DESCRIPTION TRANSMISSION DEVICE, RECEPTION DEVICE, COMMUNICATION SYSTEM; TRANSMISSION METHOD AND RECEPTION METHOD Field [0001] The present invention relates to a transmission device to perform transmission diversity transmission, a reception device, a communication system, a transmission method and a reception method. Background [0002] In a communication field, specifically in a wireless communication field, SCBT (single-carrier block transmission) in which DFT (discrete Fourier transform) and IDFT (inverse DFT) are performed on a transmission side or a reception side and a single carrier signal is processed in a block unit is actively considered. The SCBT realizes high frequency utilization efficiency equivalent to that of a multi-carrier transmission system represented by OFDM (orthogonal frequency division multiplexing). Moreover, the SCBT has low peak power and high transmission power efficiency since a base of the SCBT is a single carrier signal. Note that in the following, PAPR (a peak-to-average power ratio) that is a ratio of a peak to average power is an index of a peak power characteristic. Also, the SCBT has multi-path resistance equivalent to that of OFDM since FDE (frequency-domain equalization) is performed on a reception side to thereby compensate for frequency distortion due to a multi-path transmission channel. [0003] The SCBT is also called SC-FDE (single carrier-FDE), SC-FDMA (single carrier-frequency division multiple access), SC-OFDM (single carrier-OFDM), or DFT-Spread OFDM. [0004] On the other hand, in a wireless communication system including a plurality of transmission antennas, called a MISO (a multiple-input single-output) system or a MIMO (multiple-input multiple-output) system, a several transmission diversity methods to improve transmission quality are proposed. In the following, a conventional transmission diversity method that can be applied to SCBT will be described. In order to simplify a description, a transmission diversity method of transmitting one signal stream from two transmission antennas that are a transmission antenna #1 and a transmission antenna #2 will be described as an example in the following description. However, a similar technology is also disclosed in a case where three or more transmission antennas are included. [0005] As a simple transmission diversity method, a delay diversity technology is disclosed in Non Patent Literature 1. In the delay diversity technology disclosed in Non Patent Literature 1, a signal stream is transmitted from a transmission antenna #1 and a signal stream identical to the signal stream transmitted from the transmission antenna #1 is delayed and transmitted from a transmission antenna #2. Also, in Patent Literature 1, a CDD (cyclic delay diversity) technology in which a signal stream transmitted from a transmission antenna #2 is subjected to a cyclic delay in an IDFT block and transmitted is disclosed. When the delay diversity technology or the cyclic delay diversity technology is used, it is observed on a reception side that signal streams equivalently pass through a multi-path transmission channel with a delay. Thus, a multi-path diversity effect is acquired. Also, since a signal transmitted from a transmission antenna #2 is the same as a signal transmitted from a transmission antenna #1 and is only delayed, there is an advantage that transmission power efficiency can be kept without deterioration of a PAPR characteristic of a transmission signal waveform. [0006] In Non Patent Literature 2, an STBC technology using STBC (space-time block code) is disclosed. In the STBC technology disclosed in Non Patent Literature 2, temporally-successive two blocks are a time block #1 and a time block #2. In the time block #1, signal streams are spread to two transmission antennas and are transmitted simultaneously. In the time block #2, the signals transmitted in the time block #1 are switched between the transmission antennas, and complex conjugate and sign inversion of a positive/negative sign with respect to the one transmission antenna only are performed. The above-described signal processing in the transmission means that orthogonal coding is performed in two dimensions of time and space on a transmission side and is generally called Alamouti coding. On a reception side, it is possible to easily perform decoding by performing linear combination of two received time blocks by using transmission channel information. Accordingly, diversity gain for the number of transmission antennas, that is, transmission full diversity is acguired. [0007] In Non Patent Literature 3, unlike the above-described STBC that is orthogonally coded in two dimensions of space and time, an SFBC technology using SFBC (space-frequency block code) that performs similar coding in two dimensions of space and a frequency is disclosed. On the premise of OFDM and on the assumption that a freguency variation in a transmission channel between adjoining two sub-carriers can be ignored, in the SFBC technology, signal switching, complex conjugate, and sign inversion are performed in two transmission antennas between two adjoining sub-carriers in one time block. Accordingly, in the SFBC technology, there is an advantage that transmission full diversity is acquired similarly to the STBC technology and that time variation resistance of a transmission channel is high compared to the STBC technology since the coding is performed in the,one time block. This SFBC technology can be applied not only to OFDM but also to the SCBT technology. [0008] In Non Patent Literature 4, with respect to the SFBC technology disclosed in Non Patent Literature 3, an SFBC technology for SCBT with which technology PAPR of a transmission time waveform is not deteriorated at all by successful utilization of a property of discrete Fourier transform is disclosed. Also, in Patent Literature 