JP2001520492A - Method and receiver for channel estimation in a communication system - Google Patents

Abstract

(57) [Summary] According to the present invention, a reception signal composed of data symbols is received by a receiving station. This received signal is decomposed on the receiving side into individual sampled values and compared with known data symbols for the determination of the channel coefficients. In this case, the receiving station stores individual known data symbols of the received signal. By determining the reduced number of channel coefficients for a mobile station moving at high speed, the estimation accuracy for the channel coefficients is increased.

Description

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to a method and a receiving apparatus for channel estimation in a communication system having a transmission channel between a plurality of mobile stations at least one of which moves.

In a communication system, communication (eg, language, image information, or other data)
Is transmitted over a transmission channel, and in a wireless communication system this is done via a wireless interface using electromagnetic waves. The radiation of the electromagnetic waves takes place here at a carrier frequency in the frequency band set for the respective system. In GSM (Global System for Mobile Communication), the carrier frequency is 90
It is in the region of 0 MHz. Future wireless communication systems, such as UMTS (Universal
For mobile telecommunication systems or other third generation systems, the frequency is set to a frequency band of about 2000 MHz.

[0003] The emitted electromagnetic waves are attenuated based on reflections, diffractions and radiation due to the curvature of the ground and the like. As a result, the reception power that can be used in the radio station on the receiving side decreases. This attenuation is location dependent and, in a moving radio station, also time dependent. As the mobile station moves very quickly, the channel conditions of the transmission channel change significantly, even for short time intervals. In the case of multipath propagation, a plurality of signal components are delayed differently and arrive at a receiving-side radio station. The effects described above describe the transmission channels unique to the connection.

[0004] From DE 195 49 148 a radio communication system is known which uses subscriber separation according to the CDMA system (CDMA code division multiple access), wherein the radio interface additionally has a time division multiple subscriber separation (TDMA time division multiple access). Multiple access)
have. On the receiving side, a JD method (Joint Detection) is used to recognize the CDMA codes of a plurality of subscribers and to perform an improved detection of the transmitted data. It is known that at least two data channels can be assigned to the connection via the radio interface, wherein each data channel can be distinguished by a separate spreading code.

[0005] It is known from the GSM mobile radio network that data to be transmitted is transmitted as radio blocks (bursts) in time slots, in which a symbol known at the receiving radio station is replaced in the radio block. The transmitted midamble is transmitted. This type of midamble can be used in the sense of a training sequence for tuning the receiving side of the wireless station. The receiving radio station performs channel estimation based on the midamble, i.e., estimates the channel pulse response for different transmission channels. Alternatively, channels represented by different CDMA codes can be provided in parallel with the training sequence and the valid data.

[0006] If the channel conditions change rapidly, the channel pulse response of the conduction channel also changes within the transmitted radio block, so that only a portion of the radio block is used for channel estimation. A fixed number of equally spaced channel coefficients are determined regardless of channel conditions. However, such channel estimation is inaccurate.

[0007] It is an object of the present invention to provide a method and a receiving apparatus for channel estimation that can reliably determine a channel coefficient despite rapid movement of a mobile station. This object is achieved by a method having the features of claim 1 and claim 1.
The above problem is solved by a receiving device having a configuration described in one characteristic portion. Advantageous embodiments of the invention emerge from the dependent claims.

According to the invention, in a method for channel estimation in a communication system comprising a transmission channel between at least one moving radio station, a reception signal comprising data symbols is received by a receiving station. You. These data symbols are modified by the transmission method, for example by spreading, modulation and channel distortion.

At the receiving end, the received signal is decomposed into individual sampled values and compared with known data symbols to determine channel coefficients, wherein the individual known data symbols of the received signal are stored at the receiving station. ing. By determining a reduced number of channel coefficients for a rapidly moving mobile station, the estimation accuracy for these channel coefficients is improved. An erroneous relationship between the short sub-block of the received signal and the long desired channel pulse response does not occur, and the accuracy of the channel pulse response of the estimated channel is improved. For mobile stations that move slowly, a relatively large number of channel coefficients are determined.

This type of channel estimation also shifts the limits on the speed of the wireless station and the allowed channel conditions, without the need to simultaneously extend the training sequence or reduce the data transmission rate. The present invention is applicable to transmission between a base station and a mobile station as well as transmission between mobile stations.

