted by the (−k+1)th subcarrier will be described next. Only the FFT unit 202a connected to the reception antenna 201a is taken into consideration. Let Haa be the channel response value from the transmission antenna 101a to the reception antenna 201a in the (−k+1)th subcarrier. Let Hba be the channel response value from the transmission antenna 101b to the reception antenna 201a. When the polarities of the pilot subcarriers are represented by equations (4) and (5), polarities corresponding to the (−k+1)th pilot subcarrier are Sa(1)=1 and Sb(1)=1. Since the pilot signal in which the signals from the two transmission antennas are multiplexed is multiplied by the channel response value Haa+Hba, the two points (1,1) and (−1,−1) are received. Hence, the residual phase error detector calculates the channel response value Haa+Hba by using the channel response values Haa and Hba and creates the reference signal points (1,1) and (−1,−1).
Assume that (1,1) is transmitted by the next OFDM symbol, and the received signal point at this time is “next symbol” in
In the first embodiment, residual phase error detection using pilot subcarriers is executed without using interference suppression. Instead, the residual phase error may be detected after interference suppression is executed. In this case, the received signal points of pilot subcarriers appear equal in number to the transmission signal points from the transmission antennas 101a and 101b. When residual phase error detection using pilot subcarriers is executed after interference suppression, the S/N ratio of the pilot subcarriers is low. Hence, the estimation accuracy degrades.
The phase compensation units 205a and 205b execute phase rotation for the reception signals in correspondence with the residual phase error, thereby compensating for the phase. The two reception signals after phase compensation are converted into a serial signal by the serial-parallel converter 206 and decoded by the decoder 207 so that a reception signal corresponding to the transmission signal is obtained.
As described above, the pilot subcarriers are used to detect the residual phase error. If the signal to noise ratio of the pilot subcarriers are low, the performance of the residual phase error detection will degrade. In this case, since the phase compensation units 205a and 205b execute phase compensation on the basis of the wrong residual phase error detection result, all data subcarriers are erroneously received at a high probability. Hence, it's no exaggeration to say that the reception power of pilot subcarriers determines the reception performance of the OFDM signal reception apparatus. To solve this problem, in this embodiment, the first pilot subcarrier transmitted from the transmission antenna 101a and the second pilot subcarrier transmitted from the transmission antenna 101b have different polarities, as described above.
Sb(1)=1, Sb(2)=1, Sb(3)=1, Sb(4)=−1 (7)
The transmission antennas 101a and 101b are assumed to be omni-directional antennas as shown on the upper side of
In accordance with equations (4) and (5), for example, the phase difference between the first pilot subcarrier controlled in accordance with the polarity data Sa(1) and the second pilot subcarrier controlled in accordance with the polarity data Sb(1) is 0. To the contrary, the phase difference between the first pilot subcarrier controlled in accordance with the polarity data Sa(2) and the second pilot subcarrier controlled in accordance with the polarity data Sb(2) is 180°. As a result, as shown in, e.g., the lower side of
As is apparent from
On the receiving side, even when the reception power of one pilot subcarrier is low, the reception powers of the remaining pilot subcarriers are high at a high probability by using the first embodiment. Since the dead zone where the reception powers of all the pilot subcarriers decrease simultaneously can be reduced, the area where high-quality reception is possible widens.
In the first embodiment, it is not necessary to especially increase the transmission powers of subcarriers for all the transmission antennas 101a and 101b since the pilot subcarriers are transmitted from all antennas.
For this reason, the composite triple beat does not increase, and the input dynamic range of the D/A converter need not particularly be widened.
In the above description, the OFDM signal transmission apparatus 100 has the two transmission antennas 101a and 101b. However, the present invention can also be extended to an OFDM apparatus having three or more transmission antennas.
In pilot subcarrier polarity 1 for the four transmission antennas shown in
s(1)=[1, 1, 1, 1]
s(2)=[1, −1, −1, 1]
s(3)=[1, 1, −1,−1]
s(4)=[−1, 1, −1, 1] (8)
In this case, s(1) to s(4) are vectors different from each other. For example, the vector s(1) does not change to another vector even when it is multiplied by scalar value. When the vector of a pilot subcarrier transmitted from a certain frequency is different from that of a pilot subcarrier transmitted from another frequency, the directional beams of the respective pilot subcarriers are directed in different directions. Hence, the dead zone can be reduced. Note that s(1) to s(4) are orthogonal to each other. Even when they are not orthogonal to each other, the directional beams can be directed in different directions.
In pilot subcarrier polarity 2 for the four transmission antennas shown in
In pilot subcarrier polarity 1 for the three transmission antennas shown in
In pilot subcarrier polarity 2 for the three transmission antennas shown in
When the locations of pilot subcarriers are determined in this way, the directional beams of the respective pilot subcarriers are directed in different directions. For this reason, the dead zone can be reduced.