2, in an SFBC technology similar to that in Non Patent Literature 4, a technology of coding in a time domain is disclosed. Citation List Patent Literature [0009] Patent Literature 1: Japanese Patent No. 4988137 Patent Literature 2: Japanese Patent No. 5106250 Non Patent Literature [0010] Non Patent Literature 1: Y. Li, J.C. Chuang, and N. R. Sollenberger, "Transmitter Diversity for OFDM Systems and Its Impact on High-Rate Data Wireless Networks," IEEE J Sel. Areas Commun., VOL. 17, NO. 7, pp. 1233-1243, JULY 1999. Non Patent Literature 2: S. M. Alamouti, "A Simple Transmit Diversity Technique for Wireless Communications," IEEE J. Sel. Areas Commun., VOL. 16, NO. 8, pp. 14 51-1458, OCTOBER 1998. Non Patent Literature 3: 3GPP, Rl-071607, Ericsson, "Tx Diversity in LTE DL," March 2007. Non Patent Literature 4: C. Ciochina, D. Castelain, D. Mottier, and H. Sari, ,vNew PAPR-Preserving Mapping Methods for Single-Carrier FDMA with Space-Frequency Block Codes," IEEE Trans. Wirel. Commun., VOL. 8, NO. 10, pp. 5176-5186, OCTOBER 2009. Summary Technical Problem [0011] However, originally, both of a delay diversity technology and a cyclic delay diversity technology disclosed in Non Patent Literature 1 and Patent Literature 1 are to artificially generate a multi-path causing interference at a reception point. Thus, there is a problem in that acquired diversity gain is small compared to a system of performing spatial coding such as the STBC technology. [0012] Also, the STBC technology described in Non Patent Literature 2 is a technology on the premise that there is no time variation in a transmission channel state between transmission and reception in a transmission period of two successive time blocks and there is a problem in that a transmission characteristic is deteriorated in a case where there is a large time variation due to high-speed movement or the like. [0013] Also, the SFBC technology disclosed in each of Non Patent Literature 3, Non Patent Literature 4, and Patent Literature 2 performs coding between sub-carriers away from each other in a frequency domain and performs coding between sub-carriers away from each other for about a half of a signal band at maximum. Thus, there is a problem in that a transmission characteristic is greatly deteriorated in a case where a frequency variation in a transmission channel cannot be ignored. [0014] Also, the SFBC technology disclosed in Non Patent Literature 3 has an advantage that resistance to a frequency variation in a transmission channel is high since SFBC coding is performed between adjoining sub-carriers. However, since operations of signal switching, complex conjugate and sign inversion are performed for each sub-carrier in a frequency domain, there is a problem in that PAPR of a transmission signal is deteriorated and transmission power efficiency is decreased. [0015] The present invention is provided in view of the forgoing and is to acquire a transmission device, a reception device, a communication system, a transmission method and a reception method that can prevent a decrease in efficiency of transmission power while preventing deterioration of a transmission characteristic even in an environment in which a transmission channel temporally varies at high speed. Solution to Problem [0016] To solve the above problem and the object, the present invention provide a transmission device comprising: two or more transmission antennas; and a coding unit to perform space frequency block coding of a signal stream of a first time length, wherein the coding unit includes a signal dividing unit to equally divide the signal stream into sub-streams of second time lengths and perform a cyclic shift of one or more sub-streams among the sub-streams in a time length in which a value divided by a primary modulation symbol interval of the sub-streams beeomes a non-integer va1ue, a complex signal processing unit to output the sub-streams as they are, as output signals, perform complex signal processing on the sub-streams , the complex signal processing being time-axis inversion processing, complex conjugate processing and sign inversion processing, or the time-axis inversion processing and complex conjugate processing, and output as output signals the sub-streams on which the complex signal processing has been performed, a phase rotation unit to generate a repetitive signal that is a signal of a third time length in which signal arrangement is repeated in a time direction, for each of the output signals output from the complex signal processing unit, and give a phase rotation to the repetitive signal, and a multiplexing unit to generate, for each of the transmission antennas, a transmission signal transmitted from the transmission antenna, by multiplexing the repetitive signals to which the phase rotation has been given, and the phase rotation unit gives different phase rotations to the repetitive signals multiplexed to an identical transmission signal. Advantageous Effects of Invention [0017] According to the present invention, even in an environment in which a transmission channel temporally varies at high speed, an effect of preventing a decrease in efficiency of transmission power while preventing deterioration of a transmission characteristic is acquired. Brief Description of Drawings [0018] FIG. 1 is a diagram illustrating a configuration example of a transmission device of a first embodiment. FIG. 2 is a diagram illustrating a configuration example of a reception device of the first embodiment. FIG. 3 is a diagram illustrating a configuration example of an SFBC coding unit of the first embodiment. FIG. 4 is a diagram illustrating processing contents of a signal dividing unit of the first embodiment. FIG. 5 is a diagram illustrating processing contents of a complex signal processing unit of the first embodiment FIG. 6 is a diagram illustrating processing contents