JP2003218826A - Method for receiving orthogonal frequency division multiplexed signal and receiver using the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for receiving an orthogonal frequency division multiplexing (OFDM) signal and a receiver having sufficient followability to fluctuations in a propagation path, capable of suppressing both intersymbol interference prior to the present time and intersymbol interference between the carriers with substantially the same receiver scale and transmission format as in a conventional system. <P>SOLUTION: The reception system includes Fourier transformation processing for Fourier-transforming by multiplying the received OFDM signal by a window function, first adaptive equalization processing for eliminating an intersymbol interference component prior to the present time from the signal posterior to the Fourier transformation, second adaptive equalization processing for eliminating an intersymbol interference between the carriers of the present time, and channel estimation processing for estimating a channel impulse for use in the first and the second adaptive equalization processing. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to wireless communication using an orthogonal frequency division multiplexing (hereinafter referred to as "OFDM") system such as wireless LAN, mobile communication, terrestrial digital TV broadcasting, etc. The present invention relates to a reception system and a receiver that improve the interference caused by codes before the present time and the deterioration of transmission characteristics caused by intersymbol interference between carriers at the present time.

[0002]

2. Description of the Related Art In a transmission system using OFDM, the influence of multipath delay spread in high-speed digital signal transmission can be reduced by using multi-carrier and inserting a guard interval.
It has attracted attention as a transmission method used for broadcasting and mobile communication methods in the future.

This transmission system using OFDM is very effective in a radio propagation path in which a delay time spread shorter than the guard interval length occurs, and is adopted in, for example, a wireless LAN system and a terrestrial digital TV system.
On the receiving side, methods that use the synchronous detection method and the differential detection method are summarized as specifications.

1. IEEE Std 802.11a, “High-speed Phy
sical Layer in the 5 GHz Band ”, 1999. 2. ARIB STD-B31,“ Transmission system for terrestrial digital television broadcasting ”, Japan Radio Industries Association, May 2001 The OFDM signals used in these methods are multi-carrier type. It has the common property of inserting guard intervals.To clarify the structure of the OFDM signal,
First, the structure and operation of the modulator will be briefly described.

FIG. 9 shows the basic structure of an OFDM transmitter.
The data signal sequence is serial-parallel converted and then modulated for each carrier, and the inverse fast Fourier transform (IFF) is performed.
It is converted by T) into a modulated signal in the time domain.
The guard interval inserter inserts the latter half of the modulated signal as a guard interval to generate a transmission signal. FIG. 10 shows the basic structure of an OFDM signal, which is composed of a guard interval and a modulated signal.

The basic structure of a receiver for processing the OFDM signal thus transmitted will be described with reference to FIG. First, the guard interval remover 1 removes the guard interval part of the received signal to extract the signal in the FFT section of the OFDM signal shown in FIG. After that, the Fourier transformer 2 performs Fourier transform in the FFT section to transform the received signal from the time domain to the frequency domain. Next, the amplitude and phase components distorted by the propagation path are obtained by the channel estimator 3 using a pilot signal that is a known signal in the transmitting and receiving organization, and the signal is obtained by the synchronous detector 4 using the obtained amplitude and phase components. Demodulate. Further, in order to simplify the configuration of the receiver, a configuration in which a delay detector is used instead of the synchronous detector 4 is also considered. In this case, the channel estimator 2 is unnecessary, but the transmission characteristic is deteriorated as compared with the synchronous detection method.

These receivers are also considered to be used outdoors, and in this case, since the propagation space is widened, it is probabilistic that the delay spread due to the multiple wave propagation environment peculiar to mobile radio communication is equal to or longer than the guard interval. It is thought to occur in. At this time, the following intersymbol interference appears in the received signal.

Inter-code interference due to the code before the present time being superimposed on the current code with a delay, which includes intra-carrier interference and inter-carrier interference.

Intersymbol interference caused by superimposing the code interference of another carrier on the code of one carrier at the present time, which is only intercarrier interference.

FIG. 12 shows these intersymbol interferences. It is known that when these intersymbol interferences occur, the propagation characteristics are significantly deteriorated. In an environment in which the delay spread is longer than the guard interval, a method of lengthening the guard interval can be considered, but in that case, the transmission efficiency is significantly deteriorated. The following Documents 1 and 2 show methods capable of suppressing the above three types of intersymbol interference.