According to an advantageous embodiment of the invention, the known data symbols are compared with the sampled values of the data symbols in the transmitted training sequence, or the known data symbols are stored after data compression and subsequently first It is compared with the received sampling value. This makes it possible to rely on known data symbols which are not distorted, and therefore a relatively low effective data rate allows a more accurate channel estimation. In a second variant, accurate data estimation is a prerequisite for the subsequent adjustment of the channel coefficients. But this gives
Channel estimation can be improved during evaluation of portions of the radio block that do not include a training sequence.

According to another aspect of the invention, the values of the known data symbols, ie the training sequence or the reduced number of estimated data symbols, are taken into account when determining the channel coefficients. I have. This provides an accurate and actual (real-time) estimate of the channel coefficients, even if the time interval for the sampled value to be evaluated is reduced.

It is further advantageous to select the channel coefficient to be determined such that it has a higher power than a value on average that is not selected. This ensures that even if the number of channel coefficients is reduced, the transmission channel is described sufficiently accurately and forms an excellent basis for subsequent data estimation using the channel coefficients. For example, the selected value may be derived from a relatively large number of previously determined channel coefficients. Thus, according to the invention, only significant channel coefficients are repeatedly calculated from the preceding channel estimates. At relatively large intervals, the selection of the channel coefficients to be determined is repeated so that the development of the meaning of the individual, not always calculated channel coefficients is taken into account.

According to an advantageous embodiment of the invention, the intervals between the selected values of the channel coefficients are not equal. In this way, a large temporal spread of the channel pulse response (caused by multipath propagation) can also be represented, corresponding to the channel pulse response of the unique channels of the different transmission channels. This is achieved by selecting the channel coefficients as significant at the beginning and end of the channel pulse response as needed. At this time, it is not necessary to make the intervals equal.

Due to the reduced number of determined channel coefficient values, significantly more redundancy (ue
berbestimmt) A simultaneous equation e = Gh + n is established. Here, G is a chip sequence matrix and n is a noise component. This redundancy improves the accuracy of the channel estimation. The accuracy of the decision can then be inferred physically. It is also advantageous to reduce the number of sampling values considered in channel estimation depending on the channel conditions of the transmission channel. This reduces redundancy but makes the channel estimation more actual.

In addition, it is advantageous to extend the spacing between the channel coefficients associated with the signal delay in order to more effectively take into account the long signal delays, but in this case the number of channel coefficients is reduced by the same amount. No need to increase. This allows the channel pulse response to take into account special channel conditions (e.g. in fjord terrain) and nevertheless require only a small sampling value for channel estimation.

Embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block circuit diagram of a mobile radio network.

FIG. 2 is a diagram showing a frame structure of the wireless interface.

FIG. 3 is a diagram showing the structure of the radio block.

FIG. 4 is a block circuit diagram of a receiver in a wireless station.

FIG. 5 is a block circuit diagram of the digital signal processing means.

FIG. 6 is a diagram showing a channel estimation problem.

FIG. 7 is a flowchart for channel estimation.

The radio communication system shown in FIG. 1 corresponds to a GSM mobile radio network known for its structure, which consists of a number of mobile switching centers MSC, which are either networked together or fixed. It forms the gate to the network PSTN. Furthermore, the mobile switching centers MSC are each connected to at least one base station controller BSC. Further, each base station controller BSC
As a result, a connection to at least one base station BS is realized. This type of base station BS is a radio station that can establish a radio connection to the mobile station MS via a radio interface.

FIG. 1 shows an example in which valid information ni and signaling information si are stored in three mobile stations M.
S1 and three radio connections V1. .
. V3 is depicted, where one mobile station MS has two data channels DK
1 and DK2 are assigned, and one data channel DK3 or DK4 is assigned to each of the other mobile stations MS. The operation and maintenance center OMC implements control and maintenance functions for the mobile radio network or a part thereof. Although the function of such a structure is utilized by the wireless communication system according to the present invention, the present invention can be applied to a different wireless communication system, and the present invention can be used therein.

The base station BS is connected to an antenna device, which comprises, for example, three individual radiators. Each of the individual radiators radiates towards one sector of the radio cell served by the base station BS. Alternatively, however, more individual radiators (such as adaptive antennas) can be used, in which case spatial subscriber separation is used according to the SDMA (Space Division Multiple Access) scheme. It is also possible to do.

Effective information ni and signaling information si between the base station BS and the mobile station MS
Are affected by multipath propagation, which is caused by reflections in buildings, for example, in addition to direct propagation paths. For this reason, the signal components of the same subscriber signal that may be associated with different propagation paths arrive at the receiving station at different times (delay spread).

