CN101814931A

CN101814931A - Doppler frequency shift estimation and compensation method in TD-SCDMA (Time Division-Synchronization Code Division Multiple Access) system

Info

Publication number: CN101814931A
Application number: CN200910006579A
Authority: CN
Inventors: 朱昀; 王曼
Original assignee: ZTE Corp
Current assignee: ZTE Corp
Priority date: 2009-02-19
Filing date: 2009-02-19
Publication date: 2010-08-25
Anticipated expiration: 2029-02-19
Also published as: CN101814931B; WO2010094191A1

Abstract

The invention relates to a Doppler frequency shift estimation and compensation method in a TD-SCDMA (Time Division-Synchronization Code Division Multiple Access) system. The method comprises the following step of: after a mobile terminal receives a time slot containing the data of the mobile terminal, obtaining a Doppler frequency offset value of a receiving signal of the current time slot through comparing the phase differences of channel estimation sequences of the current time slot and a comparison time slot. The compensation method of the Doppler frequency shift in the TD-SCDMA system comprises the following steps of: after the mobile terminal receives the time slot containing the data of the mobile terminal, obtaining the Doppler frequency offset value of the receiving signal of the current time slot through comparing the phase differences of the channel estimation sequences of the current time slot and the comparison time slot; and correcting the receiving signal in joint detection by combining with the joint combination. The invention improves the accuracy of channel estimation, eliminates the influences of the Doppler frequency shift on correctly receiving demodulation, reduces the error rate and improves the system performance and is particularly suitable for TD-SCDMA terminals under high-speed motion scenes.

Description

Doppler frequency shift estimation and compensation method in TD-SCDMA system

Technical Field

The present invention relates to mobile communication technology, and more particularly, to a method for estimating and compensating doppler shift in a time division synchronous code division multiple access (TD-SCDMA) system.

Background

The TD-SCDMA (time division synchronous code division multiple access) mobile communication system is a third generation mobile communication system based on time division duplex and code division multiple access, and has the characteristics of flexible uplink and downlink configuration, low code rate, high spectrum utilization rate and the like.

As shown in fig. 1, in the TD-SCDMA system, the basic unit for transmitting data is a radio frame. 7 normal time slots (TS 0-TS 6) and 3 special time slots (downlink synchronization code segment DwPTS, guard interval GP and uplink synchronization code segment UpPTS) form a subframe, and two subframes (subframe #1 and subframe #2) form a wireless frame with the duration of 10 ms. The mobile terminal user's data is transmitted in regular time slots, one of which is 675 mus long and is made up of 864 chips (chips), each of which is 0.78125 mus long. These chips are divided into four parts: two data segments (352 chips each), a midamble (144 chips) and a guard interval GP of 16 chips length; each data segment is formed by spreading, scrambling and aliasing data of a plurality of mobile terminal code channels to be transmitted; the training code segment is formed by shifting basic midamble code allocated by the system, and the training code segment is used as a training sequence for channel estimation.

According to the current TD-SCDMA technology, the maximum moving speed of the mobile terminal supported by the TD-SCDMA technology is 120 km/h. However, in real life, high-speed trains with the speed per hour being as high as 250 km or more have appeared, and high-speed railway networks with the higher speed per hour being 300 km/h-500 km/h can be spread across the whole country in the near future, which provides great challenges for the existing TD wireless technology. When the mobile terminal is in a high-speed motion state, the doppler shift suffered by the spatially transmitted wireless signal becomes very serious, and the doppler shift increases with the increase of the carrier frequency, and the doppler shift is in a direct proportion relationship with the carrier frequency and the speed of the mobile terminal, that is:

wherein f is_dRepresenting the Doppler frequency offset, f_RFRepresenting the carrier frequency, v the speed of movement of the mobile terminal, c the speed of light, equal to 3 x 10⁸m/s, θ represents the angle between the direction of motion of the mobile terminal and the incidence of radio waves on the mobile terminal. On the mobile terminal side, the doppler shift effect causes a deviation between the frequency of the locally demodulated carrier and the signal actually received by the mobile terminal, and such deviations in frequency, when accumulated over time, cause the demodulated symbols to be deflected (distorted) in phase with the standard modulated symbols. Especially for TD system, because it is a low code rate system, the time duration of each chip is long, and the phase distortion of the demodulation symbol accumulated on each chip becomes larger, which will greatly affect the correct reception and demodulation of the mobile terminal.