In the complex domain, the pilot subcarrier polarities shown in
where sk(i) is the polarity of the pilot subcarrier, j is the imaginary unit, i is the number of the pilot subcarrier, and k is the antenna number of the transmission antenna. For example, the first element of k represents the signal transmitted from the transmission antenna 101a, and the second element of k represents the signal transmitted from the transmission antenna 101b.
According to equation (9), the phase difference between the first to fourth pilot subcarriers transmitted from the antenna 101b of k=2 is −90°. The phase difference between the pilot subcarriers transmitted from the antenna 101c of k=3 is −180°. The phase difference between the pilot subcarriers transmitted from the antenna 101d of k=4 is −270°. When the transmission antenna changes, the phase difference between pilot subcarriers of the antenna changes. Hence, as described above, since the directional beams corresponding to the respective pilot subcarriers are directed in different directions, the dead zone can be reduced.
The phase difference of −90° equals the phase difference of 270° and the phase difference of −270° equals the phase difference of 90°. To express
According to
[Sk(1), Sk(2), Sk(3), Sk(4)]=[sk(1), sk(2), sk(3), −sk(4)] (10)
In other words, the polarities of the first to fourth pilot subcarriers are given by
More generally, depending on the pilot subcarrier number i, the polarity of a pilot subcarrier is given by
In the above description, i is the number of a pilot subcarrier. However, i may be changed to the frequency of the pilot subcarrier. More specifically, for example, −21, −7, +7, and 21 are used as the value i. In consideration of the periodicity of the Fourier transform function, the pilot subcarrier polarity can also be expressed by
where sk(i) is the polarity of the pilot subcarrier, j is the imaginary unit, i is the frequency of the pilot subcarrier, k is the antenna number of the transmission antenna, and N is the number of input points in inverse Fourier transform.
Even in equation (14), when the transmission antenna changes, the phase difference between the pilot subcarriers changes. Hence, the dead zone can be reduced. In consideration of a Fourier transform pair, the expression of equation (14) is equivalent to cyclic shifting the transmission signal sample by sample for each transmission antenna along the time domains.
Other embodiments of the present invention will be described next. The other embodiments to be described below are different from the first embodiment in the pilot subcarrier insertion unit 106 in the OFDM signal transmission apparatus 100.
A pilot subcarrier insertion unit 106 according to the second embodiment has ROMs 121a and 121b which store polarity data for first and second pilot subcarriers, and a subcarrier pattern controller 122, as shown in
In the second embodiment, the polarities of the first and second pilot subcarriers transmitted from transmission antennas 101a and 101b are not fixed but change for each wireless packet. More specifically, polarity data of different patterns are read out from the ROMs 121a and 121b for each wireless packet and multiplied, by multipliers 111a to 111d and 112a to 112d, by a PN sequence generated by a PN sequence generator 110.
A wireless packet counter 123 arranged outside the pilot subcarrier insertion unit 106 shown in
For example, when a wireless packet is transmitted, the polarity data of the pattern A is read out. When the next wireless packet is transmitted, the polarity data of the pattern B is read out. When the third wireless packet is transmitted, the polarity data of the pattern C is read out. As a result, the polarity pattern of pilot subcarriers is changed for each wireless packet. Polarity data pattern change by the subcarrier pattern controller 122 is done, e.g., at random for each wireless packet.
The polarity data read out from the ROMs 121a and 121b are input to the multipliers 111a to 111d and multiplied by the PN sequence generated by the PN sequence generator 110 so that first and second pilot subcarriers are generated, as in the first embodiment. The generated first and second pilot subcarriers are inserted between first data subcarriers and between second data subcarriers, respectively, so that first and second subcarriers are generated.
When the first and second subcarriers are input to IFFT units 107a and 107b shown in
According to the second embodiment, the pattern of directional beams formed by the antennas 101a and 101b changes between transmission of pilot subcarriers controlled by, e.g., the polarity data of the pattern A and transmission of pilot subcarriers controlled by, e.g., the polarity data of the pattern B.
In an OFDM signal reception apparatus which is placed in a zone where the received power is low for the pilot subcarriers of, e.g., the pattern A, the probability that the received power of pilot subcarriers recovers is high when pilot subcarriers of the pattern B different from the pattern A are transmitted. Hence, when the pilot subcarrier pattern is changed for each wireless packet, the dead zone can be reduced.
The pattern of polarity data of pilot subcarriers need not always be changed at random. For example, of various patterns of polarity data of pilot subcarriers, a pattern which ensures a satisfactory reception performance may be stored in correspondence with each OFDM signal reception apparatus. In this case, pilot subcarriers can be transmitted by using the pattern stored in correspondence with the OFDM signal reception apparatus at the transmission destination.