Reference 1. Toru Nishikawa, Yoshitaka Hara, Shinsuke Hara, "A Study on Adaptive Array for OFDM in Mobile Communications",
IEICE Research Technical Report, RCS2000-232, 2001 3
Monthly literature 2. Yoichi Maeda, Masayuki Mogi, Ryuji Kono, "OFDM
Study on Optimization of Reflection Coefficient of Multidimensional Lattice Filter for Automotive ", IEICE Technical Report, RCS2000-25
0, March 2001 In the method described in the above-mentioned document 1, a receiver is provided with a plurality of antennas, and the received signals of the respective antennas are weighted and combined so that a delayed wave longer than the guard interval or an incoming wave with a large Doppler shift is obtained. By suppressing the above, the intersymbol interference before the present time and the inter-carrier interference intersymbol interference at the present time are suppressed. However, this method has a drawback in that the size of the receiver becomes large because it is necessary to install a plurality of antennas in the receiver.

In the method described in the above-mentioned document 2, a received signal including intersymbol interference before the present time point and intersymbol interference between carriers at the present time point is returned to a signal without interference by passing through a multidimensional lattice filter. Is the way. In this method, the inverse characteristics of the generation process of the three types of intersymbol interference are realized by the multidimensional lattice filter. However, this method has drawbacks in that convergence is slow in obtaining the tap coefficient of the lattice filter, and followability to fluctuations in the propagation path is not good.

In addition, the paper “Kirsten Matheus, Klaus Kno
che, Martin Feuersanger Karl-Drik Kammeyer, ”Two-
Demensional (recursive) Channel Equalization for Mu
lticarrier Systems with Soft Impulse Shaping (MCSI
S) ”, Proc. IEEE Globecom Conference, pp.956-961, N
ovember, 1998 ”discloses adaptive equalization processing in the time / frequency domain in a multi-carrier transmission system. The technique described in this paper relates to the multi-carrier transmission system and is different from the OFDM system. The points different from the OFDM system are that the spectrum of each carrier is shaped in the transmitter, that the guard interval is not inserted on the transmitter side, and that each carrier is not orthogonalized as in the OFDM system. It is a point.

[0014]

When the conventional receiver is used outdoors, the propagation space is widened. Therefore, the delay spread due to the multipath propagation environment peculiar to mobile radio communication is stochastically longer than the guard interval. Thought to occur,
At this time, three types of intersymbol interference occur. The inter-symbol interference occurring as shown in FIG. 12 includes (1) inter-symbol interference (intra-carrier interference and inter-carrier interference) due to delay and superimposition of a code before the current time, and (2)
There is inter-code interference (only inter-carrier interference) due to the code interference of another carrier being superimposed on the code of one carrier at the present time.

Furthermore, these interferences are time-varying in wireless transmission. In the environment where the intersymbol interference occurs as described above, the conventional synchronous detection method and differential detection method have a problem that the transmission characteristics are significantly deteriorated.

The present invention has been made under the circumstances as described above, and the object of the present invention can be realized in a receiving system of an OFDM signal with a receiver scale and a transmission format similar to the conventional one, and it is possible to prevent fluctuations in the propagation path. An object of the present invention is to provide a receiving system and a receiver which have good followability and can suppress intersymbol interference before the present time and intersymbol interference between carriers at the present time.

[0017]

According to the present invention, a first adaptive equalization process for removing an intersymbol interference component before the present time from a signal after Fourier transform by multiplying an OFDM received signal by a window function By performing both the second adaptive equalization processing for removing the inter-carrier interference between the current carriers, all the inter-symbol interference is removed. Decision feedback adaptive equalization is used as the first adaptive equalization processing, and maximum likelihood sequence estimation adaptive equalization or delayed decision feedback sequence estimation adaptive equalization is used as the second adaptive equalization processing. For the channel impulses used in the two adaptive equalization processes, a replica of an interference signal of either or both of the intersymbol interference before the present time point and the intersymbol interference between carriers at the present time point is generated and estimated by the recursive least squares method.

By doing so, it becomes possible to use the adaptive algorithm with good convergence to follow the high-speed fluctuation of the propagation path, and use the maximum likelihood sequence estimation adaptive equalizer with the best performance. You can Further, there is an advantage that the size of the receiver can be made almost the same as the conventional one and the signal transmission efficiency is not lowered.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings.

