JP2005151377A - Method and apparatus for estimating transmission line characteristics in ofdm communication system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and apparatus for estimating transmission line characteristics in an OFDM communication system in which noise components can be remarkably eliminated. <P>SOLUTION: The apparatus for estimating the transmission line characteristics in the OFDM communication system comprises: a DCT circuit 7 that receives a preamble signal; a noise elimination circuit 8 that receives a signal from the DCT circuit 7; and an IDCT circuit 9 that receives a signal from the noise elimination circuit 8 and transmits the signal to a frequency base equivalent circuit 10. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a method and an apparatus for estimating transmission path characteristics on a frequency axis that are necessary for synchronous detection demodulation in an OFDM (Orthogonal Frequency Division Multiplexing) communication system.

  In recent years, there has been an increasing demand for multimedia wireless communication services for exchanging data integrating various information such as voice and images by wireless communication.

  In order to meet these requirements, it is essential to realize a wireless communication system that enables high-speed and high-quality data communication. However, in wireless communication, there is a problem that communication quality is significantly deteriorated due to multipath fading generated by a plurality of reflected waves. Therefore, OFDM (Orthogonal Frequency Division Multiplexing) communication schemes that transmit using a plurality of narrowband subcarriers having an orthogonal relationship on the frequency axis as a countermeasure for multipath fading are attracting attention (the following non-intervals). (See Patent Documents 1 and 2). By adopting the frequency axis equalization method, the OFDM method can use synchronous detection as a demodulation method even in a multipath fading environment. High-speed, high-quality data communication using a high-efficiency modulation method such as multilevel QAM It is possible to provide.

  The frequency axis equalization of the OFDM system is premised on high-accuracy estimation of transmission path characteristics. As a channel characteristic estimation method, a scheme using a preamble symbol composed of a known pilot signal is generally used.

  However, it has been pointed out that the method using this preamble symbol deteriorates the estimation accuracy due to the noise component contained in the transmission path band. On the other hand, a channel characteristic estimation method using a DFT (Discrete Fourier Transform) method that removes noise components on the time axis and improves estimation accuracy has been proposed (Patent Document 1 and Non-Patent Document 3 below). reference).

  This transmission path characteristic estimation method uses the feature that all impulse responses on the time axis of the transmission path characteristics exist within the guard interval (GI). However, the DFT method is based on ideal Nyquist sampling in which the number of FFT (Fast Fourier Transform) points on the transmission side is equal to the number of subcarriers, and the improvement effect deteriorates in the case of a general sampling interval. There's a problem.

  On the other hand, since the OFDM signal has an orthogonal relationship between subcarriers on the frequency axis, by using a scattered pilot (SP) in which pilot subcarriers are arranged at a certain interval, a plurality of preambles are used using one preamble. It is possible to collectively estimate the transmission path characteristics of independent lines.

  This technique is based on an OFDMA (OFDM-Access) method (see Non-Patent Document 4 below) or MIMO (Multi-Input Multi-Output) that requires efficient estimation of transmission path characteristics of a plurality of independent communication lines in a short time. ) Use in systems (see Non-Patent Document 5 below) and the like is being studied. Further, it is also used as a fading compensation method with time fluctuation by inserting SP in a data symbol (Non-Patent Document 6 below).

In the transmission line characteristic estimation method using these SPs, it is necessary to estimate the transmission line characteristic between pilot signals arranged at certain intervals by the frequency axis interpolation method. As a frequency axis interpolation method, a DFT interpolation method using a sampling function has been proposed so far. It is known that this DFT interpolation method causes a large distortion particularly in the transmission line characteristics at both ends when the number of FFT points and the number of subcarriers are different as in the above DFT method.
JP 2003-46473 A L. J. et al. Cimini, Jr., “Analysis and simulation of digital mobile channel using ortho frequency division multiplexing,” IEEE Trans. Comm. , Vol. COM-33, pp. 665-675 (1985). A. S. Bahai and B.I. R. Saltzberg, “Multi-Carrier Digital Communications,” Kluwer Academic Pub. (1999). Y. Zhao and A.M. Huang, “A Novel Channel Estimate Method for OFDM Mobile Communication Systems Based on Pilot Signals and Transform-Domain Processing,” Proc. of IEEE VTC '97 pp. 2089-2093. Y. R. Teng, A.M. Ito, T .; Nagaosa, K .; Mori and H.M. Kobayashi, “Proposal of OFDMA Scheme with Adaptive Subchannel Allocation for Wireless LAN Systems,” Proc. of ISITA 2002, pp. 531-534. P. Vandenameele, L.M. Van Der Perre, M.M. G. E. Engels, B.M. Gyselinckx and H.M. J. et al. De Man, “A Combined OFDM / SDMA Approach,” IEEE J. Select. Areas Commun. , Vol. 18, no. 11, pp 2312-2321, Nov. 2000. P. Dambacher, “Digital Terrestrial Television Broadcasting,” Springer, 1998. Translated by Tsuneo Saito “Digital Image Processing”, Science and Technology Publishing, 2002.