The third embodiment of the present invention will be described next. In the third embodiment, the polarity pattern of pilot subcarriers is changed only when an error has occurred in a preceding transmitted wireless packet, and the packet is for retransmission. As shown in
As shown in
A transmission signal input to an encoder 102 shown in
In the third embodiment, whether the wireless packet is a retransmission packet is determined by analyzing the retransmission field. Instead, an upper-level layer (medium access control (MAC) layer in the IEEE 802.11a standard) which executes wireless access control may directly notify the subcarrier pattern controller that the wireless packet is for a retransmission.
As described above, in the third embodiment, the retransmission packet is transmitted to the same transmission partner by using a polarity pattern different from that of the pilot subcarriers of the wireless packet preceding transmitted from an OFDM signal transmission apparatus 100. As a result, the pattern of directional beams formed by the plurality of transmission antennas changes in retransmission. Hence, the OFDM signal reception apparatus can correctly receive the retransmitted packet at a high probability.
A pilot subcarrier insertion unit 106 according the fourth embodiment of the present invention includes two PN sequence generators 110a and 110b, as shown in
The polarity patterns of the first and second subcarriers may be either the same or different. A case in which the same polarity data S are used for the first and the second subcarriers, will be described here. The first pilot subcarriers are modulated as follows in accordance with the PN sequence PNa and pilot subcarrier polarity data S.
Pa(i, j)=PNa(i)×S(j) (15)
The second pilot subcarriers are modulated as follows in accordance with the PN sequence PNb and pilot subcarrier polarity data S.
Pb(i, j)=PNb(i)×S(j) (16)
In the fourth embodiment, for example, a case in which the combination of (1,−1) and (−1,1) is transmitted and a case in which the combination of (1,1) and (−1,−1) is transmitted can be considered. In the former combination, the phase difference between the signals from the transmission antennas 101a and 101b is 180°. In the latter combination, the phase difference is 0°. Since the directional beam of transmission changes between the transmission of the former combination and the transmission of the latter combination, the reception power changes. As described in the first embodiment, the four candidate reception signal points (1,−1), (−1,1), (1,1), and (−1,−1) can be obtained by combining the channel response from the transmission antennas to the reception antennas with the signals transmitted by the pilot subcarriers.
The residual phase error measuring method will be described next. Assume that the received symbol of a pilot subcarrier is (1,1). In this case, in the first embodiment, (1,1) is transmitted again as the next symbol, or (−1,−1) is transmitted. It looks on the receiving side as if a BPSK signal were received from a single antenna. Hence, the reception power does not change.
In the fourth embodiment, (−1,1) can also be transmitted as the next symbol. Hence, the received symbol containing a phase error can be “next symbol 1” or “next symbol 2” shown in
Phase compensation units 205a and 205b multiplex the phase rotation for the reception signals in correspondence with the residual phase error, thereby compensating for the phase. The two received signals after phase compensation are converted into a serial signal by a serial-parallel converter 206 and decoded by a decoder 207 so that a reception signal corresponding to the transmission signal is obtained. In the fourth embodiment, the residual phase error is measured before interference suppression. The residual phase error can also be detected after interference suppression. In this case, a pilot subcarrier transmitted from a single antenna appears as the output after interference suppression. Hence, only two reception signal points appear.
As described above, according to the fourth embodiment, as shown in
As shown in
Pa(i, j)=PNa(i)×Sa(j) (17)
Pb(i, j)=PNb(i)×Sb(j) (18)
In this case, pilot subcarriers shown in
In a pilot subcarrier insertion unit 106 according to the sixth embodiment of the present invention, as shown in
Pa(2i−1, j)=PN(2i−1) (19)
Pa(2i, j)=−PN(2i)* (20)
Pb(2i−1, j)=PN(2i) (21)
Pb(2i, j)=−PN(2i−1) (22)
where * is the complex conjugate. As indicated by equations (19) to (22), the PN sequences are transmitted by using transmission diversity using two transmission antennas 101a and 101b and two symbols. The transmission diversity method represented by equations (19) to (22) is the same as that disclosed in U.S. Pat. No. 6,185,258B1.
The detailed pilot subcarrier signals given by equations (19) to (22) are applied to the (−k+1)th subcarrier and (k−4)th subcarrier shown in
Pa(2i−1, j)=PN(2i (23)
Pa(2i, j)=−PN(2i−1) (24)
Pb(2i−1, j)=PN(2i−1) (25)
Pb(2i, j)=−PN(2i)* (26)
The detailed pilot subcarriers given by equations (23) to (26) are applied to the (−k+4)th subcarrier and (k−1)th subcarrier shown in
As is apparent from equations (19) to (26), in this embodiment, transmission diversity is executed by using a two-OFDM symbol duration. No diversity gain is obtained when only one symbol is received.