Assuming that the mobile station MS is moving, multipath propagation is combined with yet another obstacle, and the signal components of various propagation paths by one subscriber signal at the receiving mobile station MS. Are superimposed depending on time. And in this case,
It is assumed that the subscriber signals of the various base stations BS are superimposed on the received signal rx in one frequency channel at the receiving point. The role of the receiving mobile station MS is to perform a sufficiently accurate and timely channel estimation on the transmission channel of the subscriber signal, even if the mobile station MS is moving at high speed, as well as to be transmitted in the subscriber signal. That is to detect valid information ni, signaling information si, and data symbols of organization information data.

FIG. 2 shows a frame structure of the wireless interface. According to the TDMA component, a wideband frequency range, for example, B = 1.6 MHz, is used for a plurality of time slots ts, for example, eight time slots ts1.
~ Ts8. Each time slot ts in the frequency range B is 1
Make up one frequency channel. Information on a plurality of connections is transmitted as a radio block in a frequency channel provided for valid data transmission.
According to the FDMA (Frequency Division Multiple Access) component, a plurality of frequency ranges B are allocated to the wireless communication system.

As shown in FIG. 3, these radio blocks for valid data transmission consist of a plurality of data parts dt with data symbols d, in which the midamble m known at the receiving end is contained. Section is embedded. The data d is spread by a connection-specific fine structure spreading code (CDMA code). As a result, for example, K data channels DK1, DK2, DK3,. . . DK4 can be separated by their CDMA components. These data channels DK1, DK2, DK3,. . . A predetermined energy E is assigned to each of the DKs 4 on a transmission side for each symbol.

By spreading each symbol of data d with Q chips, Q sub-intervals of period Tc can be transmitted within symbol period Ts. In this case, the Q chips form a unique CDMA code. The midamble m with the training sequence tseq also consists of L chips of period Tc. Further, a guard time guard of the period Tg is provided in the time slot ts, whereby various signal propagation times in the connection of the successive time slots ts are compensated.

In the broadband frequency domain B, successive time slots ts are divided according to a frame structure. Thus, eight time slots ts are combined into one frame, and certain time slots of the frame form a frequency channel for valid data transmission and are used repeatedly by the group of connections. Another frequency channel, for example for the frequency synchronization or time synchronization of the mobile station MS, is not introduced in each frame, but at a predetermined point in one multiframe.

The parameters of the wireless interface are as follows, for example.

Radio block duration 577 μs Number of chips per midamble m 243 Guard time Tg 32 μs Data symbols per data part N 33 Symbol duration Ts 6.46 μs Chips per symbol Q 14 Chip duration Tc 6/13 μs rise The parameters can be adjusted differently in the direction (MS → BS) and in the descending direction (BS → MS).

The receiver of FIG. 4 is associated with a radio station, which can be a base station BS or a mobile station MS. At the receiver, the receiving device according to the invention is used for channel estimation. A corresponding transmitter implementation is described, for example, in German patent specification DE 197.
34936.

FIG. 4 shows the receiving path of the device in detail. In the partial module E1, the reception signal rx is converted from the transmission frequency band to the low-pass region, and is divided into a real component and an imaginary component. In the partial module E2, analog low-pass filtering is performed, and subsequently, in the partial module E3, the received signal is oversampled twice by 13/3 MHz and a word width of 12 bits.

In the partial module E4, digital low-pass filtering is performed with a bandwidth of 13 /
The filtering is performed by a 6 MHz filter so that the sharpness of the edge becomes as large as possible for channel separation. Subsequently, the signal oversampled twice is reduced to 2: 1 in the partial module E4.

The received signal e obtained in this way substantially consists of two parts. That is, it is composed of a component em (training sequence tseg using known data symbols) for channel estimation and components e1 and e2 for data estimation. In the submodule E5, the estimation of the channel coefficient h ^(k) of the channel pulse response is performed using the known midamble basic code of all data channels transmitted in the respective time slots.

In the sub-module E 6, the parameters b ^(k) for the adapted filter for each data channel are detected using the CDMA code c ^(k) . In the sub-module E7, the interference caused by the midamble m ^(k) in the reception block e1 / 2 used for data estimation is removed. This is possible if h ^(k) and m ^(k) are known.

In the sub-modules E 8 to E 12, the detection of the data symbol d is performed using the pseudo inverse of the combined channel matrix A. An alternative solution is a singular value decomposition of the combined channel matrix A. Another solution is based on another optimization criterion, for example a minimum mean square error criterion (MMSE) instead of a zero insertion (ZF) criterion. Further, a feedback-coupled structure is also possible. These solutions can also be combined with one another.