Embodiment 1 FIGS. 1 and 2 show an example of an embodiment for carrying out the present invention, and the detailed configuration of the channel estimator in FIG. 1 is shown in FIG. The features of the present invention are, as shown in FIG. 1, a Fourier transformer for performing a Fourier transform by multiplying a received signal subjected to orthogonal frequency division multiplexing by a window function, and an intersymbol interference before the present time from the signal after the Fourier transform. It is composed of an adaptive equalizer for removing components, an adaptive equalizer for removing inter-carrier interference components at the present time, and a channel estimator for estimating channel impulses used in the above two adaptive equalization processes.

When the delay spread due to the multipath propagation environment peculiar to mobile radio communication is stochastically longer than the guard interval, intersymbol interference before the present time point and intersymbol interference between carriers at the present time point occur. Here, “current time” refers to an arbitrary OFDM symbol period, and “before current time” refers to a symbol period before that. A signal multiplexed by the same inverse Fourier transform in the transmitter is called an "OFDM symbol". The receiver receives an arbitrary OFDM symbol, multiplies the received signal by a window function, and performs Fourier transform. Here, the "window function" is an arbitrary time function including a rectangular window.

Next, there is intersymbol interference from before the present point in time, actually there is intercarrier interference and intracarrier interference as shown in FIG. 9, but these are collectively removed by the decision feedback adaptive equalizer. In the decision feedback adaptive equalizer, if the received signal is a signal or pilot signal known in the transceiver, it is used, and at the time of the other data section, the already determined result is fed back and used. Generate replicas of intersymbol interference from before this time. This process is performed by the ISI replica generator of FIG. In the ISI replica generator, the signal of each carrier from before the present time (pilot signal or signal after determination) and the amplitude when the signal is received,
The estimated value of the phase is used to generate a replica of intersymbol interference on all carriers. By subtracting the replica from the Fourier-transformed received signal, it is possible to obtain a received signal in which inter-symbol interference from before the present time is removed in all carriers. The estimated values of amplitude and phase used in the ISI replica generator are estimated by the channel estimator.

Further, the inter-carrier interference between the current carriers is removed by the maximum likelihood sequence estimation type adaptive equalizer to determine the data signal sequence. As shown in FIG. 1, using the transmission signal candidate generated by the maximum likelihood sequence estimator and the estimated values of the amplitude and the phase when the signal is received, the ICI replica generator uses the present inter-carrier code. A replica of the received signal including inter-interference is generated, and a sequence in which the absolute value squared error between the generated replica and the received signal from which the inter-symbol interference before the present time is removed is minimized is output as the transmission signal sequence. Since the maximum likelihood sequence estimation type adaptive equalizer cannot perform batch processing on all carriers, processing is performed for each carrier.
Inter-carrier inter-code interference leaking into a certain carrier can be limited to the extent of adjacent carriers by using a window function during Fourier transform. Therefore, a replica of a received signal of a certain carrier can be generated with signal candidates of several carriers including the adjacent carrier and the own carrier. In fact, since the processing is performed sequentially, the maximum likelihood sequence estimator with a small number of states can be used for sequence estimation, and the delay decision feedback sequence estimation adaptive equalizer with a reduced amount of calculation can also perform processing. is there. In addition, ISI
Similar to the replica generator, the amplitude and phase estimation values used in the ICI replica generator are estimated by the channel estimator.

As shown in FIG. 2, in the channel estimator for estimating the amplitude and phase of the received signal used in the above two adaptive equalizers, as shown in FIG. Interference, (2) inter-carrier intersymbol interference at the present time generated by a delayed wave of GI or more and a window function, (3) a replica of a received signal of its own carrier is generated in each carrier, and an absolute value 2 with an actual received signal is generated. The channel impulse is estimated by controlling the recursive least squares method so that the power error is minimized. By updating the process in each carrier, the process converges in about several OFDM symbols. Channel estimation is performed using the known pilot signal when the received signal is in the pilot signal section, and the determined result when it is in the data symbol section.

The above basic operation will be described in detail using mathematical expressions. Symbol i, time in OFDM system
The transmission signal s _i (t) is transmitted from the transmitter in iT _s ≦ t ≦ (i + 1) T _s .

[0026]

[Equation 1] However, z _{i, n} is a data symbol modulated in each carrier in the OFDM system, N is the number of carriers, Δ _f is a carrier interval, and T _G is a guard interval.

The received signal r (t) observed at the receiver through the multiple wave propagation path is given by the following equation 2 where N _i is the number of transmission symbols and h (t) is the impulse response of the propagation path. Note that n (t) is thermal noise.

[0028]

[Equation 2] Here, h (t) is a propagation path that can be discretized with a sampling interval _Δt.