As described above, according to the sixth embodiment, pilot subcarriers are transmitted by using transmission diversity. Hence, the residual phase error can accurately be detected, and the performance at the receiver can be improved.
A pilot subcarrier insertion unit 106 according to the seventh embodiment of the present invention includes subcarrier locating devices 126a and 126b to make the locations of pilot subcarriers and data subcarriers change between transmission antennas 101a and 101b, as shown in
Modulated signals for the transmission antenna 101a, which are obtained by multiplying a PN sequence by polarities Sa(1) to Sa(4), are input to the subcarrier locating device 126a as pilot subcarriers Pa(1) to Pa(4). The subcarrier locating device 126a rearranges the data subcarriers and pilot subcarriers and inputs them to an IFFT unit 107a. The processing of signals for the transmission antenna 101b is also the same, and a description thereof will be omitted.
Since data signals can be assumed to be random, the correlation between the first data subcarriers transmitted from the transmission antenna 101a and the second data subcarriers transmitted from the transmission antenna 101b is generally low. For this reason, the phase difference between the subcarriers transmitted from the transmission antenna 101a and those transmitted from the transmission antenna 101b changes between the −kth subcarrier and the (−k+2)th subcarrier. Hence, the directional beam of the pilot subcarrier transmitted by the kth subcarrier is different from the directional beam of the pilot subcarrier transmitted by the (−k+2)th subcarrier at a high probability.
According to the seventh embodiment, since the probability that the reception powers of all the pilot subcarriers decrease simultaneously due to the influence of null points is very low, no dead zone is formed. In addition, even when the power of pilot subcarriers happens to be low in one symbol interval, the data signal changes at a high probability between the current symbol and the next symbol. Hence, the reception power of the pilot subcarriers can recover in the next symbol at a high probability. As described above, according to the seventh embodiment, the pilot subcarrier reception probability can be increased, and the dead zone can be reduced.
The pilot subcarrier location can also be changed in wireless packets. In, e.g., the IEEE 802.11a standard, a pilot signal to estimate the channel response of all subcarriers is inserted in the unique word shown in
According to the present invention, compensation of the residual phase error can accurately be executed even when a wireless packet containing both a portion to transmit data from a single antenna and a portion to transmit data from a plurality of antennas is received. According to a wireless communication preamble signal system proposed in Jan Boer et al, “Backwards Compatibility”, IEEE 802.11-03/714r0, as shown in
The wireless communication preamble signal shown in
In the IEEE 802.11a wireless packet shown in
Detailed control in receiving the wireless packet shown in
The receiver which has received the short preamble x01 detects the start of the long preamble sequence x02 by using an AGC and time synchronization means (not shown) to detect an FFT window. Simultaneously, estimation and compensation of the frequency offset are done. The receiver which has received the long preamble sequence x02 measures the channel responses of all subcarriers by using known pilot subcarriers. Especially, the channel response of a pilot subcarrier is transferred to the residual phase error detector 204. The above processing can be implemented by a known technique, and a description thereof will be omitted.
Next, the SIGNAL field x03 is received. The SIGNAL field is subjected to FFT by FFT units 202a and 202b. The FFT output is input to an interference suppression circuit. However, since the SIGNAL field is output from a single antenna, interference suppression need not be performed. Hence, the processing executed by the interference suppression circuit is processing for multiplying a unit matrix or processing for weighting and combining the outputs from the FFT units 202a and 202b to increase the Signal to Noise ratio. The output from the interference suppression circuit is input to the residual phase error detector 204.
When the received point of the pilot subcarrier of the SIGNAL portion is “next symbol in single antenna transmission” shown in
The demodulator 207 demodulates the SIGNAL2 portion and transfers the decoding result to the SIGNAL analysis unit 208. The SIGNAL analysis unit 208 analyzes the second signal field, analyzes the number of multiplexed transmission antennas, and the information that there will be a transmission from plurality of antennas after the long preamble. The SINGAL analysis unit 208 transfers the analyzed result to the residual phase error detector 204.
Next, the receiver receives the long preamble from the transmission antenna 101b and measures the channel response from the transmission antenna 101b. A case in which the DATA portions x08 and x09 are received will be described next. In the following description, signals from the two transmission antennas 101a and 101b are multiplexed in the DATA portions, and the multiplexed pilot subcarriers have polarities represented by equations (4) and (5). As described above, the number of multiplexed transmission antennas can be recognized by the signal from the SIGNAL analysis unit 208.