  As described above, the conventional transmission path characteristic estimation apparatus uses an estimation method using a preamble symbol which is a known pilot signal. Since the transmission path characteristics estimated by this method are affected by all of the noise added on the transmission line, there is a problem that the estimation accuracy is greatly deteriorated.

  As a method for improving this, a noise reduction method using the DFT method has been proposed (see Patent Document 1 and Non-Patent Document 3). In this method, the estimated channel characteristics on the frequency axis are subjected to DFT processing to obtain a time axis signal. The time axis signal obtained here corresponds to the time axis impulse response of the fading transmission path. Here, when ideal Nyquist sampling is assumed, since power concentrates on the signal component of the low-order impulse response, only the section in which the signal component of the impulse response exists is extracted and other noise components can be removed.

  However, in an actual system, it becomes impossible to set Nyquist sampling due to aliasing, and each delay time of the multipath delay wave has an arbitrary value. In other words, the noise component is removed and the signal component of the desired impulse response is removed at the same time, so that the estimated transmission path characteristics are distorted and the estimation accuracy is deteriorated.

  Hereinafter, the problems of the prior art will be described in detail.

First, problems of the conventional transmission path characteristic estimation method will be described.
[1] Transmission path characteristic estimation method using preamble Transmission path characteristics are generally estimated using a preamble symbol added to the head of a frame. Here, _assuming that M pilot data information on the frequency axis to be transmitted is an, the preamble signal b _k on the time axis subjected to IFFT (Inverse Fast Fourier Transform) is expressed by the following equation (1). .

ここで、ＮはＦＦＴポイント数、Ｍはサブキャリア数、（Ｎ−Ｍ）はエイリアス除去のために周波数軸上で挿入されるゼロパディング数を示す。また、複数の遅延波で構成されるマルチパスフェージングの時間軸上の伝送路インパルス応答ｈ_k は次式（２）によって表される。

Here, N is the number of FFT points, M is the number of subcarriers, and (N−M) is the number of zero padding inserted on the frequency axis for alias removal. Further, the transmission path impulse response h _k on the time axis of multipath fading composed of a plurality of delayed waves is expressed by the following equation (2).

ここで、Ｌは遅延波数、ρ_l ，θ_l ，τ_l は遅延波の振幅、位相、遅延時間を示す。式（１）、（２）よりマルチパスフェージング及び雑音を含む受信側での時間軸信号ｒ_k は次式（３）によって与えられる。

Here, L represents the delay wave number, and ρ _l , θ _l , and τ _l represent the amplitude, phase, and delay time of the delay wave. From the equations (1) and (2), the time axis signal r _k on the receiving side including multipath fading and noise is given by the following equation (3).

式（３）の受信信号ＲｍはＮポイントＦＦＴにより周波数軸信号に変換され次式（４）となる。

The received signal Rm in Expression (3) is converted into a frequency axis signal by N-point FFT, and becomes the following Expression (4).

ここで、Ｈ_m ，Ｚ_m は、周波数軸上での伝送路特性と雑音成分を示す。ここで、ａ_m はプリアンブルシンボルの情報であり受信側では既知であることを利用し、伝送路特性Ｈ_m は次式（５）によって推定できる。

Here, H _m and Z _m indicate transmission path characteristics and noise components on the frequency axis. Here, a _m is information on the preamble symbol and is known on the receiving side, and the transmission path characteristic H _m can be estimated by the following equation (5).