[Equation 3] Becomes Note that h _d is a complex amplitude and DΔ _i is the maximum delay time. Here, as a delay wave exceeding a guard interval _{T G} (= GΔ _t) is present, it is set to D> G.

Further, the Fourier transform section iT _s + T of the symbol i
_The received signal r _i (t) observed in _G ≤ t ≤ (i + 1) T _s is

[Equation 4] It can be expressed as. In addition, the second right side of the above equation
The term is a component that causes inter-carrier interference of the symbol i, and the third term on the right side is an inter-symbol interference component that leaks from the previous symbol (i-1).

At the receiver, immediately after the guard interval, iT
Fourier transform is performed at N points obtained by discretizing the section from _s + T _G to (i + 1) T _s at the sampling interval Δ _t . At that time, there is an advantage that the spread of the inter-carrier interference and the inter-symbol interference component in the frequency domain can be suppressed by performing the Fourier transform after multiplying the received signal by the window function.

Further, the received signal R _i (m) after the Fourier transform observed with the carrier m is given by the following expression 5.

[0032]

[Equation 5] It should be noted that W _n (m), W _{0, d, n} (m), W _{-1, d, n} (m), and N (m) are the following Equation 6, Equation 7, Equation 8, and Equation 9, respectively. Where p _w (kΔ _t )
Is the discretized window function.

[0033]

[Equation 6]

[Equation 7]

[Equation 8]

[Equation 9] Here, as the processing performed in the receiver, the complex amplitude h _d of each path is estimated by the channel estimator using the received signal after the Fourier transform. Next, a replica of the intersymbol interference component, which is the third term on the right side of the above equation 5, is generated by the decision feedback adaptive equalizer using the post-decision signal, and subtracted from the received signal after the Fourier transform. Finally, replicas of the first term and the second term on the right-hand side of Equation 5 are generated by the maximum likelihood sequence estimation adaptive equalizer, subtracted from the received signal after the processing of the decision feedback adaptive equalizer, and their absolute values are obtained. The sequence that minimizes the squared value of is output as the transmission signal sequence.

Embodiment 2 The operation and algorithm of the channel estimator will be described. The channel estimator estimates the complex amplitude h _d of each path using the received signal after the Fourier transform. At that time, a pilot signal or a determined signal which is a known signal between the transmitter and the receiver is used. Now, it is assumed that the received signal is a pilot signal for channel estimation and the transmitted signals of all carriers are known.

FIG. 2 shows a configuration example of the channel estimator. The channel estimator generates a replica of the received signal on the carrier m, and operates according to the standard that minimizes the square of the absolute value of the error from the actual received signal. Channel estimator input vector
Let x (m) be the following Expression 10.

[0036]

[Equation 10] However, for 1 <G

[Equation 11] And for 1 ≧ G

[Equation 12] Is. Then, the number of elements L + 1 vectors to allow channel estimation to the maximum delay time d? _T should be D + 1 or more.

The complex amplitude of each path to be estimated Then the vector w to be estimated is

[Equation 13] And the replica of the received signal at carrier m Is

[Equation 14] It can be expressed as. Here, if the error from the received signal R _i (m) is e (m), the error e (m) is

[Equation 15] Becomes Reference “S. Haykin, Adaptive filter Theory, 3
rd ed., Prentice-Hall, 1993 ", describes the RLS algorithm in detail.

Here, when the optimum tap vector w is sequentially calculated using the exponentially weighted RLS algorithm,
Inverse matrix P (m) of autocorrelation matrix and Kalman gain vector k
It is updated by the following Expressions 16 to 18 using (m).

[0039]

[Equation 16]

[Equation 17]

[Equation 18] As described above, in the RLS algorithm, the tap vector at the carrier m is multiplied by the complex conjugate of the pre-estimation error and the time-varying Kalman gain vector k (m), and the vector is added to the tap vector one symbol before. Estimated by The initial value P (0) of the inverse matrix of the autocorrelation matrix and the initial value W (0) of the estimated tap vector are as shown in the following Expressions 19 and 20, respectively. However, I is an identity matrix and δ is a small positive real number.

[0040]

[Formula 19]

[Equation 20] The RLS algorithm is updated in each carrier in the frequency domain, up to the last carrier. Then, in the next symbol, the tap vector estimated in the previous symbol
Update w and the inverse matrix P (m) of the autocorrelation matrix as initial values. At that time, in the processing of the first carrier, the forgetting factor λ is set to 1 or less so as to follow the fast fluctuation of the propagation path. For other carriers, the forgetting factor λ is set to 1 and the estimation results are averaged.