On the other hand, the joint detection technology is generally adopted when the TD-SCDMA mobile terminal performs receiving demodulation. The joint detection uses the prior information in the multiple access interference to treat the separation of all mobile terminal signals as a uniform process, and converts the received aliasing chip signals into demodulation symbols of each mobile terminal in one step, thereby reducing the mutual interference of multiple mobile terminals and increasing the system capacity. However, the effectiveness of the joint detection technology is established on the basis of accurate channel estimation, and the existing implementation of the TD channel estimation adopts a method of low-cost Fast Fourier Transform (FFT)/Inverse Fast Fourier Transform (IFFT) based on Steiner and a subsequent detection threshold noise tap removal. This technique is described in Steiner B.BAIEP, Low cost channel estimation in the uplink receiver of CDMAmobile radio systems, Frequz 1993, 47 (12): 292 & 298 & kang Shao Li et al, communication bulletin, volume 23, No. 10, No. 125-130, of the improvement of the low cost channel estimation method in TD-SCDMA System. This channel estimation technique is slot-based, i.e., the channel is considered to be fixed and invariant for one slot time; in practice, fading experienced by the mobile terminal during motion is modulated by doppler shift, that is, the channel is time-varying in a time slot, and the channel value estimated by the existing estimation method is actually the average value of the channel in a time slot period. Therefore, an error exists between the estimated channel obtained by the estimation method and a real channel, especially, the error is very large in a high-speed motion state of the mobile terminal, and if the estimated channel with the error is brought into the joint detection matrix, repeated superposition and amplification of interference can be caused, so that correct decoding of data is influenced.

Disclosure of Invention

The technical problem to be solved by the present invention is to provide a method for estimating and compensating doppler shift in TD-SCDMA system, so as to overcome the adverse effect on correctly receiving data caused by the movement of mobile terminal.

In order to solve the above problem, the present invention provides a method for estimating a doppler frequency offset value in a TD-SCDMA system, comprising:

after receiving the time slot containing the data of the mobile terminal, the mobile terminal obtains the Doppler frequency offset value of the received signal of the current time slot by comparing the phase difference of the channel estimation sequence of the current time slot and the comparison time slot.

Further, the above method may also have the following features:

the mobile terminal is stored with n preset range intervals of absolute values of Doppler frequency offset values and a comparison time slot corresponding to each range interval, wherein the n range intervals cover all possible absolute values of the Doppler frequency offset values section by section, and n is more than or equal to 1;

the mobile terminal obtains the doppler frequency offset value of the received signal of the mobile terminal by comparing the phase difference between the current time slot and the channel estimation sequence of the comparison time slot, which means that:

judging which range interval the absolute value of the prejudged value of the Doppler frequency offset value of the current time slot is in;

determining a corresponding comparison time slot according to the judged range interval;

and estimating the phase offset of the received signal caused by the Doppler frequency shift in each chip length of the time slot by comparing the phase difference of the current time slot and the comparison time slot on the corresponding path.

Further, the above method may also have the following features:

the prejudged value of the Doppler frequency offset value of the current time slot is the Doppler frequency offset value of a last processing time slot of the mobile terminal;

the doppler frequency offset value of the last processing slot of the mobile terminal is: the Doppler frequency offset value of the mobile terminal on the strongest path of the time slot, or the maximum Doppler frequency offset value of the mobile terminal in each path of the time slot, or the weighted sum of the two items, which is stored in the mobile terminal and is estimated when the time slot containing the mobile terminal data is processed.

Further, the above method may also have the following features:

the step of comparing, by the mobile terminal, the phase difference between the current time slot and the channel estimation sequence of the comparison time slot to obtain the doppler frequency offset value of the signal received by the mobile terminal at the current time slot specifically includes:

intercepting the training sequence of the comparison time slot, performing channel estimation, and estimating the average channel impulse response of each mobile terminal in the comparison time slot;

intercepting a training sequence of the current time slot, performing channel estimation, and estimating the average channel impulse response of each mobile terminal in the current time slot;

and the mobile terminal part of the average channel impulse response of the current time slot and the comparative time slot performs phase subtraction according to the corresponding path, and then divides the chip interval length between the two time slots to obtain the phase offset of the received signal of the mobile terminal on each path and each chip caused by Doppler frequency shift.

Further, the above method may also have the following features:

when n is 2, the range interval includes: a ═ 0, X), B ═ X, + ∞), the comparison time slot corresponding to the range interval a is the time slot with the same time slot number as the current time slot in the last subframe, the comparison time slot corresponding to the range interval B is the TS0 time slot in the subframe where the current time slot is, wherein the value of X is determined according to engineering practice.

Further, the above method may also have the following features:

when the comparison time slot is the TS0 time slot in the frame, intercepting the training sequence of the TS0 time slot, performing channel estimation, and estimating the average channel impulse response of each mobile terminal in the comparison time slot; when the comparison time slot is the time slot which is in the same time slot as the current time slot in the previous frame, the average channel impact response of the comparison time slot is obtained according to the storage record of the mobile terminal;

intercepting a training sequence of the current time slot, performing channel estimation, and estimating the average channel impulse response of each mobile terminal in the current time slot period;

and the mobile terminal performs phase subtraction on the channel estimation value of the strongest path in the current time slot and the comparison time slot, and then divides the phase subtraction by the chip interval length of the current time slot and the comparison time slot to obtain the phase offset of the received signal of the mobile terminal on each chip caused by Doppler frequency shift.