ここで、式（５）の第２項は雑音成分を示し、その平均電力は次式（６）によって求められる。

Here, the second term of the equation (5) indicates a noise component, and the average power is obtained by the following equation (6).

式（５）より、推定された伝送路特性には式（６）の雑音電力が含まれるために推定精度が劣化することがわかる。
〔２〕ＤＦＴ法を用いた伝送路特性推定方式
上記〔１〕の伝送路特性推定法に対して、推定精度の改善法（雑音成分の軽減法）としてＤＦＴ法が提案されている。これは、式（５）で得られた伝送路特性の時間軸信号は式（２）に示す伝送路の時間軸インパルス応答に相当することを利用している。

From Equation (5), it can be seen that the estimated transmission path characteristics include the noise power of Equation (6), so that the estimation accuracy deteriorates.
[2] Transmission Line Characteristic Estimation Method Using DFT Method In contrast to the transmission line characteristic estimation method of [1] above, the DFT method has been proposed as a method for improving estimation accuracy (noise component reduction method). This utilizes the fact that the time-axis signal of the transmission path characteristic obtained by Expression (5) corresponds to the time-axis impulse response of the transmission path shown in Expression (2).

  FIG. 9 is a diagram showing a configuration of a transmission path characteristic estimation apparatus using a conventional DFT method.

  In this figure, 101 is a GI elimination circuit, 102 is an FFT circuit, 103 is a switch, 104 is a transmission line characteristic estimation device, 105 is a transmission line characteristic estimation unit, 106 is an IDFT (Inverse Discrete Fourier Transform) circuit, and 107 is noise removal. The circuit, 108 is a DFT circuit, 109 is an equalizer, and 110 is a decision circuit.

  As shown in this figure, the GI removal circuit 101 that has received the received signal includes an FFT circuit 102, a switch 103, a transmission path characteristic estimation unit 105, an IDFT circuit 106, a noise removal circuit 107, a DFT circuit 108, and an equalizer 109. Through the determination circuit 110, the demodulated data can be output.

  In FIG. 9, the time axis signal after IDFT by the IDFT circuit 106 is expressed by the following equations (7) and (8) using the above equation (5).

式（８）において、特に（Ｍ／Ｎ）τ_l ＝η_l （整数）の関係を満足する場合は、式（８）は次式（９）によって表される。

In the equation (8), particularly when the relationship of (M / N) τ _l = η _l (integer) is satisfied, the equation (8) is expressed by the following equation (9).

式（９）が成立するのは、例えばＮ＝Ｍのナイキストサンプリングの場合や、各遅延波の遅延時間τとＮ，Ｍとの間に特殊な関係が有る場合である。このような場合には、式（９）より明らかなように時間軸上の伝送路特性のインパルス応答は上記式（２）と同様にｎ＝０からη_L-1 までの間でのみ存在することになる。

Formula (9) is established, for example, in the case of N = M Nyquist sampling or when there is a special relationship between the delay time τ of each delay wave and N, M. In such a case, as is clear from the equation (9), the impulse response of the transmission line characteristic on the time axis exists only between n = 0 and η _L-1 as in the above equation (2). It will be.

図９でＤＦＴ回路１０８によりＤＦＴ後の周波数軸上の伝送路特性は式（１０）をＤＦＴすることにより、次式（１１）のように与えられる。

In FIG. 9, the transmission path characteristic on the frequency axis after the DFT by the DFT circuit 108 is given by the following equation (11) by DFT the equation (10).

式（１１）の第１項は、式（２）で与えられる理想的な伝送路特性であり、第２項は雑音成分を示し、その平均電力は次式（１２）によって求められる。

The first term of the equation (11) is an ideal transmission path characteristic given by the equation (2), the second term indicates a noise component, and the average power is obtained by the following equation (12).

式（６），（１２）より、ＤＦＴ法を用いた伝送路特性推定方式では従来方式〔１〕と比較して、Δ＝１０ｌｏｇ（Ｍ／η_L-1 ）ｄＢの雑音電力が軽減されることになり、伝送路特性推定精度の改善が可能となる。

From equations (6) and (12), the transmission line characteristic estimation method using the DFT method reduces the noise power of Δ = 10 log (M / η _L−1 ) dB compared to the conventional method [1]. As a result, it is possible to improve the transmission path characteristic estimation accuracy.