For each symbol, the estimated complex amplitude of each path , The inter-symbol interference before the present time and the inter-symbol interference between the current carriers are suppressed by the adaptive equalizer. By using the RLS algorithm in the frequency domain, it is possible to estimate the tap vector with a few symbols, and thus the convergence is excellent. Also, by adaptively changing the forgetting factor λ,
There is an advantage that followability to high-speed fluctuation is also improved.

Embodiment 3: A decision feedback type adaptive equalizer will be described. As shown in FIG. 1, the decision feedback adaptive equalizer feeds back the decision result, multiplies it by a tap coefficient to generate a replica of intersymbol interference, and subtracts it to suppress intersymbol interference. A decision feedback adaptive equalizer is described in, for example, the document “John G. Proakis, Digital Communication, 3rd ed.,
McGraw-Hill, 1995 ”describes its operation in detail.

Here, since the intersymbol interference component in the received signal can be represented by the third term on the right side of the equation 5, the judgment result in the symbol i-1 or the pilot signal z _{i-
1, n} and the complex amplitude of each path estimated by the channel estimator , The replica of the intersymbol interference component Is

[Equation 21] Becomes By subtracting this from the received signal R _i (m) of each carrier, the intersymbol interference component before the present time can be removed.

Fourth Embodiment A maximum likelihood sequence estimation type adaptive equalizer will be described. The maximum likelihood sequence estimation adaptive equalizer is similar to the decision feedback adaptive equalizer in the document “John G. Proakis, Digital.
Communication, 3rd ed., McGraw-Hill, 1995 ”describes its operation in detail. As shown in FIG. 1, the maximum likelihood sequence estimation adaptive equalizer is a transmission generated by the maximum likelihood sequence estimator. A replica of the received signal is generated using the signal candidates, and the sequence having the smallest absolute value squared error from the actual received signal is output as the transmitted signal sequence.

In this case, the orthogonality between carriers is broken due to inter-code interference between carriers, and the spread extends to all carriers, so that the number of transmission signal candidates generated by the maximum likelihood sequence estimator becomes enormous. Will end up. However, by using a window function during the Fourier transform, it is possible to suppress the spread of inter-carrier intersymbol interference to the extent of adjacent carriers, which can significantly reduce the computational complexity of the maximum likelihood sequence estimation adaptive equalizer. There are advantages. Thereby, the power consumption of the receiver can be suppressed.

Considering the case where only the influence of the carrier m observed by the window function from the adjacent carriers m-1 and m + 1 needs to be considered, the first term and the second term on the right side of the above equation 5 are: It can be approximated by the following formula 22.

[0047]

[Equation 22] Therefore, a replica of the component obtained by subtracting the intersymbol interference component from the received signal R _i (m) of each carrier Is the transmission signal candidate And channel estimate Using and, the following equation 23 is obtained.

[0048]

[Equation 23] The observed replica of carrier m is carrier m.
Since it can be approximately expressed only by the transmission signal candidates of -1, m, m + 1, the transmission signal can be determined by the maximum likelihood sequence estimator with 16 states. The maximum likelihood sequence estimation is performed in each carrier by the maximum likelihood sequence estimation type adaptive equalizer in the frequency domain to determine the transmission signals of all carriers.

Fifth Embodiment: In the description of the basic operation of the first embodiment, it is assumed that the Fourier transform for the symbol i is performed in the section from iT _s + T _G to (i + 1) T _S immediately after the guard interval. Since the present invention can eliminate intersymbol interference,
The following two methods are possible. The sections of the Fourier transform of the following two methods are shown in FIG.

Method 1: It is also possible to perform Fourier transform in the section from iT _s to (i + 1) T _s and positively use the signal power in the guard interval.

Method 2: From the same point of view, Fourier transform iT
_s+ T_GTo (i + 1) T_sAnd iT_sTo (i + 1) T_s-T _GIn two sections of
Between both Fourier transform results before and after the current time
Remove the interference component and use the maximum likelihood sequence estimation type adaptive equalizer to
Minimize the sum of both absolute value squared errors when calculating
To control.

As described above, since the signal energy of the guard interval is also taken in and used, the transmission error rate characteristic is improved.