In the sub-module E 8, a cross-correlation matrix A ^{* TA} is calculated. Since A ^{* TA} has a Toeplitz structure, only a small portion of the matrix needs to be calculated here. This small portion can then be used to expand to full size. When the movement of the mobile station MS is slow, this matrix A ^{* TA} is large. This is because a large partial block is selected. From this matrix A ^{* TA} only the small parts are calculated. If the movement is fast, the matrix A ^{* TA} becomes smaller. Therefore, in some cases, a perfect calculation is chosen to obtain a low-noise solution. In the submodule E9, the Cholesky decomposition of A ^{* TA} is H ^* TH
Executed in Here, H is an upper triangular matrix. Based on the Toeplitz structure of A ^{* TA} , H also has an approximate Toeplitz structure and does not need to be completely calculated. The vector s represents the reciprocal of the diagonal element of H. Diagonal elements can advantageously be used in solving equations.

In the partial module E 10, matched filtering (matched filter) of the received symbol sequence e 1 / e 2 is executed by b _(k) . The partial module E11 implements the device 1 for solving the equation for H ^{* T} * z1 / 2 = e1 / 2, and the partial module E12 implements the device 2 for solving the equation for H * d1 / 2 = z1 / 2. In the partial module E13, the estimated data d1 / 2 is demodulated, decoded, and subsequently convolutionally decoded by a Viterbi decoder. Decrypted data block

[0044]

[Outside 1]

[0045]

[Outside 2]

On the receiving side (see FIG. 5), after analog processing, that is, after amplification, filtering, and conversion to the inverse band in the high frequency section, digital low-pass filtering of the received signal rx is performed by the digital low-pass filter. The digitized part of the received signal e is represented by a vector em of length L = M * W and the data part dt
Does not include interference. This digital received signal portion is transmitted to digital signal processing means including a channel estimator KS. The data estimation in the joint detection data estimator is performed in common for all connections. For further details on this, please refer to German Patent Specification DE19734936.

Next, channel estimation will be described in detail. The partial module E5 has a channel estimator KS, a storage device SP and a control device SE. Information can be exchanged between them. The storage device SP stores the data symbols t of the training sequence tseq known from the midamble m, as well as the data symbols d to follow the already detected estimation. The control device SE can switch between the various modes of channel estimation on a subscriber-by-subscriber basis, where the channel estimation is performed by a channel estimator KS. The channel coefficient h is detected from the sampling values e1 to e16 of the received signal e by the channel estimation.

The channel estimation problem is shown in FIG. The transmitted data symbol d is transformed by the transmission scheme, is represented by the combined channel matrix A, and exists at the receiving wireless station as a sampling value e. The receiving wireless station attempts to detect the estimated value d ＾ for the transmitted data symbol d. This detection is performed by detecting an estimated and combined channel matrix A ＾ during channel estimation. The optimization criterion is, for example, that the square error of the deviation of the sampling value e with respect to the estimated received signal e ＾ is minimized. In the channel estimation from the training sequence tseq, d ＾ = t, and at the time of tracking, the already estimated d is feedback-coupled and d ＾
= D.

In the channel estimation in the mobile radio system, the motion of the mobile station MS must be considered. The motion causes a Doppler shift, which can be represented by changes in the amplitude and phase of the individual channel coefficients h (complex numbers). For this, see the channel model in German Patent Specification DE 196 35 271. With a number of equally spaced channel coefficients h, the value of each channel coefficient h changes, but the profile of the channel pulse response remains constant. Thus, along with the tracking of the channel coefficient h, the evaluation of the detected data symbol d
Even if the speed of the mobile station MS is relatively high, channel estimation can be further performed during the radio block.

If the speed is slow, the value of the channel coefficient h is based on the average over multiple radio blocks of the channel estimate associated with the midamble m. In the case of frequency hopping, only radio blocks of the same carrier frequency are considered. When the speed is medium, the average is suppressed, and when the speed is high, the tracking of the channel coefficient is also performed outside the midamble m formed by the training sequence tseq.
The mode switching of the channel estimation is performed by the control unit SE in the receiving device depending on the accuracy of the estimation, the measured rate of change of the channel pulse response, or the rate of change of the anticipation time (timing advance) which coincides with the signal propagation time. .