However, such an improvement can be obtained when an ideal Nyquist sampling (N = M) is assumed or when there is a special relationship between the delay time τ of each delay wave and N, M. . In an actual apparatus configuration, zero padding is necessary on both sides of information data on the frequency axis for the purpose of removing aliases generated after D / A conversion on the transmission side, and generally N ≠ M. Further, each delay time of the multipath delay wave has an arbitrary value. In this case, the relationship of (M / N) τ ₁ = η ₁ (integer) does not hold from the equation (8), and the transmission line impulse response on the time axis extends over the entire time zone as shown in FIG. Will exist. This is because the transmission line impulse response based on the N-point periodic function as shown in the equation (4) is reproduced with the M-point periodic function.

Therefore, when removing the section N _P of the central portion for the purpose of improving estimation accuracy, as shown in FIG. 10, the estimation of channel characteristics will be also removed some of the desired signal components at the same time impulse response with noise components Accuracy will deteriorate.
[3] Frequency axis interpolation method using DFT method In the estimation of transmission path characteristics in a fading environment with time fluctuation, a scattered pilot (SP) in which pilot subcarriers are inserted at regular intervals between data subcarriers is used. The As described above, the use of this SP as a channel characteristic estimation method in the MIMO system shown in FIG. 11 is being studied. That is, as shown in FIG. 11, SPs interleaved with each other from each antenna are multiplexed and transmitted in one common preamble, and on the receiving side, the frequency axis interpolation method is used to make the connection between all antennas. The transmission path characteristics can be estimated.

  In addition, in an OFDMA system using the OFDM method as a multiple access method, the transmission path characteristics of independent communication lines from a plurality of user terminals to the base station are estimated using one preamble, as in the MIMO system. It becomes possible. Here, as a frequency axis interpolation method for SP, a DFT interpolation method using a sampling function is mainly studied. However, the DFT interpolation method is based on Nyquist sampling, and as described in [2] above, there is a problem that a large distortion occurs in the reproduced transmission path characteristics in the actual system configuration.

  The present invention is characterized by using a DCT (Discrete Cosine Transform) method that has already been adopted as a compression technique for an image signal in place of the DFT method in order to solve the problems of the conventional DFT method. . The DCT method is a kind of the same orthogonal transformation method as the DFT method, but enables a signal waveform to be expressed with high accuracy using only lower-order signal components as compared with the DFT method.

  That is, in the present invention, in the case of non-Nyquist sampling, which is a drawback of the transmission path characteristic estimation method using the DFT method, there is no problem that the time axis impulse response spreads over the entire time zone, and the time axis impulse response is low order. Therefore, the noise component can be largely removed without almost removing the signal component of the desired impulse response, and the estimation accuracy of the transmission path characteristics can be greatly improved. Similarly, the DCT method can be applied to the frequency axis interpolation method, and high-accuracy frequency axis interpolation can be realized.

  In view of the above situation, an object of the present invention is to provide a transmission path characteristic estimation method and apparatus in an OFDM communication system that can remove a significant noise component.

In order to achieve the above object, the present invention provides
[1] In the transmission path characteristic estimation method in the OFDM communication system, the impulse response on the time axis of the transmission path characteristic on the frequency axis utilizes the characteristic that power is concentrated on the low order component, and the low order component is Transmission path characteristic estimation is performed using a DCT (Discrete Cosine Transform) method that can be efficiently reproduced.

  [2] In the transmission path characteristic estimation method in the OFDM communication system described in [1] above, even if the noise component is largely removed on the time axis, the desired component of the time axis impulse response is removed compared to the DFT method. It is characterized by improving the estimation accuracy of the transmission path characteristics on the regenerated frequency axis.

  [3] The transmission path characteristic estimation method in the OFDM communication system described in [1] is characterized in that the DFT system extended to twice the sampling points equivalent to the DCT system is used.

  [4] In the transmission path characteristic estimation method in the OFDM communication system described in [1] above, when a scattered pilot signal is used, frequency axis interpolation is performed with high accuracy using a DCT system.