Of the four pilot subcarriers, pilot subcarriers having the leftmost polarities in equations (4) and (5) will be focused. The polarities of pilot subcarriers transmitted by this frequency are Sa(1)=1 and Sb(1)=1. When the channel response from the transmission antenna 101a to the reception antenna 201a is Haa, and the channel response from the transmission antenna 101b to the reception antenna 201a is Hba, the channel response value of a pilot subcarrier received by the DATA portion is Haa+Hba shown in
More specifically, since the pilot subcarriers are transmitted by BPSK signals modulated by a PN sequence, the points (1,1) and (−1,−1) shown in
As described above, the operation of the residual phase error detector 204 must be switched between single antenna transmission and multiple antenna transmission. In the eighth embodiment, this operation is implemented by analyzing the second signal field x04 (SIGNAL2) shown in
In the above embodiments, OFDM signal transmission apparatuses which transmit different signals in correspondence with transmission antennas. However, the present invention can also be applied to an OFDM signal transmission apparatus using a transmission method of transmitting different signals in correspondence with a plurality of transmission beams.
According to the ninth embodiment, subcarriers obtained by inserting pilot subcarriers between data subcarriers by the pilot subcarrier insertion unit 106 are input to a beam forming device 108. The beam forming device 108 weights and combines the outputs from the serial-parallel converter 105a and the outputs from the serial-parallel converter 105b and outputs the signals to IFFT units 107a and 107b. The outputs after IFFT are output from transmission antennas 101a and 101b.
The beam forming device 108 is a device which executes processing for forming (beam forming) a plurality of transmission beams. The beam forming device 108 can be implemented by using a known technique. Referring to
Known beam forming schemes are roughly classified into two shames. In the first beam forming mode, the channel response between a radio transmission apparatus and a radio reception apparatus is completely known. A beam forming weight is calculated in accordance with the channel response in order to steer the beam toward the reception apparatus. In the second beam forming mode, the channel response is not completely known and a predetermined beam forming weight is used. Therefore, the beam may not be steered toward the reception apparatus. John Ketchum et al have described the Eigenvector steering (ES) scheme as one of the first beam forming schemes and the Spatial spreading (SS) scheme as one of the second beam forming schemes in reference “ftp://ieee:wireless@ftp.802wirelessworld.com/11/04/11-04-0870-00-000n-802-11-ht-system-description-and-operating-principles.doc.” According to a description in the reference by John Ketchum et al, the diversity effect can be obtained by using beam forming.
The beam forming device 108 has, e.g., the ES scheme and SS scheme to cope with the first beam forming scheme and second beam forming scheme. For data subcarriers, beam forming can effectively be done by using either the ES scheme or SS scheme, as described by John Ketchum et al. On the other hand, for pilot subcarriers, beam forming by the method instructed by John Ketchum et al is not preferable. This is because the pilot subcarriers are known information on the receiving side and must always correctly be received on the receiving side.
Pilot subcarrier forming methods using the ES scheme and SS scheme will be described next.
In the ES scheme, the channel response between a radio transmission apparatus and a radio reception apparatus is measured, and the transmission beam is directed in the direction of the radio reception apparatus, as described in p. 10 of the reference by John Ketchum et al. Since the pilot signal arrives at the radio reception apparatus regardless of pilot subcarriers generated by the pilot subcarrier insertion unit 106, no dead zone is formed. Hence when a beam should be formed by using the ES scheme, the same setting as in the first to eighth embodiment is done for the polarities of pilot subcarriers. On the other hand, the SS scheme is a beam forming scheme when the channel response is not completely known and the predetermined value is used as a beam forming weight. Hence, it cannot guarantee that the pilot subcarriers should always arrive at the radio reception apparatus.
Pilot subcarriers when the number of transmission antennas is 4 is represented by a matrix P.
The number of columns of the matrix P equals the number of pilot subcarriers on the frequency axis. The first row of the matrix P represents the polarity of the first pilot subcarrier transmitted from each antenna. The number of rows of the matrix P equals the number of input ports of the beam forming device 108. That is, the matrix P represents
The number of columns of the matrix Q equals the number of input ports of the beam forming device 108, i.e., the number of transmission beams. The number of rows of the matrix P equals the number of transmission antennas. The i-th row of the matrix Q represents the weight corresponding to the i-th transmission beam (in this case, i=1, 2, 3, 4). The beam forming device 108 multiplies the matrix P by the matrix Q to obtain a matrix QP given by
The number of columns of the matrix QP equals the number of pilot subcarriers on the frequency axis. The number of rows of the matrix QP equals the number of transmission antennas. For example, the first column of the matrix QP will be focused. This corresponds to transmitting a pilot with a power of “16” (i.e., amplitude is 4) from only the first transmission antenna. The second column of the matrix QP will be focused. This corresponds to transmitting pilot signals with a power of “8” (i.e., amplitude is root 8) from only the third and fourth transmission antennas. As described above, especially for the first pilot subcarrier, it will be transmitted with high power only from the first transmission antenna. Regarding the second pilot subcarrier, a pilot signal is transmitted only from the third and fourth transmission antennas. As a result, the transmission power varies in the frequency band of the OFDM signal. As a result, the problems pointed out in the prior art, i.e., composite triple beat and an increase in dynamic range of the transmission signal occur.