of a phase rotation unit of the first embodiment. FIG. 7 is a diagram illustrating processing contents of a multiplexing unit of the first embodiment. FIG. 8 is a flowchart illustrating an example of a processing procedure in the SFBC coding unit of the first embodiment. FIG. 9 is a diagram schematically illustrating a frequency spectrum of an output signal of each of the signal dividing unit, the complex signal processing unit, the phase rotation unit, and the multiplexing unit of the first embodiment. FIG. 10 is a diagram illustrating a relationship between frequency signals of transmission signals transmitted from transmission antennas of the first embodiment. FIG. 11 is a diagram schematically illustrating an SFBC coding method for four transmission antennas of the first embodiment. FIG. 12 is a diagram illustrating a configuration example of a transmission device of the first embodiment in a case where SFBC for two transmission antennas and delay processing are combined. FIG. 13 is a diagram illustrating an example of a control circuit of the first embodiment. FIG. 14 is a flowchart illustrating an example of a processing procedure in an SFBC coding unit of a second embodiment. FIG. 15 is a diagram illustrating an example of processing in a signal dividing unit of the second embodiment. FIG. 16 is a diagram illustrating a configuration example of an SFBC coding unit of a third embodiment. FIG. 17 is a diagram illustrating processing of a DFT unit in a transmission device of the third embodiment. FIG. 18 is a diagram illustrating a configuration example of a reception device of a fourth embodiment. FIG. 19 is a diagram schematically illustrating processing of a sub-block combination unit of the fourth embodiment. FIG. 20 is a diagram illustrating a configuration example of a reception device of a fifth embodiment. Description of Embodiments [0019] In the following, embodiments of a transmission device, a reception device, a communication system, a transmission method, and a reception method according to the present invention will be described in detail with reference to the drawings. Note that this invention is not limited to these embodiments. [0020] First Embodiment. FIG. 1 is a diagram illustrating a configuration example of a transmission device of the first embodiment according to the present invention. As illustrated in FIG. 1, a transmission device 1 of the present embodiment includes a mapping unit 11 that performs primary modulation of a transmission bit seguence, an SFBC coding unit 12 that is a coding unit to perform SFBC coding, which is coding using an SFBC technology, of a signal stream that is a result of the primary modulation, CP adding units 13-1 and 13-2 to add a CP (cyclic prefix) to an SFBC-coded transmission signal output from the SFBC coding unit 12, and transmission antennas 14-1 and 14-2. The CP adding unit 13-1 is connected to the transmission antenna 14-1 and adds CP to a transmission signal transmitted from the transmission antenna 14-1. The CP adding unit 13-2 is connected to the transmission antenna 14-2 and adds CP to a transmission signal transmitted from the transmission antenna 14-2. Here, an example in which there are two transmission antennas is described. In a case where there are three or more transmission antennas, a CP adding unit is included for each transmission antenna and each CP adding unit adds CP to a transmission signal transmitted from a corresponding transmission antenna, that is, a connected transmission antenna. The SFBC coding unit 12 performs SFBC coding corresponding to the number of transmission antennas. [0021] In FIG. 1, a component related to baseband signal processing in the transmission device 1 is illustrated. However, the transmission device 1 may include components that are not illustrated in FIG. 1. For example, the transmission device 1 may include a filter, and an analog unit that performs analog signal processing. [0022] Here, a premise and a definition of words in the following description are described. It is assumed that SFBC coding in the present embodiment is Alamouti coding disclosed in each of Non Patent Literature 2 and 3. Also, in the present embodiment, it is assumed that SCBT is used as a transmission system. [0023] Also, in the present embodiment, a unit of adding a cyclic prefix, that is, CP is called a "block" and a time length of one block excluding CP is N. Note that in the present embodiment, time t indicates time discretized in a unit of a primary modulation symbol interval. That is, t is an integer. A time length N of one block is also an integer. A primary modulation result, that is, a modulation symbol sequence transmitted in one block is called a "signal stream" or is simply called a "stream" and a time length of the signal stream is M. In SFBC in a case of two transmission antennas, that is, SFBC for two transmission antennas, N = M. In Alamouti-type SFBC with four transmission antennas, N = M * (4/3) . SFBC coding is performed in a closed manner in one block. In the SFBC coding, a signal stream is equally divided. A signal stream generated by the equal devision is called a "signal sub-stream" or is simply called a "sub-stream." When the number of times of division in a case of equally dividing a signal stream is C, a time length of a sub-stream is Msub ^ M/C. The number of times of division C is C = 2 in SFBC for two transmission antennas and is C = 3 in SFBC for four transmission antennas. [0024] Next, a whole operation in the transmission device 1 will be described. In the transmission device 1, first, the mapping unit 11 performs mapping of a transmission bit sequence into a PSK (phase shift keying) modulation symbol sequence, a QAM (quadrature amplitude modulation) modulation symbol sequence, or the like, that is, performs primary modulation. Note that the transmission bit sequence may be a