  [5] The method for estimating channel characteristics in the OFDM communication system according to [4] above, wherein frequency axis interpolation is performed with high accuracy even when the pilot signal interval K is large.

  [6] In a transmission path characteristic estimation apparatus in the OFDM communication system, a DCT circuit that receives a preamble signal, a noise removal circuit that receives a signal from the DCT circuit, a frequency axis equalization circuit that receives a signal from the noise removal circuit And an IDCT circuit for transmission to the network.

  According to the present invention, the following effects can be achieved.

  (1) A significant noise component can be removed.

  (2) By performing the IDCT operation, the transmission path characteristics on the frequency axis with reduced noise components can be reproduced.

  (3) It can be realized only by changing IDFT and DFT to DCT and inverse discrete cosine transform (IDCT), respectively, in the configuration of the conventional system.

  (4) High accuracy frequency axis interpolation is possible. That is, the frequency axis interpolation method using the DCT method of the present invention uses the characteristic that power concentrates on low-order signal components of the DCT method, and the frequency axis of the SP, which was difficult with the conventional DFT method, is used. It is possible to estimate the channel characteristics using interpolation with high accuracy.

  In the transmission path characteristic estimation apparatus in the OFDM communication system, a DCT circuit 7 that receives a preamble signal, a noise removal circuit 8 that receives a signal from the DCT circuit 7, and an IDCT circuit 9 that receives a signal from the noise removal circuit 8 are provided. It is possible to remove a significant noise component.

  Hereinafter, embodiments of the present invention will be described in detail.

  First, a transmission path characteristic estimation method using the DCT method will be described.

  FIG. 1 is a block diagram of an OFDM demodulator showing an embodiment of the present invention.

  In this figure, 1 is an OFDM demodulator, 2 is an A / D converter, 3 is a GI removal circuit, 4 is an FFT circuit, 5 is a switch, and 6 is a transmission path characteristic estimation apparatus according to the present invention. The characteristic estimation device 6 includes a DCT circuit 7, a noise removal circuit 8, and an IDCT circuit 9. Further, 10 is a frequency axis equalization circuit, and 11 is a determination circuit.

  An OFDM reception signal input to the OFDM demodulator 1 is A / D converted by an A / D converter 2 and then passed through a DCT circuit 7 that receives a preamble signal via a GI removal circuit 3, an FFT circuit 4, and a switch 5. The frequency axis transmission line characteristic sent to the removal circuit 8 and reproduced via the IDCT circuit 9 can be sent to the frequency axis equalization circuit 10 and demodulated data can be output from the decision circuit 11.

According to the present invention, a DCT method capable of efficiently reproducing a low-order component using the characteristic that the time-axis impulse response of the transmission path characteristic is concentrated on the low-order component is adopted, and the problems of the conventional DFT method To solve.
(1) Discrete cosine transform (DCT)
Similar to the DFT method, the DCT method is a kind of orthogonal transformation method and has already been adopted as a high-efficiency compression technique for image signals (see Non-Patent Document 7 above).

  FIG. 2 is a diagram comparing the conventional method (a) using DFT and the method (b) of the present invention using DCT.

  As shown in FIG. 2A, in the DFT method, N pieces of data are processed as signals that are periodically repeated. Therefore, as described in [2] above, higher-order signal components are generated due to discontinuities at both ends. On the other hand, as shown in FIG. 2B, the DCT method is processed as a signal having a cycle of 2N data in which N data are arranged on a mirror object. As a result, both ends are continuous, and the signal after DCT has a characteristic that power is concentrated on low-order signal components. Here, it is expected that a significant noise component can be removed by using the DCT method from the characteristic that the transmission path characteristics on the time axis are generally composed only of delayed waves in the GI. Is done.

(2) Transmission path characteristic estimation method using DCT method Transmission path characteristic estimation system using DCT method is the same as that in FIG. 1 for IDFT circuit 106 and DFT circuit 108 in the configuration of the conventional system shown in FIG. This can be realized simply by changing to the DCT circuit 7 and the IDCT circuit 9. That is, the transmission line characteristics on the frequency axis estimated by the equation (5) are subjected to DCT processing by the following equation (14).

ただし、Ｗ_n はＤＣＴの係数であり次式（１５）によって表される。

Here, W _n is a DCT coefficient and is expressed by the following equation (15).