Sixth Embodiment: A case where a linear process for orthogonalization is used as the second adaptive equalization process will be described with reference to the configuration example of FIG. A signal from which intersymbol interference from before the present time is removed by the first adaptive equalization processing is defined as R _{i, ICI} (m), and further defined as a vector y as shown in the following Expression 24.

[0054]

[Equation 24] When the component from which the intersymbol interference from before the present time is removed is expressed by a matrix in the equation 5, a vector z having the signal z _{i, n} modulated in each carrier as an element

[Equation 25] Using, the following equations 26 and 27 are obtained.

[0055]

[Equation 26]

[Equation 27] However, the p rows and q columns of the matrices W _d , W _{o, d} are expressed by the following formula 28 and formula 2 using W _n (m) of formula 6 and W _{o, d, n} (m) of formula 7, respectively.
It becomes like 9.

[0056]

[Equation 28]

[Equation 29] Further, n is a vector having as an element the noise N (m) observed in each carrier.

Here, the method of obtaining the transmission modulation vector z from the vector y in the receiver can be realized by multiplying the vector y by the inverse matrix of A when noise can be ignored. However, since the inverse matrix does not always exist,
General inverse matrix as shown in the following Equation 30 To use.

[0058]

[Equation 30] This general matrix Performs the operation of orthogonalizing the inter-carrier interference between the current carriers and returning to the original modulation vector z.

Further, since noise actually exists, the matrix for orthogonalization is Is 31 below.

[0060]

[Equation 31] However, I is a unit matrix, and σ ² is a minute raw real number.
Generalized inverse matrix Output vector when is multiplied by the vector y Becomes the following Expression 32, and the transmission signal is determined by detecting this.

[0061]

[Equation 32] This process is called a second adaptive equalization process. Reference "S. Haykin,
Adaptive filter Theory, 3 ^rd ed., Prentice-Hall, 1
In 993 ", orthogonalization using a generalized inverse matrix is described in detail.

Embodiment 7: A case where the first adaptive equalization processing is used for a received signal in the time domain will be described with reference to FIG.

Since the received signal is as shown in the equation (4), the intersymbol interference component before the present time is the third term on the right side. By generating a replica of the received signal in the time domain using the determination result or the pilot signal z _{i-1, n} in the symbol (iI) before the present time, and subtracting from the received signal, the intersymbol interference before the present time is obtained. Suppress. Replicas of intersymbol interference components from the time before in the time domain Is the following Expression 33 and Expression 34.

[0064]

[Expression 33]

[Equation 34] In addition, Is the complex amplitude of each path estimated by the channel estimator.
A replica of the intersymbol interference component from the received signal r _i (t) of the symbol i before the present time Signal with subtracted Becomes the following Expression 35, which is input to the Fourier transformer and the second adaptive equalization processing is performed on the output.

[0065]

[Equation 35] Eighth Embodiment: A case in which a channel estimation process is combined with a process in the second adaptive equalization process will be described with reference to FIG. When maximum likelihood sequence estimation type adaptive equalization or delay decision feedback sequence estimation type adaptive equalization is used as the second adaptive equalization processing, the maximum likelihood sequence estimation is performed when the transmission signal sequence is estimated by the maximum likelihood sequence estimator. A channel estimation process is sequentially performed for each transmission signal candidate generated by the device. By doing so, the ability to follow changes in the propagation path is improved.

[Embodiment 9] A case will be described in which channel estimation is performed using the determined signal or the signal that has been error-correction decoded and then re-encoded in the channel estimation processing. As described in the second embodiment, when a pilot signal, which is a known signal between transmission and reception, is inserted in all carriers, channel estimation is performed by using it. In the data section, the channel estimation value is updated using the determined signal. However, when the pilot signals are inserted only discretely as in the terrestrial digital TV broadcasting, that is, the scattered pilot signals lack the pilot signals necessary for channel estimation. This is because the arrangement of the scattered pilot signals is designed by considering only the multipath delay spread within the guard interval.

Therefore, in carriers other than pilots, channel estimation is performed by using a signal determined by synchronous detection or differential detection, or by error-correction decoding the determined signal and then using the re-encoded signal as a substitute for the pilot signal. To do.

The operation of the receiver is summarized as above. (1) Perform signal determination by synchronous detection or differential detection on carriers other than pilot signals, (2) Use the result of determination as a scattered pilot signal (the signal that has been error-correction decoded and then re-encoded) Channel estimation is performed, (3) two types of adaptive equalization processing are operated using the channel estimation result, and (4) both synchronous detection and differential detection and adaptive equalization processing are operated in several symbols, After improving the estimation accuracy, synchronous detection and differential detection are stopped and only the present invention is operated.