More specifically, when beam forming based on the SS scheme is applied to pilot subcarriers, the orthogonality of the pilot subcarriers is damaged. For certain pilot subcarriers, some transmission antennas transmit no pilot signals. Hence, the space diversity effect decreases.
In the ninth embodiment, when the beam forming device 108 uses beam forming based on the SS scheme, the SS scheme is used for data subcarriers, as described in the reference by John Ketchum et al. However, no beam forming is executed for pilot subcarriers. More specifically, for pilot subcarriers, the matrix Q should be the identify matrix Q′ as follows.
The first row of the matrix Q′ represents the weight corresponding to the i-th transmission beam (in this case, i=1, 2, 3, 4). In this case, since the matrix P directly appears in the matrix Q′P, pilot signals are transmitted by pilot subcarriers which are orthogonal to each other on the frequency axis and spatial axis.
When SS scheme is used, cyclic delayed diversity (CDD) is applied to the transmission signal from each antenna, respectively. Specifically, the transmission signal from the first transmission antenna is generated to the transmission signal from the second transmission antenna by 50 nsec cyclic shifting. Since the CDD scheme is described in the reference by John Ketchum et al, explanation is omitted.
As described above, in the ninth embodiment, when beam forming is executed in accordance with the second beam forming mode and, for example, the SS scheme without using channel response information, pilot signals are transmitted without applying beam forming to pilot subcarriers. As a result, pilot signal transmission from only a single antenna can be prevented, and any increase in dynamic range of the transmission signal can be prevented. In addition, any dead zone on the receiving side can be prevented by the space diversity effect.
In the above-mentioned description, beam forming not applied to the pilot subcarrier in the second beam formation mode (multiplied by the identify matrix). However, for example, when beam forming matrix Q′ which multiplied to matrix P with the beam formation device 107 is assumed given by
Matrix Q′P given by
From each antenna, the pilot subcarrier which orthogonal to each other on the frequency axis and the space axis is transmitted so that matrix Q′P may show. This is a polarity pattern of the pilot subcarrier shown in
Thus, in the second mode that is the mode which is not based on the channel response but forms the transmission beam, by making the wait of the transmission beam to the data subcarrier, and the wait of the transmission beam to the pilot subcarrier into a different value. It is avoided that a pilot signal inclines and is transmitted from single antenna, and since the pilot subcarrier which are orthogonal to each other on the frequency axis and the space axis is transmitted from each antenna, it is avoidable that the dead zone is generated in the receiving side.
In this embodiment, the polarity pattern of pilot subcarriers is fixed in a wireless packet for the descriptive convenience. Even when the polarity pattern is changed in a wireless packet, i.e., for each OFDM symbol, the effect of the present invention can be obtained. As shown in
The 10th embodiment of the present invention will be described next. This embodiment is different from the first to ninth embodiments in that the polarity of a pilot subcarrier changes for each OFDM symbol, and the frequency of the pilot subcarrier, i.e., the position of the pilot subcarrier periodically changes. When the position of a pilot subcarrier periodically changes, the channels of all subcarriers can be estimated during the data portions.
In this embodiment, the pilot subcarriers are orthogonal to each other in a plurality of unit times between transmission antennas. In the example shown in
By using the same description as in equation (9), the polarity of each pilot subcarrier shown in
where Sppos(i,t),k,t indicates that the (Sppos(i,t))th subcarrier is used as a pilot subcarrier, k is the transmission antenna number, and t is the OFDM symbol number. In equation (27), the right-hand member of equation (9) is multiplied by polarity data Prot(i,t) given by
where Ppos(i,t) is a conversion function to convert the subcarrier number from the pilot subcarrier number i and OFDM symbol number t. More specifically, Ppos(i,t) can be expressed by
As is apparent from equation (29), the position (frequency) of the pilot subcarrier changes in accordance with the period of equation (28). For example, in subcarriers transmitted from the first to fourth transmission antennas, first, pilot subcarriers are located at positions corresponding to the subcarrier numbers of −20, −7, +7, and +20 in the first 4-OFDM-symbol interval (interval from DATA1 to DATA4) where t=1 to 4.
For example, the pilot subcarriers located at the subcarrier number of −20 will be focused. The polarities of pilot subcarriers transmitted from the first transmission antenna are “1, 1, 1, 1”. The polarities of pilot subcarriers transmitted from the second transmission antenna are “1, j, −1, −j”. The polarities of pilot subcarriers transmitted from the third transmission antenna are “1, −1, 1, −1”. The polarities of pilot subcarriers transmitted from the fourth transmission antenna are “1, −j, −1, j”. The pilot subcarriers transmitted from the first to fourth transmission antennas are orthogonal to each other in the 4-OFDM-symbol interval. This also applies to the pilot subcarriers located at the remaining subcarrier numbers of −7, +7, and +20. The pilot subcarriers are orthogonal to each other in the 4-OFDM-symbol interval.