式（１４）で得られる時間軸上のインパルス応答は、ＤＦＴ法の場合と異なりナイキストサンプリング以外の任意のサンプリング間隔においても図３に示すようにＰ₁ 以下の低次の時間成分に集中する。従って、任意のサンプリング間隔においても次式の関係が成立する。

Unlike the case of the DFT method, the impulse response on the time axis obtained by Expression (14) concentrates on low-order time components of P ₁ or less as shown in FIG. 3 even at an arbitrary sampling interval other than Nyquist sampling. Accordingly, the relationship of the following equation is established at an arbitrary sampling interval.

式（１６）の関係を利用して、ＩＤＣＴ操作を行うことにより雑音成分を軽減した周波数軸上の伝送路特性を次式のように再現することができる。

By utilizing the relationship of Expression (16), the transmission path characteristics on the frequency axis with the noise component reduced by performing the IDCT operation can be reproduced as the following expression.

ＤＣＴ法を用いた伝送路特性の推定精度については後述する。

The estimation accuracy of transmission path characteristics using the DCT method will be described later.

(3) Frequency Axis Interpolation Method Using DCT Method FIG. 4 shows an arrangement example of pilot signals composed of SP.

Here, when the SP interval is K, K types of independent SP arrays are possible. Further, when the SP interval is K, the amplitude value of each pilot signal can be set to √K times by assuming that the average signal power of the preamble symbol is constant. With the configuration of FIG. 4, 2 ^K types of independent transmission path characteristics can be estimated using one preamble symbol in a MIMO system including K transmission / reception antennas. Here, the frequency number of the first pilot signal for the J-th SP array is J. J is a number from 0 to K-1. Here, the channel characteristic on the frequency axis estimated using the J-th SP array is expressed by the following equation (18) by setting m = iK + J in equation (5).

ただし、式（１８）では雑音成分は省略している。

However, the noise component is omitted in Equation (18).

Next, when DCT processing is performed on the equation (18) composed of M _P (= M / K) points, the following equation (19) is obtained.

ここで、式（１９）で得られる時間軸に相当するＭ_P 個のインパルス応答はＤＣＴ法の特徴から低次の時間成分に集中している。従って、Ｍ_P 付近のインパルス応答は十分小さな振幅値となる。これら関係を利用することにより、式（１９）は次式（２０）のように表される。

Here, M _P impulse responses corresponding to the time axis obtained by Equation (19) are concentrated on low-order time components due to the characteristics of the DCT method. Thus, the impulse response of the near M _P becomes sufficiently small amplitude. By using these relationships, the equation (19) is expressed as the following equation (20).

式（２０）の関係を利用して、式（１９）をＩＤＣＴ処理することにより周波数軸補間されたＭポイントの伝送路特性は次式（２１）によって求めることができる。

Utilizing the relationship of equation (20), the transmission path characteristics of M points subjected to frequency axis interpolation by subjecting equation (19) to IDCT processing can be obtained by the following equation (21).

ただし、ｂはＳＰの最初のパイロット信号番号Ｊの値によって異なる値を取る補正項とする。式（２１）は、上で設定したｍ＝ｉＫ＋Ｊの関係を利用すると、次式（２２）となる。

However, b is a correction term that varies depending on the value of the first pilot signal number J of SP. Equation (21) becomes the following equation (22) when the relationship of m = iK + J set above is used.

ここで、２番目のｃｏｓ中の（２Ｊ＋ｂ）／Ｋが１を取る時に直交関数の関係からｃｏｓ項の積の和は１となり、式（２２）は、式（４）で与えられるＭポイントから構成される伝送路特性と同じとなる。すなわち、ｂ＝Ｋ−２Ｊに設定し式（２０）にＭ_P ポイントのＩＤＣＴ処理を行うことにより、高精度の周波数軸補間が可能となる。ここで本発明のＤＣＴ法を利用した周波数軸補間法は、ＤＣＴ法の有する低次の信号成分に電力が集中する特徴を利用しており、従来のＤＦＴ法では困難であったＳＰの伝送路特性の推定を高精度に実現可能としている。
（４）特性評価
ここでは、上記（３）で提案したＤＣＴ法を用いた伝送路特性推定方式の各種伝送特性について計算機シミュレーションで評価する。表１にシミュレーション仕様を示す。