bit sequence on which pre-processing such as error correction coding has been performed. Then, the SFBC coding unit 12 performs SFBC coding of a modulation symbol sequence, that is, a signal stream and outputs, as a result of the SFBC coding, transmission signals for the number of transmission antennas. A detail of the SFBC coding of the present embodiment will be described later. The CP adding units 13-1 and 13-2 respectively add CPs to the transmission signals output from the SFBC coding unit 12 and respectively output these signals to the transmission antennas 14-1 and 14-2. The transmission antennas 14-1 and 14-2 transmit the CP-added transmission signals output from the CP adding units 13-1 and 13-2. [0025] FIG. 2 is a diagram illustrating a configuration example of a reception device of the present embodiment. As illustrated in FIG. 2, a reception device 2 of the present embodiment includes a reception antenna 21, a CP removing unit 22 to remove CP from a reception signal received in the reception antenna 21, and a DFT unit 2 3 that is a discrete Fourier transform unit to transform the CP-removed reception signal into a frequency domain by DFT. The reception device 2 further includes an SFBC decoding unit 24 that is a decoding unit to perform SFBC decoding of the reception signal after DFT, an FDE unit 25 that is a frequency-domain equalization unit to perform FDE, that is, frequency-domain equalization on an SFBC decoding result, an IDFT unit 26 that is an inverse discrete Fourier transform unit to perform IDFT on the signal on which the frequency-domain equalization has been performed, and a demapping unit 27 that calculates an estimation bit sequence by performing demodulation on the signal on which the IDFT has been perforfmed, in response to primary modulation. [0026] The reception device 2 receives a signal transmitted from the above-described transmission device 1. That is, with the transmission device 1, the reception device 2 configures a communication system. In a case where pre-processing such as error correction coding is performed on a transmission bit sequence in the transmission device 1, the reception device 2 may perform decoding processing that corresponds to the pre-processing on a transmission side, such as deinterleaving or error correction decoding, in a subsequent stage to the demapping unit 27. Also, in a case where soft determination error correction decoding is performed in the subsequent stage, the demapping unit 27 may calculate a soft determination value. [0027] Also, although an example in which there is one reception antenna is illustrated in FIG. 2, there may be a plurality of reception antennas. In this case, the SFBC decoding unit 24 or the FDE unit 25 combines a plurality of reception signals received by the plurality of reception antennas. [0028] In FIG. 2, a component related to baseband signal processing in the reception device 2 is illustrated. However, the reception device 2 may include components that are not illustrated in FIG. 2. For example, the reception device 2 may include a filter, and an analog unit that performs analog signal processing. Also, time synchronous processing, frequency synchronous processing, transmission channel estimation and the like are performed in an actual digital circuit in the reception device 2 but illustration thereof is omitted here. In the following embodiments, it is assumed that the time synchronous processing, the frequency synchronous processing, the transmission channel estimation and the like are operated ideally. [0029] Here, before a detail description of the present embodiment, a relationship between a time signal and a frequency signal thereof is described. In the following, when a time signal v[t] and a frequency signal V[f] are a Fourier transform pair, a relationship between the two is expressed in a manner of an expression fl). Note that f indicates a discretized frequency. v[t] « V[f] ... fl) [0030] A frequency sign a], of a signal v[-t] a time axis of which is inverted with respect to v[t], that is, a Fourier transform result has a property that a frequency axis is inverted with respect to V[f] that is a frequency signal of v[t], as expressed in an expression (2). v[-t] O v[-f] ... (2) [0031] As expressed in the following expression (3), a frequency signal of complex conjugate v* [t] of v[t] has a property that complex conjugate and a frequency axis are inverted with respect to V[f] that is a frequency signal of v[t] . [0032] v*[t] O V*[-f] ... (3) [0033] When a signal to which a phase rotation in proportional to time with respect to a time signal g[t] is given is v[t], a frequency signal of v[t] is that which is generated by performing frequency shift on a frequency signal G[f] of g[t] , as expressed in an expression (4). Af indicates a frequency shift amount. [0034] v[t] - g[t] x exp(-j2nAft) O V[f] - G[f+Af] ... (4) [00 35] As expressed in an expression (5) , in a frequency signal V[f] of the signal v[t] generated by shifting time of g[t] by At, a phase rotation in proportional to a frequency with respect to the frequency signal G[f] of g[t] occurs. [0036] v[t] = g[t+At] «• V[f] - G[f] * exp(j2nfAt) ... (5) [0037] It is known that a frequency signal of a repetitive time signal becomes a comb-shaped spectrum. For example, a signal of a time length N in which signal a signal g[t] (t - 0, 1, ..., (N/2) - 1) of a time length N/2 is repeated twice, that is, reproduced and arranged in a temporally cascade manner is v[t] (t = 0, 1, ..., N - 1). Here, when a frequency signal in which N/2-length DFT is applied with respect to g[t] is G[f] (f =0, 1, ..., (N/2) - 1) and k is an integer, a DFT result of v[t] becomes V[k] = G[f] in a case where k = 2f and becomes V[k] ~- 0 in a case where k = 2f + 1. When this is expressed in an expression, the following expression (6) is acquired with mod as a remainder