In the next 4-OFDM-symbol interval (interval from DATA5 to DATA8) where t=5 to 8, the pilot subcarriers at the subcarrier numbers of −17, −4, +10, and +23 are transmitted. The polarities of the pilot subcarriers transmitted from the first to fourth transmission antennas are the same as in the first 4-OFDM-symbol interval where t=1 to 4 so that the pilot subcarriers are orthogonal to each other.
As shown in
The period of equation (28) need not always equal the change period of the pilot subcarrier position indicated by equation (29). The period of equation (29) only needs to be an integer multiple of the period of equation (28). Equation (29) is generated from Fourier coefficients. They only need to be sequences orthogonal with respect to the time axis t.
A pilot subcarrier insertion unit 106 according to this embodiment will be described next with reference to
First pilot subcarriers generated by causing multiplying units 111a to 111d to obtain the products between a PN sequence from a PN sequence generator 110 and polarity data Sa(j) of the first pilot subcarriers, which are stored in a ROM 121a, are input to the subcarrier locating device 126a. Similarly, second pilot subcarriers generated by causing multiplying units 112a to 112d to obtain the products between the PN sequence from the PN sequence generator 110 and polarity data Sb(j) of the second pilot subcarriers, which are stored in a ROM 121b, are input to the subcarrier locating device 126b. Prot(i,t) represented by equations (27) to (29) are stored in the ROMs 121a and 121b as the polarity data Sa(j) and Sb(j).
The processing for changing the polarities or positions of pilot subcarriers for each OFDM symbol is executed in the following manner. In transmitting DATA1, a counter 131 counts a clock signal 130 of the OFDM symbol period and recognizes that it is time to transmit DATA1. The counter 131 outputs, to a subcarrier locating controller 132 and pilot pattern controller 133, instructions to control the subcarrier locating and pilot pattern for DATA1. In transmitting DATA2, the counter 131 recognizes by the clock signal 130 that it is time to transmit DATA2. The counter 131 outputs, to the subcarrier locating controller 132 and pilot pattern controller 133, instructions to control the subcarrier locating and pilot pattern for DATA2. More specifically, the pilot pattern controller 133 reads out values according to equations (27) and (28) from the ROMs 121a and 121b. The subcarrier locating controller 132 inserts pilot subcarriers at positions according to equation (29). In this way, the pilot subcarriers shown in
An OFDM signal reception apparatus having a function of receiving the pilot subcarriers indicated by equation (27) and detecting a residual phase error will be described next with reference to
An OFDM signal received by a reception antenna 201a is input to an FFT unit 202a through a radio reception unit (not shown) and divided into subcarrier signals by Fourier transform. An OFDM signal received by a reception antenna 201b is also subjected to Fourier transform by an FFT unit 202b and divided into subcarrier signals by Fourier transform.
Of the subcarriers output from the FFT units 202a and 202b, data subcarriers are input to an interference suppression circuit 203, and pilot subcarriers are input to a residual phase error detector 204 and channel variation detector 210. The residual phase error detected by the residual phase error detector 204 is compensated by phase compensators 205a and 205b. The channel variation detected by the channel variation detector 210 is compensated by channel variation compensators 211a and 211b. Elements except the residual phase error detector 204, channel variation detector 210, and channel variation compensators 211a and 211b are the same as in the OFDM reception apparatus shown in
In this embodiment, the phase compensator 204 and channel compensators 211a and 211b are independently arranged. These devices can be regarded as devices to compensate the distortion of the signals. Hence, the phase compensator 204 and channel compensators 211a and 211b can be implemented by a single compensator. In this embodiment, the phase compensator 204 and channel compensators 211a and 211b are arranged on the output side of the interference suppression circuit 203. Instead, they can also be arranged on the input side.
The residual phase error detector 204 detects the residual phase error by using pilot subcarriers. The polarity data of pilot subcarriers in each OFDM symbol and the PN sequence to be multiplied by the polarity data are known in the OFDM signal reception apparatus. Hence, the residual phase error detector 204 can detect the residual phase error by using one of the detection principle shown in
Propagation path estimation using pilot subcarriers will be described next. For example, the 20th pilot subcarrier will be described. Let Haa be the channel response value from the transmission antenna 101a to the reception antenna 201a, Hba be the channel response value from the transmission antenna 101b to the reception antenna 201a, Hca be the channel response value from a transmission antenna 101c (not shown) to the reception antenna 201a, and Hda be the channel response value from a transmission antenna 101d (not shown) to the reception antenna 201a.