Here, when (2J + b) / K in the second cos takes 1, the sum of the products of the cos terms becomes 1 from the relationship of the orthogonal function, and the equation (22) is obtained from the M point given by the equation (4). It is the same as the configured transmission path characteristics. That is, by performing the IDCT processing of M _P point set b = K-2J formula (20), thereby enabling the frequency axis interpolation precision. Here, the frequency axis interpolation method using the DCT method of the present invention uses the characteristic that the power concentrates on the low-order signal components of the DCT method, and the SP transmission line, which was difficult with the conventional DFT method, is used. It is possible to estimate the characteristics with high accuracy.
(4) Characteristic Evaluation Here, various transmission characteristics of the transmission path characteristic estimation method using the DCT method proposed in (3) above are evaluated by computer simulation. Table 1 shows the simulation specifications.

マルチパスフェージングとしては、４０波指数関数遅延プロファイルを想定し、各遅延波は独立にレイリー変動するものとした。また、特性評価に際してはシンボル同期と周波数同期を理想とした。

As multipath fading, a 40-wave exponential function delay profile is assumed, and each delayed wave is assumed to undergo Rayleigh fluctuation independently. In the characteristic evaluation, symbol synchronization and frequency synchronization were ideal.

  FIG. 5 shows amplitude levels of the time axis impulse response in the DTF method shown in FIG. 10 and the time axis impulse response in the DCT method according to the present invention shown in FIG. The figure also shows the noise level. As is apparent from FIG. 5, the DCT method concentrates power on lower-order signal components as compared to the DFT method, and the amplitude value of the impulse response component rapidly decreases as the order is increased. I understand.

  As a result, even if the noise component is largely removed on the time axis, the DCT method can reduce the rate of removal of the desired component compared to the DFT method, and can improve the estimation accuracy of the transmission path characteristics on the reproduced frequency axis. It can be seen that it can be improved.

FIG. 6 is a diagram showing BER (Bit Error Rate) characteristics when the modulation method is the QPSK method. In this figure, the delay dispersion of multipath fading is 100 ns, the number of FFT points N is 512, the number of subcarriers M is 384, and the number of zero padding inserted for the purpose of alias removal is 128 (= 512-384). noise removal section N _P is the result of the case of fixing utilized irrespective of 40 C / N, the horizontal axis represents C / N (frequency-time axis) (dB), the ordinate indicates the BER, a line The BER characteristic in the case of using an ideal transmission line characteristic not including noise, the B line indicates the BER characteristic by the DCT method of the present invention, and the C line indicates the BER characteristic by the conventional preamble method. FIG. 6 shows that the method of the present invention can achieve almost the same characteristics as the ideal BER characteristics. Also, it considered an effective method in the actual apparatus configuration because it can set a fixed N _p value regardless of C / N.

  FIG. 7 shows the BER characteristics when the pilot signal interval (K) is changed when the modulation method is the QPSK method, and the delay dispersion of multipath fading is 100 ns. The horizontal axis represents the pilot signal interval (K), the vertical axis represents BER, the A line represents the BER characteristic according to the DCT method of the present invention when C / N is 12 dB, and the B line represents the present invention when C / N is 24 dB. BER characteristics by DCT method, C line is BER characteristic by DCT method of the present invention when C / N is 40 dB, D line is BER characteristic by conventional DFT method when C / N is 12 dB, E line is C The BER characteristic by the conventional DFT method when / N is 24 dB, and the F line shows the BER characteristic by the conventional DFT method when C / N is 40 dB.

FIG. 8 is a diagram showing C / N vs. BER characteristics when the modulation method is the QPSK method. The multipath fading delay dispersion is 100 ns, the FFT point number N is 512, the subcarrier number M is 384, and the noise removal interval N _p. Is 40 and the pilot signal (SP) interval K is 8. The horizontal axis represents C / N (dB), the vertical axis represents BER, the A line represents the BER characteristic when the transmission line characteristic is ideal, the B line represents the BER characteristic according to the DCT method of the present invention, and the C line represents the conventional DFT. The BER characteristic by a method is shown.