operator. v[t] - g[t mod (N/2)] (t - 0, ..., N-l) O V[2f] = G[f], V[2f+1] =0 (f = 0, ..., (N/2)-l) (6) [0038] In a case where a property expressed by the above expression (6) is generalized, when L is a positive number, a DFT result of a signal v[t] (t = 0, 1, ..., N-l) of a time length N in which signal a signal g[t] (t = 0, 1, ---, (N/L) - 1) of a time length N/L is repeated L times in a cascade manner becomes a comb-shaped spectrum as expressed in the following expression (7) with k as an integer. v[t] = g[t mod (N/L)] (t = 0, ..., N - 1) « V[k] = G[f] (when k = Lf) , V[k] - 0 (when k * Lf) ... (7) [0039] On the premise of the above, SFBC coding of the present embodiment will be described. FIG. 3 is a diagram illustrating a configuration example of the SFBC coding unit 12 of the first embodiment. The SFBC coding unit 12 includes a signal dividing unit 121 that divides an input signal stream, a complex signal processing unit 122 that performs time-axis inversion processing, complex conjugate processing and sign inversion processing, or time-axis inversion processing and complex conjugate processing on the divided signal or that outputs the divided signal as it is, a phase rotation unit 123 that reproduces and arranges the signal output from the complex signal processing unit 122 or that gives a phase rotation to the signal output from the complex signal processing unit 122, and a multiplexing unit 124 that adds signals processed by the phase rotation unit 123. [0040] FIG. 4 is a diagram illustrating processing contents of the signal dividing unit 121. The signal dividing unit 121 equally divides a signal stream s[t] of a time length M that is input from the mapping unit 11. Here, since a description is made with SFBC for two transmission antennas as an example, the number of times of division in the equal division is C ^ 2. Thus, the signal stream s[t] is equally divided into two that are a first half and a second half, and results of the equal division are sa[t] and sb [t] respectively. Also, sa[t] and sb[t] are respectively referred to as a first sub-stream and a second sub-stream. The signal dividing unit 121 outputs the signal sub-streams sa[t] and sb[t] in parallel. A time length Msub of each of the equally-divided signal sub-streams sa[t] and sb[t] is Maub = M/C = M/2. [0041] FIG. 5 is a diagram illustrating processing contents of the complex signal processing unit 122. In FIG. 5, input signals into the complex signal processing unit 122 are illustrated as signals sl21a and sl21b and output signals of the complex signal processing unit 122 are illustrated as signals sl22a, sl22b, sl22c and sl22d. The signal s!21a is a signal sub-stream sa[t] output from the signal dividing unit 121 and the signal sl21b is a signal sub-stream Sb[t] output from the signal dividing unit 121. The complex signal processing unit 122 respectively outputs, as signals sl22a and sl22c, the signal sl21a, that is, sa[t] and the signal sl21b, that is, sb[t] as they are. Also, the complex signal processing unit 122 performs time-axis inversion, complex conjugate, and sign inversion on the signal sl21b and outputs a result of the processing, that is, -s\[-t] as a signal sl22b. Moreover, the complex signal processing unit 122 performs time-axis inversion and complex conjugate on the signal sl21a and outputs a result of the processing, that is, s'a[-t] as a signal sl22d. That is, the complex signal processing unit 122 generates and outputs -s\[-t] and s*a[-t], and outputs sa[t] and sb[t] as they are. [0042] FIG. 6 is a diagram illustrating processing contents of the phase rotation unit 123. To the phase rotation unit 123, the signals sl22a, sl22b, sl22c and s!22d output from the complex signal processing unit 122 are input. As illustrated in FIG. 6, the phase rotation unit 123 generates signals of a time length N in which the signals sl22a, sl22b, sl22c and sl22d are respectively reproduced and arranged, that is, repetitive signals of a time length N in each of which two sub-streams of the same contents are arranged, and the time length N is the same as a time length N of a signal stream. The phase rotation unit 123 outputs a signal of the time length N in which the signal sl22a is reproduced and arranged, as it is as a signal s!23a that is a first repetitive signal and outputs a signal of the time length N in which the signal sl22c is reproduced and arranged, as it is as a signal sl23c that is a second repetitive signal. Note that outputting an input signal as it is corresponds to giving a phase rotation with a phase rotation amount being 0. Also, the phase rotation unit 123 performs a phase rotation on a third repetitive signal that is a repetitive signal of the time length N in which the signal sl22b is reproduced and arranged, by multiplication by exp (j2nt/N) for the time length N and outputs the phase-rotated signal as a signal s!23b. The phase rotation unit 123 performs a phase rotation on a fourth repetitive signal that is a repetitive signal of the time length N in which the signal sl22d is reproduced and arranged, by multiplication by exp (j2nt/N) for the time length N and outputs the phase-rotated signal as a signal s!2 3d. [0043] FIG. 7 is a diagram illustrating processing contents of the multiplexing unit 124. The multiplexing unit 124 generates a signal sl24a that is a first transmission signal by adding, that is, multiplexing the signals sl23a and sl23b and outputs the signal sl24a to the CP adding units 13-1. Also, the multiplexing unit 124 generates a signal sl24b that is a second transmission signal by adding, that is, multiplexing the signals sl23c and sl23d and outputs the signal s!24b to the CP adding units 13-2. In the following, the signal sl24a is also expressed as xl[t] and the signal sl24b is also expressed as x2[t]. xl [t] corresponds to a transmission signal transmitted from the transmission