By doing so, in the scattered pilot signal in which the pilot signal is inserted only discretely, it is possible to estimate the maximum delay time or more of the channel estimation using only the pilot signal. it can. Further, even when the pilot signal is inserted in all carriers, the signal after error correction can be used for channel estimation and first adaptive equalization processing, and interference suppression capability can be improved.

Embodiment 10: In the channel estimation processing,
A case where the channel estimation process is performed again when the pilot signals are discretely arranged will be described. When a pilot signal is inserted in all carriers in the pilot section and almost no pilot signal is inserted in the data section like a wireless LAN, the channel estimation value estimated in the pilot section is updated in the data section. On the other hand, when a pilot signal is discretely inserted in all symbols as in terrestrial digital TV broadcasting, the channel estimation value is initialized periodically or irregularly to perform channel estimation again. By doing so, it is possible to avoid an increase in estimation error due to a reduction in received power and a deterioration in instantaneous channel estimation due to failure to follow a high-speed channel fluctuation.

Embodiment 11: A case where two processes are repeatedly performed in the adaptive equalization process and the channel estimation process will be described. The accuracy of channel estimation is proportional to the intersymbol interference suppression capability of the present invention. Therefore, by repeatedly performing the adaptive equalization process and the channel estimation process, a synergistic effect is obtained, and the ability to suppress intersymbol interference and the channel estimation accuracy can be improved.

The operation will be described. First, the first adaptive equalization processing and the second adaptive equalization processing are performed as usual. Then, the channel estimation is performed again. By repeating the adaptive equalization process and the channel estimation process in this way, the characteristics are improved.

Example 12: A computer simulation was conducted to confirm the performance of the present invention. The conditions are shown in Table 1.
Enumerate in.

[0074]

[Table 1] The parameters of the simulation are determined based on the specifications of the IEEE802.11a of 5 GHz band wireless LAN system. In order to perform the coherent detection, a channel estimation pilot is inserted in front of the data symbol by two symbols as shown in FIG. 4, and the length of the guard interval is twice that of the data symbol. As the propagation path model, the equal power two-path model is used, and the maximum Doppler frequency is 0H.
z. A Hanning window was used as the window function, and the number of taps for channel estimation was set to 25. FIG. 5 shows the average error rate characteristics when the delay time difference between the two paths is 1.2 μs (24Δ _t ). For comparison, the delay time difference in the coherent detection using the channel estimation pilot is shown within the guard interval and above the guard interval.

In the conventional synchronous detection, the power of the delay wave is large in the equal power two-path model, so that it can be confirmed that the transmission characteristic is significantly deteriorated when there is a delay of the guard interval or more. On the other hand, in the method of the present invention, as shown in FIG. 5, it is possible to obtain almost the same characteristics as the transmission characteristics within the guard interval, even in an environment in which a delayed wave exceeding the guard interval 0.8 μs exists. it can. In particular, since the signal band is wide, the environment where E _b / N ₀ exceeds 30 dB is stochastically low when considering outdoor use, and the E _b / N ₀ at which the system is actually operated is 10 to 10. Good characteristics are obtained in the range of 30 dB.

Note that the OFDM reception system of the present invention is not limited to the above-mentioned embodiments, and it goes without saying that various modifications can be made without departing from the scope of the present invention.

[0077]

According to the receiving method in OFDM of the present invention, good reception is possible even in the presence of inter-carrier interference at the present time and inter-symbol interference before the present time, the convergence of the algorithm is good, and the propagation path is good. There is an advantage that it can follow the high speed fluctuation of. Further, the size of the receiver can be made approximately the same as the conventional one, and since it is not necessary to change the transmission parameters, it is not necessary to change the specifications, and the excellent effect of not lowering the signal transmission efficiency is achieved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of a reception system in OFDM of the present invention.

FIG. 2 is a block diagram showing a channel estimator used in the present invention.

FIG. 3 is a diagram showing a Fourier transform section when a guard interval is incorporated.

FIG. 4 is a diagram showing a format of a pilot signal for channel estimation.

FIG. 5 is a diagram showing an average error rate characteristic of the system of the present invention.

FIG. 6 is a block diagram showing a configuration example when linear processing is used as the second adaptive equalization processing in the present invention.

FIG. 7 is a block diagram showing a configuration example when the first adaptive equalization processing is performed in the time domain according to the present invention.