From FIG. 8, it can be seen that the method of the present invention can achieve almost the same BER characteristics as ideal characteristics. Also, it considered an effective method in the actual apparatus configuration because it can set a fixed N _P value regardless of C / N. In contrast, in the DFT method, it is necessary to set the optimum N _P for each C / N, and is at C / N is high are significantly degraded compared to the characteristics of the ideal case I understand that.

  From the above results, it was proved that the frequency axis interpolation method of the present invention can be sufficiently used as a transmission path characteristic estimation method in an OFDMA or MIMO system.

  In the present invention, the problems of the conventional transmission path characteristic estimation method based on the DFT method have been clarified, and a transmission path characteristic estimation system using discrete cosine transform that solves these problems has been proposed. It was also clarified that this proposed method can be used as a frequency axis interpolation method using scattered pilots. The effectiveness of the present invention was verified by computer simulation.

  In addition, this invention is not limited to the said Example, Based on the meaning of this invention, a various deformation | transformation is possible and these are not excluded from the scope of the present invention.

  The present invention is not limited to terrestrial digital TV transmission systems and wireless LAN systems that currently employ the OFDM communication system as a standardized communication system, but is also considered as a promising transmission system for next-generation wireless communication systems. It can be used as a transmission path characteristic estimation method such as an OFDMA or MIMO system.

It is a block diagram of an OFDM demodulator showing an embodiment of the present invention. It is the figure which contrasted the conventional system (a) using DFT, and the system (b) of this invention using DCT. It is a figure which shows the time-axis impulse response by DCT method. It is a figure which shows the example of arrangement | positioning of a scattered pilot signal. It is a figure which shows the comparison of the amplitude convergence characteristic of the time-axis impulse response in DCT method and DFT method. It is a figure which shows about the BER (Bit Error Rate) characteristic at the time of using this invention, when a modulation system is a QPSK system. It is a figure which shows the relationship between the pilot signal space | interval (K) and the BER characteristic in the frequency-axis interpolation system using this invention, when a modulation system is a QPSK system. It is a figure which shows the BER characteristic at the time of using the frequency-axis interpolation method of this invention in the case of a QPSK system. It is a figure which shows the structure of the transmission-line characteristic estimation apparatus using the conventional DFT method. It is a figure which shows the transmission-line impulse response on the conventional time axis. It is a figure which shows the transmission-line characteristic estimation method in a MIMO system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 OFDM demodulator 2 A / D converter 3 GI removal circuit 4 FFT circuit 5 Switch 6 Transmission path characteristic estimation device 7 DCT (discrete cosine transform) circuit 8 Noise removal circuit 9 IDCT (inverse discrete cosine transform) circuit 10 Frequency axis, etc. Circuit 11 Judgment circuit

Claims (6)

  A DCT (Discrete Cosine Transform) method that can efficiently reproduce the low-order component using the characteristic that the impulse response on the time axis of the transmission line characteristic on the frequency axis concentrates power on the low-order component. A method of estimating transmission path characteristics in an OFDM communication system, characterized in that transmission path characteristics are estimated.   2. The method of estimating channel characteristics in an OFDM communication system according to claim 1, wherein even if the noise component is largely removed on the time axis, the ratio of removal of the desired component of the time axis impulse response is small compared to the DFT method. A method of estimating transmission path characteristics in an OFDM communication system, which improves the estimation accuracy of transmission path characteristics on a reproduced frequency axis.   2. The method of estimating transmission path characteristics in an OFDM communication system according to claim 1, wherein the DFT system is extended to double sampling points equivalent to the DCT system.   2. The method for estimating channel characteristics in an OFDM communication system according to claim 1, wherein when a scattered pilot signal is used, frequency axis interpolation is performed using the DCT method. .   5. The method of estimating channel characteristics in an OFDM communication system according to claim 4, wherein the frequency axis interpolation is accurately performed even when the pilot signal interval K is large. (A) a DCT circuit that receives a preamble signal;
(B) a noise removal circuit for receiving a signal from the DCT circuit;
(C) An IDCT circuit that receives a signal from the noise removal circuit and transmits the signal to a frequency axis equalization circuit, and a transmission path characteristic estimation apparatus in an OFDM communication system.

Images