antenna 14-1 that is a first transmission antenna and x2[t] corresponds to a transmission signal transmitted from the transmission antenna 14-2 that is a second transmission antenna. [0044] The transmission signals xl [t] and x2[t] that are acquired by the series of processing in the signal dividing unit 121, the complex signal processing unit 122, the phase rotation unit 123, and the multiplexing unit 124 can be expressed by next expressions (8) and (9) using the signals sa[t] and sb[t] after the division by the signal dividing unit 121. Note that as described above, in SFBC for two transmission antennas, M = N and Msub = N/2. Note that the time length M is also referred to as a first time length, the time length Msub is also referred to as a second time length, and the time length N is also referred to as a third time length. xittj - sa[t mod Msub] - s*b[(N-t) mod Msub] * exp (j2nt/N) (t = 0, . . ., N-l) ... (8) x2ft] - sb[t mod MSUb) + s\[(N-t) mod Msub] x exp(j2nt/N) (t = 0, . . ., N-l) ... (9) [004 5] As it is understood from the expressions (8) and (9), and FIG. 7, in this manner, the phase rotation unit 123 gives a phase rotation to each signal in such a manner that a rotation amount of the phase rotation varies between signals added in the multiplexing unit 12 4. [0046] FIG. 8 is a flowchart illustrating an example of a processing procedure in the SFBC coding unit 12 of the present embodiment. As illustrated in FIG. 8, the SFBC coding unit 12 equally divides a signal stream s[t] of the time length M (step SI). More specifically, the signal dividing unit 121 equally divides the signal stream s[t] of the time length M that is input from the mapping unit 11. Then, the SFBC coding unit 12 calculates s*a[-t] and -s*b[-t] based on sa[t] and sb[t] (step S2). More specifically, the complex signal processing unit 122 generates sa*[~t] by performing time-axis inversion and complex conjugate on sa[t] and generates -s*b[-t] by performing time-axis inversion, complex conjugate and sign inversion on sb[t]. Then, as described above, the complex signal processing unit 122 outputs sa[t] , sb[t] , s*a [-t], and -s*b[-t] to the phase rotation unit 123. [0047] Then, the phase rotation unit 123 of the SFBC coding unit 12 reproduces and arranges each of sa[t], sb[t], s*a[-t] and -s'b[-t], and generates repetitive signals thereof (step S3}. More specifically, the signals of the time length N in which signals sa[t], sb[t], s a [ -t ] , and -s*b[-tj are respectively reproduced and arranged are generated, that is, the repetitive signals each of which has the two sub-streams of the same contents arranged and the same length N as the time length N of a signal stream are generated. Then, the phase rotation unit 123 of the SFBC coding unit 12 performs phase rotation processing on each of a repetitive signal corresponding to s'a[-t] and a repetitive signal corresponding to ~s*b[-t] (step S4). [0048] Then, the multiplexing unit 124 of the SFBC coding unit 12 adds a repetitive signal corresponding to sa[t] and the repetitive signal corresponding to -s'b[~t], and adds a repetitive signal corresponding to Sb[t] and the repetitive signal corresponding to s*a[-t] (step S5). Signals xi[t] and x2[t] acquired by the addition are respectively output to the CP adding units 13-1 and 13-2. [0049] Here, how the series of processing in the SFBC coding unit 12 is observed as a frequency spectrum will be described. FIG. 9 is a diagram schematically illustrating a frequency spectrum of an output signal of each of the signal dividing unit 121, the complex signal processing unit 122, the phase rotation unit 123, and the multiplexing unit 124. Here, frequency signals that are results of performing Msub-point DFT on the signal sub-streams sa[t] and sb[t] of the time lengths Msub are Sa[f] and Sb[f] respectively and frequency signals that are results of performing N-point DFT on xl[t] and x2[t] that are output signals of the time length N from the multiplexing unit 124 are XI[f] and X2[f] respectively. sl21a or the like illustrated under each frequency spectrum in FIG. 9 indicates a time signal corresponding to the spectrum. [0050] Frequency spectrums of output signals output from the signal dividing unit 121 have the time lengths Msub, and are results of performing Msub-point DFT on the signal sub-streams sa[t] and sb[t], that is, Sa[f] and Sb[f], as illustrated in FIG. 8. [0051] Among frequency spectrums of the output signals output from the complex signal processing unit 122, frequency spectrums corresponding to the signals sl22a and s!22c are Sa[f] and Sta[f] respectively. On the other hand, among frequency spectrums of the output signals output from the complex signal processing unit 122, a frequency spectrum corresponding to the signal sl22b becomes -S*b(f] according to the properties expressed by the above expressions (2) and (3). Similarly, a frequency spectrum corresponding to the signal sl22d becomes S*a[f] according to the properties expressed by the above expressions (2) and (3). [0052] Since an output signal from the complex signal processing unit 122 is repeated twice in a cascade manner in each of frequency spectrums respectively corresponding to the signals sl23a and s!23c among frequency spectrums of the output signals output from the phase rotation unit 123, the frequency spectrum becomes a comb-shaped spectrum according to the properties expressed by the above expressions (6) and (7). Moreover, since repetition is performed twice in a cascade manner and a phase rotation of exp(j2nt/N) is given for the time length N in a time domain with respect to each of the signals sl23b and s123d, a spectrum becomes a comb-shaped spectrum a frequency of which is shifted by one frequency interval in a frequency domain. In the following, a description will be made on the assumption that one interval of a discrete frequency, that