In this case, when noise components are moved, signals receives in the intervals of DATA1, DATA2, DATA3, and DATA4 can be given by
rDATA1=Haa+Hba+Hca+Hda (30)
rDATA2=(Haa+jHba−Hca−jHda)exp(jθ) (31)
rDATA3=(Haa−Hba+Hca−Hda)exp(j2θ) (32)
rDATA4=(Haa−jHba−Hca+jHda)exp(j3θ) (33)
These are signals before the residual phase error is removed, and exp(jnθ) (n=1, 2, 3) represents the residual phase error. The channel estimation value of Haa can be obtained by executing processing given by
Since the phase difference between several symbols represented by the residual phase error exp(jnθ)(n=1, 2, 3) can be assumed to be sufficiently close to 1, the estimation value of Haa can be obtained by equation (34). This is because the pilot subcarriers are transmitted at a period (4-OFDM-symbol interval in this embodiment) at which the pilot subcarriers are orthogonal to each other, as shown in
The estimation value of Hba can also be obtained by
Ĥba=rDATA1−jrDATA2−rDATA3+jrDATA4 (35)
Generally, when calculation using the complex conjugate of the sequences shown in
In the above description, the channel variation value is estimated by using signals before residual phase error compensation. The channel variation can also be estimated by using signals after residual phase error compensation. In this case, since the residual phase error components in equations (30) to (33) can be neglected, the channel variation value can accurately be obtained.
As described above, in this embodiment, the positions of pilot subcarriers are changed for each orthogonal interval of the pilot subcarriers. The orthogonal interval indicates an interval where the pilot subcarriers are orthogonal to each other between the transmission antennas and 4-OFDM-symbol interval in the example shown in
In the above description, since the number of transmission antennas is four, sequences (e.g.,
As described above, according to the 10th embodiment, pilot subcarriers are orthogonal to each other in a plurality of unit times between a plurality of transmission antennas. In addition, the frequencies (positions) of the pilot subcarriers are changed at a period corresponding to an integer multiple of the unit time. Furthermore, the pilot subcarriers are orthogonal to each other even on the frequency axis. Hence, directional patterns formed by the transmission antennas in transmitting pilot subcarriers can variously be changed.
As in the first to ninth embodiments, the dead zone where the received powers of all four pilot subcarriers decrease at the same time, can be reduced. The area where high-quality reception is possible can be widened. Simultaneously, any composite triple beat at the transmitter can be prevented. On the receiving side, the channel estimation value corresponding to each transmission antenna can easily be obtained during the data portions.
In this embodiment, subcarriers to transmit pilot signals change for each time. Even when the subcarriers to transmit pilot signals decrease due to fading to make it difficult to measure the residual phase error, pilot signals are transmitted from other subcarriers in the next OFDM symbol. Hence, the pilot signals can accurately be received at a high probability during the data portions.
According to the present invention, even when the received power of a pilot subcarrier is low, the received power of another pilot subcarrier becomes high at a high probability. Hence, the dead zone where the reception powers of all pilot subcarriers decrease simultaneously can be reduced, and the area where high-quality reception is possible widens. In addition, since the transmission powers of pilot subcarriers from the transmission antennas are the same as those of data subcarrier, any composite triple beat can be avoided. In addition, the input dynamic range of the D/A converter need not particularly be wide.
Additional advantages and modifications will readily occur to those skilled in the art. Therefore, the invention in its broader aspects is not limited to the specific details and representative embodiments shown and described herein. Accordingly, various modifications may be made without departing from the spirit or scope of the general inventive concept as defined by the appended claims and their equivalents.
Number | Date | Country | Kind |
---|---|---|---|
2004-071732 | Mar 2004 | JP | national |
2004-176096 | Jun 2004 | JP | national |
2004-235349 | Aug 2004 | JP | national |
2004-256247 | Sep 2004 | JP | national |
2005-052949 | Feb 2005 | JP | national |
Number | Name | Date | Kind |
---|---|---|---|
7263133 | Miao | Aug 2007 | B1 |
20020136176 | Abeta et al. | Sep 2002 | A1 |
20030016621 | Li | Jan 2003 | A1 |
20050152314 | Sun et al. | Jul 2005 | A1 |
20070237262 | Hwang et al. | Oct 2007 | A1 |
Number | Date | Country |
---|---|---|
2002-374224 | Dec 2002 | JP |
2003-283441 | Oct 2003 | JP |
2003-304215 | Oct 2003 | JP |
2003-304216 | Oct 2003 | JP |
2004-173258 | Jun 2004 | JP |
2004328464 | Nov 2004 | JP |
2004039022 | May 2004 | WO |
Number | Date | Country | |
---|---|---|---|
20050201268 A1 | Sep 2005 | US |