For tracking, the data part of the radio block is divided into partial blocks. This makes it possible to assume that the channel conditions change only negligibly during the partial block. The data detection uses the previously detected channel coefficients h, and the detected known data symbols d (e _{13 in} FIG. 4) of the partial block are fed back for channel estimation. The sub-blocks can be overlapped to minimize the delay of the channel estimation, so that, for example, the first sub-block contains data symbols 1 to 10, the second sub-block contains data symbols 2 to 11, and so on.

A problem arises in the context of a large number of long channel pulse responses in a short sub-block with known data symbols d or for a training sequence tseq with a limited number of known data symbols t in the midamble m. . In order to construct the channel estimation sufficiently accurately, a highly redundant system of equations must be determined.

The channel estimation (step 1 in FIG. 7) evaluates the time-dependent received signal during the training sequence tseq:

[0054]

(Equation 1)

Here, n (t) is a noise component, and x is a symbol for the convolution operation. In a matrix representation, this equation is written as e = Ad + n. The combined channel matrix A reproduces the effects of spreading by the CDMA code c and the modulation by the transmission channel having the channel coefficient h to be estimated.

This matrix equation can be solved by the method of least squares (Step 2 in FIG. 7).

[0057]

(Equation 2)

Here, A ⁺ is a generalized inverse matrix of matrix A (generalisierten Inverse der Matrix A)
Where A ′ is the Hermitian matrix of matrix A and d ＾ is the reconstructed vector of data symbols d (＾ is the symbol for the estimate.

[0059]

[Outside 3]

After the data estimation has been carried out, if necessary at the rapidly moving mobile station MS (step 3 in FIG. 7), in order to follow the channel coefficient h, the received signal e containing the valid data is The partial block of the part is newly evaluated.

[0061]

(Equation 3)

Here, c and d are combined as a combined chip sequence g, and a matrix expression: e = Gh + n is obtained. As a result, the channel coefficient h becomes

[0063]

(Equation 4)

(Step 6 in FIG. 7).

A large number of transmission channels can be described sufficiently accurately by a small number of relatively high power channel coefficients h. These channel coefficients are distributed over large time intervals, possibly due to signal delays.

The delay (position) of these important channel coefficients h is determined in step 4 (see FIG. 7). In the following, only these reduced numbers of channel coefficients h are determined in the evaluation of the partial blocks.

[0067] Sixteen sampling values e1, of the three subscriber signals A, B, C and the received signal e
. . . An example of a system of equations to solve having e16 has the form:

[0068]

(Equation 5)

This system of equations has only a little increased redundancy and therefore results in an estimation disturbed by noise. Chip sequence G and channel coefficient h
May be punctured. This leaves only significant channel coefficients h _s :

[0070]

(Equation 6)

Removing the punctured elements results in a simplified system of equations:

[0072]

(Equation 7)

[0073] high extremely redundancy for the channel coefficients h = h _s number the system of equations which is reduced. Therefore, the accuracy of the estimation (Step 5 in FIG. 7) is

[0074]

(Equation 8)

It is improved by

This procedure can also be used to increase the signal delay that can be exhibited by the channel pulse response. The number of channel coefficients h does not change. However, the spacing of the channel coefficients h is increased.

[Brief description of the drawings]

FIG. 1 is a block circuit diagram of a mobile radio network.

FIG. 2 is a diagram showing a frame structure of a wireless interface.

FIG. 3 is a diagram showing a structure of a radio block.

FIG. 4 is a block circuit diagram of a receiver in a wireless station.

FIG. 5 is a block circuit diagram of a digital signal processing means.

FIG. 6 is a diagram illustrating a channel estimation problem.

FIG. 7 is a flowchart for channel estimation.

[Explanation of symbols]

 MS Mobile station BS Base station DK Data channel BSC Base station controller OMC Operation and maintenance center

[Procedural Amendment] Submission of translation of Article 34 Amendment of the Patent Cooperation Treaty

[Submission date] April 14, 2000 (2000.4.14)

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Claims

[Correction method] Change

[Correction contents]

[Claims]

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0003

[Correction method] Change

[Correction contents]

[0003] The emitted electromagnetic waves are attenuated based on reflections, diffractions and radiation due to the curvature of the ground and the like. As a result, the reception power that can be used in the radio station on the receiving side decreases. This attenuation is location dependent and, in a moving radio station, also time dependent. As the mobile station moves very quickly, the channel conditions of the transmission channel change significantly, even for short time intervals. In the case of multipath propagation, a plurality of signal components are delayed differently and arrive at a receiving-side radio station. The effects described above describe the transmission channels unique to the connection. U.S. Pat. No. 5,323,422 states that the step size of the adaptive algorithm is:
A method is disclosed for channel estimation in a mobile radio station that is adjusted and set according to the noise present and the speed at which the mobile station moves. However, in this method, the total number of channel coefficients used for the algorithm always remains constant.