is, a frequency interval is one sub-carrier. [0053] A frequency spectrum corresponding to the signal sl24a among output signals output from the multiplexing unit 124, that is, XI[f] is the frequency spectrum of the signal sl23a and the frequency spectrum of the signal sl23b added to each other. A frequency spectrum corresponding to the signal sl24b, that is, X2[f] is the frequency spectrum of the signal sl23c and the frequency spectrum of the signal s!23d added to each other. Since a frequency of the signal sl23b is shifted by one sub-carrier from that of the signal sl23a, the signal sl23b is orthogonal to sl23a and is multiplexed without interference. Similarly, the signals sl23c and sl23d are orthogonal to each other and are multiplexed without interference. [0054] FIG. 10 is a diagram illustrating a relationship between a frequency signal XI[f] of xl[tj transmitted from the transmission antenna 14-1 and a frequency signal X2[f] of x2[t] transmitted from the transmission antenna 14-2. The frequency signals XI[f] and X2[f] of the transmission signals from the transmission antennas 14-1 and 14-2 are in a relationship of being SE'BC-coded for each two successive sub-carriers. That is, expression can be made with the following expressions (10) and (11). Xl[2f] = Sa[f], X2[2f] - Sb[f] ... (10) Xi[2f + 1] = -S*b[f], X2[2f+1] = S*a[fj ... (11) [0055] In such a manner, Sa[f] is mapped, that is, assigned to a position of an even number in Xi[f], that is, a sub-carrier of an even number, -S*b[f] is mapped to a position of an odd number in Xi[f ] , that is, a sub-carrier of an odd number, Sb[f] is mapped to a position of an even number in X2[f] , and Sa[f] is mapped to a position of an odd number in X2[f] . [0056] By the above signal processing in the SFBC coding unit 12, as expressed in FIG. 10 and the expressions (10) and (11), SFBC between adjoining sub-carriers can be realized unlike the conventional SFBC technologies disclosed in Non Patent Literature 4 and Patent Literature 2. Thus, it is possible to suppress deterioration in a transmission characteristic. Moreover, the processing in the SFBC coding unit 12 is performed only by time-domain processing of signal equal division, time-axis inversion, complex conjugate, sign inversion, reproduction, phase rotation and combination, with respect to temporally- successive signal streams. Compared to the conventional technology in Non Patent Literature 3, great PAPR deterioration in a transmission signal waveform transmitted from the transmission antennas 14-1 and 14-2 does not occur Also, since whole SFBC coding processing is performed in a time domain, a device configuration can be simple compared to the conventional technology in Non Patent Literature 3. [0057] Next, decoding processing in the SFBC decoding unit 24 of the reception device 2 in the first embodiment will be described. In the present embodiment, as expressed in the expressions (10) and (11), the transmission device 1 transmits an Alamouti-coded transmission signal. Thus, to the reception device 2, the conventional SFBC decoding method disclosed, for example, in Non Patent Literature 2, 3, or 4 can be applied. Here, for simplification of a description, as illustrated in FIG. 2, an example in which there is one reception antenna 21 will be described. Note that as described above, to the SFBC decoding unit 24, a frequency-domain signal that is a CP-removed reception signal transformed into a frequency domain by the DFT unit 23 is input. Also, transmission channel estimation is performed by a transmission channel estimation unit that is not illustrated in FIG. 2 and a result of the estimation of a transmission channel is input into the SFBC decoding unit 24 . [0058] When a complex transmission function of a transmission channel between the transmission antenna 14-1 and the reception antenna 21 is H][f] and a complex transmission function of a transmission channel between the transmission antenna 14-2 and the reception antenna 21 is Ho [f] , frequency signals R[2f] and R[2f + 1] of reception signals at frequencies 2f and 2f + 1 are expressed by next expressions (12} and (13). Note that N[f] is a noise component an average value of which is 0 and dispersion of which is a2 at a frequency f. R[2f] = Hl[2f]Xl[2f] + H2[2f]X2[2f] + N[2f] = Hl[2f]Sa[f] + H2[2f]Sb[f] + N[2f] ... (12) R[2f+1] - Hx[2f+l]Xi[2f+l] + H2[2f+l]X2[2f+l] + N[2f+1] - Hi[2f + 1] (-Ssb[f] ) + H2[2f + l]S*a[f] + N[2f + 1] ... (13) [0059] SFBC decoding with respect to a transmitted signal sub-stream is processing in a frequency domain expressed in next expressions (14) and (15). By the following expressions (14) and (15), decoding signal sub-streams S'a [f] and S'b[f] that are decoding results can be acquired. In the following, where appropriate, SFBC decoding expressed in the expression (14) is called S'a[f] decoding and SFBC decoding expressed in the expression (15) is called S'b[f] decoding. S'a[f] - H\[2fjR[2f] + H2[2f+l]R*[2f+l] = (|Hl[2f]|2 + |H2[2f + l] |2)xSa[f] + (H\[2f]H2[2f] - H*a[2f + l]H2[2f + l])xSb[f] + Na[f] ... (14) S'b[f] = H*2[2f]R[2f] - Hi[2f + l]R*[2f + l] = (|H2[2f]|2 + |Hl[2f+l]|2) * Sb[f] + (H1[2f]Hi2[2f ] - Hi [2f+ 1] H% [2f+l] ) x Sd[f] + Nb[f] ... (15) [0060] Here, in the above expression (14), Na[f] is an equivalent noise component in the S'a[f] decoding, an average value thereof being 0 and dispersion thereof being (|H:j[2f]|2 + |H2[2f + 13j;')a2. Also, in the above expression (15), Nb[f] is an equivalent noise component in the S'b[f] decoding, an average value thereof being 0 and a dispersion value thereof being (|H2[2f]|2 + |HT[2f + l]\2)a2 [0061] As expressed in the expressions (14) and (15), the SFBC decoding unit 24 calculates S'a[f] and S'b[f] on the basis of the complex transmission functions Hl[f] and H2[f] of the transmission channels that are input as results of estimation of transmission