FIG. 8 is a block diagram showing a configuration example in the case where a channel estimation process is combined with a processing process of a second adaptive equalization process in the present invention.

FIG. 9 is a block diagram showing a basic configuration of an OFDM transmitter.

FIG. 10 is a diagram showing a basic configuration of an OFDM signal.

FIG. 11 is a block diagram showing a basic configuration of an OFDM receiver.

FIG. 12 is a diagram showing a state of intersymbol interference.

[Explanation of symbols]

1 Guard interval remover 2 Fourier transformer 3-channel estimator 4 Synchronous detector 5 Serial / parallel converter 6 output terminals

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Kazuhiko Fukawa             2-12-1, Ookayama, Meguro-ku, Tokyo Tokyo Kogyo             Inside the university F term (reference) 5K022 DD01 DD13 DD23 DD24

Claims (15)

[Claims] 1. A Fourier transform process for performing a Fourier transform by multiplying an orthogonal frequency division multiplexed received signal by a window function,
A first adaptive equalization process for removing inter-symbol interference components before the present time from the signal after the Fourier transform; a second adaptive equalization process for eliminating inter-carrier interference between the current carriers;
And a channel estimation process for estimating a channel impulse used in the second adaptive equalization process. 2. The orthogonal frequency division multiplex signal reception system according to claim 1, wherein decision feedback adaptive equalization is used as the first adaptive equalization processing. 3. The orthogonal frequency division multiplex signal reception system according to claim 1, wherein maximum likelihood sequence estimation type adaptive equalization is used as the second adaptive equalization processing. 4. The orthogonal frequency division multiplex signal reception system according to claim 1, wherein delay decision feedback sequence estimation type adaptive equalization is used as the second adaptive equalization processing. 5. The orthogonal frequency division multiplexed signal according to claim 1, wherein a linear process for orthogonalizing a received signal in which inter-carrier interference between carriers at the present time is used is used as the second adaptive equalization process. Receiving method. 6. The orthogonal frequency division multiplex signal reception system according to claim 1, wherein in the Fourier transform processing, a window function other than a rectangular window is used when intersymbol interference is large, and a rectangular window is used when intersymbol interference is small. 7. In the channel estimation process, a replica of an interference signal of either or both of intersymbol interference before the present time point and inter-carrier intersymbol interference at the present time point is generated, and their amplitude and phase components are recursive least squares method. The method for receiving an orthogonal frequency division multiplex signal according to claim 1, which is estimated by. 8. The orthogonal frequency division multiplex signal reception system according to claim 1, wherein, in the Fourier transform processing, the time interval in which the Fourier transform is performed is a time interval including a guard interval. 9. In the Fourier transform processing, a time section in which the Fourier transform is performed is made to be from the head of a guard interval and two sections are made immediately after the guard interval, and adaptive equalization processing is performed in each of the two sections. The method for receiving an orthogonal frequency division multiplexed signal according to claim 1, wherein 10. The method for receiving an orthogonal frequency division multiplexed signal according to claim 1, wherein the first adaptive equalization processing is performed not on the signal after Fourier transform but on the received signal. 11. The orthogonal frequency division multiplex signal reception system according to claim 1, wherein in the second adaptive equalization processing, the channel estimation processing is combined with the processing step to perform the processing recursively. 12. The channel estimation process according to claim 1, wherein in the channel estimation process and the first adaptive equalization process, the determined signal, or the signal obtained by performing error correction decoding and then re-encoding is performed. Orthogonal Frequency Division Multiplexed Signal Reception System. 13. In the channel estimation processing, when known signals are discretely arranged between a transmitter and a receiver,
The method for receiving an orthogonal frequency division multiplexed signal according to claim 1, wherein the channel estimation process is performed again. 14. The orthogonal frequency division multiplexed signal according to claim 1, wherein adaptive equalization processing and channel estimation processing are repeated in the first adaptive equalization processing, the second adaptive equalization processing, and the channel estimation processing. Receiving method. 15. A Fourier transformer for performing a Fourier transform by multiplying a received signal subjected to orthogonal frequency division multiplexing by a window function,
A first adaptive equalizer that removes the intersymbol interference component before the present time from the signal after the Fourier transform, a second adaptive equalizer that removes the intersymbol interference between carriers at the present time, and the first and second adaptive equalizers. An orthogonal frequency division multiplexed signal receiver comprising a channel estimator for estimating a channel impulse used in a processor.