[Procedure amendment 3]

[Document name to be amended] Statement

[Correction target item name] 0007

[Correction method] Change

[Correction contents]

An object of the present invention is to provide a method and a receiver for channel estimation that can reliably determine a channel coefficient despite rapid movement of a mobile station. This object is achieved by a method having the configuration according to the features of claim 1 and a receiving device having the configuration according to the features of claim 8. Advantageous embodiments of the invention emerge from the dependent claims.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0062

[Correction method] Change

[Correction contents]

Here, c and d are combined as a combined chip sequence g, and a matrix expression: e = Gh + n is obtained. As a result, the channel coefficient h becomes

[Procedure amendment 5]

[Document name to be amended] Statement

[Correction target item name] 0072

[Correction method] Change

[Correction contents]

[0072]

(Equation 7)

──────────────────────────────────────────────────続 き Continuation of front page (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE ), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AU, BR, CA, CN, HU, ID, IL, JP, KR, MX, NO, PL, RU, UA , US, VN

Claims (15)

[Claims] 1. A channel estimation method in a communication system having a plurality of transmission channels between a base station (BS) and a mobile station (MS), wherein the base station (BS) and the mobile station (MS) can transmit and receive. Receiving a reception signal (e) including a data symbol modulated by a transmission method, decomposing the reception signal (e) into individual sampling values (e1 to e16) on a reception side, and MS, BS), each known data symbol (d, t) of the received signal (e) is stored. To determine the channel coefficient (h), the sampled values (e1 to e16) and the known Compare the data symbols (d, t) and reduce the number of channel coefficients (h) for the mobile station (MS) moving at high speed.
Is determined. 2. The method according to claim 1, wherein the known data symbol (t) is sampled by a data symbol sampling value (e1 to e1) in a transmitted training sequence (tseq).
The method according to claim 1, wherein the method is compared with (16). 3. The method according to claim 1, wherein the known data symbols (d) are stored after data detection and compared with the first received sampling values (e1 to e16). 4. The method according to claim 1, further comprising taking into account a reduced number of chip sequence matrices (G _s ) in determining the channel coefficients. 5. The method of claim 1, wherein the reduced number of selected channel coefficients (h) is:
5. The method of claim 4, wherein the power is greater than the unselected value. 6. The method according to claim 5, wherein a relatively large number of predetermined values of the channel coefficients (h) are excluded. 7. The method according to claim 1, wherein the intervals between the selected values of the reduced chip sequence matrix (G _S ) are unequal. 8. The following equation with high redundancy: e = Gh + n (n = noise component, G = chip sequence matrix) is created by reducing the number of determined channel coefficients (h). The method according to any one of claims 1 to 7. 9. A sampling value (e1 to e1) to be considered when estimating a channel.
9. The method according to claim 1, wherein the number of items in (6) is reduced depending on channel conditions of the transmission channel.
A method according to any one of the preceding claims. 10. The method according to claim 1, wherein the interval between the channel coefficients related to the signal delay time is increased without increasing the number of the channel coefficients. the method of. 11. Storage of sampling values (e1 to e16) of a reception signal (e) including data symbols modulated by a transmission method, and storage of known data symbols (d, t) of the reception signal (e). And a channel estimator (KS) for estimating a channel for a plurality of transmission channels. , The sampling values (e1 to e16) are compared with the known data symbols (d, t), and for a mobile station (MS) moving at high speed, a reduced number of channel coefficients (h)
Is determined so as to be determined. 12. The receiving apparatus according to claim 11, wherein the selected value of the reduced number of channel coefficients (h) has an average higher power than the unselected value. 13. A sampling value (e1 to e) to be considered when estimating a channel.
16. The number of 16) is reduced depending on the channel conditions of the transmission channel.
13. The receiving device according to 1 or 12. 14. The subscriber signal according to claim 11, wherein a number of subscriber signals superimposed on the received signal (e) in the frequency channel under the receiving device (EE) are transmitted via different transmission channels. The receiving device according to claim 1. 15. The receiving device according to claim 11, wherein the training sequence (tseq) including the known data symbol (t) and the subscriber signal are transmitted on the same frequency channel.