channels, and a frequency signal that is a reception signal transformed into a frequency domain. [0062] As it is understood from the expression (14) , a coefficient of an intended Sa[f] component in the S'a[f] decoding is a maximum ratio combination of transmission channel gain of the two transmission antennas 14-1 and 14-2. and transmission diversity is acquired. Also, a coefficient of an Sb[fj component becomes substantially 0 in a case where frequency selectivity of a transmission channel can be ignored, and an interference component is removed from a decoding result. A coefficient of an intended Sb[f] component also becomes a maximum ratio combination in the expression (15) and a coefficient of an Sa[f] component that causes interference becomes substantially 0 in a case where frequency selectivity of a transmission channel can be ignored. Thus, the SFBC decoding can be realized by the expression (14) and the expression (15). As described above, this SFBC decoding is the technology disclosed, for example, in Non Patent Literature 2, 3, or 4. [0063] Next, processing in the FDE unit 2 5 of the present embodiment will be described. As described above, S'a[f] and S'b[f] for the number of sub-carriers Msub, that is, for f = 0, 1, ..., and Msub - 1 are acquired by the SFBC decoding unit 24 in a preceding stage of the FDE unit 25. Here, FDE of S'a[f] will be described. As described later, FDE of S'b[f] is similar. [0064] According to the expression (14), the coefficient of the intended Sd[f] component included in S'a[f], that is, an equivalent transmission channel value after the SFBC decoding is |H.i[2f] \'~ + jH?[2f + 1] \'~ . Here, an equivalent transmission channel value Ga[f] is a value defined in the following expression (16}. Ga[f] = |HI[2 f] | 2 + |H2[2f+l]|2 ... (16) [0065] As described above, since dispersion of the equivalent noise component included in S'a[f] is Ga[f]o2, dispersion of a noise component varies depending on a frequency f. Thus, in order to make an influence of an equivalent noise component constant regardless of a frequency, as expressed in the following expression (17), normalization is performed by division of the SFBC-decoded signal S'a[f] by V(Ga[f]) and a result of the division is set as Ya[f]. Accordingly, dispersion of an equivalent noise component included in Ya[f] becomes cT and a square root of an equivalent transmission channel value becomes V(Ga[f] ) . Ya[f] = S'a[f] / V(Ga[f]) -.- (17) [0066] The FDE unit 25 performs the processing expressed in the above expression (17) with respect to all sub-carriers, that is, all kinds of f in f = 0, 1, ..., Msub -1 and applies FDE to Ya[f] of f = 0, 1, ..., Msub - 1. As FDE, for example, FDE based on a minimum mean square error (MMSE), that is, MMSE-FDE can be applied. MMSE-FDE weight Wa[f] with respect to Ya[f] at a frequency f is calculated by the following expression (18). Wa[f] - V(Ga[f] ) / (Ga[f] + a2) . . . (18) [0067] The FDE unit 25 calculates the MMSE-FDE weight by the expression (18) with respect to all sub-carriers and multiplies Ya[f] thereby. Subsequently, the IDFT unit 26 performs Msub-point IDFT with respect to Ya[f] multiplied by the MMSE-FDE weight and acquires ya[t] that is an estimation signal sub-stream for sa[t]. [0068] With respect to S'ta[f], a coefficient of an intended Sb[f] component is similarly set as Gb[f] as expressed in expressions (19) and (20). Also, S'bff] that is divided by V(Gb[f]) and normalized is set as Y-D[f]. Accordingly/ dispersion of an equivalent noise component included in Yb[f] becomes a" and a square root of an equivalent transmission channel value becomes V(Gb[f]) . Gb[f] = |H2[2f] |2 + |H1[2f+l] |2 . . . (19) Yb[f] - S'b[f] / V{Gb[f]) ... (20) [0069] Similarly to the expression (18), MMSE-FDE weight Wb[fj is calculated with respect to all sub-carriers and Yb[f] is multiplied thereby. Subsequently, the IDFT unit 26 performs Msub-point IDFT on Yb[f] multiplied by the MMSE-FDE weight and acquires yb[t] that is an estimation signal sub-stream for sb[t]. [0070] In the above, an SFBC coding method and decoding method, and an FDE method have been described with respect to SFBC for two transmission antennas. However, the present invention can be applied not only to SFBC for two transmission antennas but also to arbitrary SFBC that realizes coding by complex conjugate and sign inversion of a signal. As an example other than the case of SFBC for two transmission antennas, an SFBC coding method and decoding method in SFBC for four transmission antennas will be briefly described focusing on a difference from the case for two transmission antennas. Similarly to the transmission device 1 in FIG. 1, a transmission device corresponding to SFBC for four transmission antennas includes a mapping unit 11 and an SFBC coding unit 12. However, the SFBC coding unit 12 performs SFBC coding for four transmission antennas, as is discussed below. Also, this transmission device includes four transmission antennas and four CP adding units. [0071] FIG. 11 is a diagram schematically illustrating an SFBC coding method for four transmission antennas. In SFBC using four transmission antennas, a signal stream of a time length M is equally divided into three. That is, the number of times of division C = 3. A time length of signal sub-streams after the equal division is set as Msub. The signal sub-streams of the time lengths Msub are set as sa[t], sb[t], and sc[t]. In SFBC using four transmission antennas, N ^ 4Msub. The four transmission antennas are set as transmission antennas #1, #2, #3, and #4 and signals transmitted from the transmission antennas #1, #2, #3, and #4 are respectively set as xl[t], x2[t], x3[t], and x4[t]. xl[t], x2[t], x3[t], and x4[t] can be expressed by the following expressions (21) to (24) using sa[t], sb[t], and sc[t] . xl[t] = sa[t mod Msub] - s*b[(N-t) mod Msubj x exp(j2nt/N) - s*c[