The invention also provides a method for compensating the Doppler frequency shift in the TD-SCDMA system, which comprises the following steps:

after receiving the time slot containing the data of the mobile terminal, the mobile terminal obtains the Doppler frequency offset value of the received signal of the current time slot by comparing the phase difference of the channel estimation sequence of the current time slot and a comparison time slot, and corrects the received signal in the joint detection by combining the joint detection.

Further, the above method may also have the following features:

the modifying the received signal of each mobile terminal in the joint detection specifically includes:

multiplying the obtained phase offset by a spreading factor SF, and then compensating the average channel impulse response of each mobile terminal in the current time slot period according to the path;

multiplying the spread spectrum codes, channel codes and scrambling codes of each mobile terminal to respectively obtain composite spread spectrum codes of each mobile terminal;

and constructing a joint detection matrix A by using the compensated average channel impulse response and the composite spread spectrum codes of all the mobile terminals, and carrying out joint detection on the data segment in the current time slot by using the joint detection matrix A to obtain a modified demodulation symbol.

Further, the method may further include:

and comparing the phase of the obtained demodulation symbol with the phase of a standard modulation symbol in a constellation diagram, smoothing the phase frequency offset value of the strongest path and/or the maximum phase frequency offset value of each mobile terminal by the obtained phase difference, and storing the processed result as the Doppler frequency offset value of the time slot.

Compared with the prior art, the invention improves the accuracy of channel estimation, eliminates the influence of Doppler frequency shift on correct receiving demodulation, and reduces the error rate, thereby improving the system performance, and being particularly suitable for the scene of the TD-SCDMA mobile terminal under high-speed motion. The accuracy of estimating the Doppler frequency offset is mediated by selecting a proper comparison time slot, the Doppler frequency offset is estimated by performing phase difference operation on the comparison time slot and the current time slot, and under the preferable condition, the mobile terminal only estimates and compensates the phase offset of the received signal caused by the Doppler frequency offset of the strongest path (main path), so that the implementation complexity of the method is reduced; in addition, the mobile terminal can further reduce the complexity of the method and the system overhead by storing the channel estimation sequence of the comparison time slot.

Drawings

FIG. 1 is a schematic diagram of a TD-SCDMA frame structure in the prior art;

FIG. 2 is a diagram illustrating the phase shift of each path of a wireless channel due to Doppler shift modulation during a conventional time slot according to an embodiment of the present invention;

FIG. 3 is a flow chart of a method in an embodiment of the invention;

fig. 4 is a schematic diagram of a Steiner channel estimation sequence and channel windows corresponding to mobile terminals in the embodiment of the present invention;

FIG. 5 is a block diagram of a joint detection A matrix according to an embodiment of the present invention;

FIG. 6 shows a simulation result of the system according to the embodiment of the present invention.

Detailed Description

The technical solution of the present invention will be described in more detail with reference to the accompanying drawings and examples.

For a TD time slot traveling over a wireless channel, it is subject to convolutional modulation of the wireless channel impulse response. The impulse response is composed of two parts: fading due to multipath effects: since the duration (675 μ s) of a TD slot is much smaller than the coherence time of the spatial channel, this portion can be considered approximately fixed within a slot period; 2 phase shift caused by Doppler effect

Offset amount

Linearly over time within the duration of a time slot, as shown in fig. 2, where c_iIndicating the ith chip in the slot, each slope representing the phase offset on the corresponding radial (the slope of each slope, i.e. the slope, due to the different angle between the radial direction and the direction of motion of the mobile terminal)

And different as well). And sequentially spreading the space channels of a certain mobile terminal on a time slot length according to chips to obtain:

<math><mrow><mfenced open='{' close=''><mtable><mtr><mtd><mi>h</mi><mrow><mo>(</mo><mi>chip</mi><mo>_</mo><mn>1</mn><mo>)</mo></mrow><mo>=</mo><mo>[</mo><msub><mi>cof</mi><mrow><mi>fadmg</mi><mo>_</mo><mn>1</mn></mrow></msub><mo>·</mo><msup><mi>e</mi><mrow><mo>-</mo><mi>j</mi><mn>2</mn><msubsup><mi>πf</mi><mi>d</mi><mn>1</mn></msubsup><mo>-</mo><msub><mi>t</mi><mrow><mi>chip</mi><mn>1</mn></mrow></msub></mrow></msup><mo>,</mo><msub><mi>cof</mi><mrow><mi>fadmg</mi><mo>_</mo><mn>2</mn></mrow></msub><mo>·</mo><msup><mi>e</mi><mrow><mo>-</mo><mi>j</mi><mn>2</mn><mi>π</mi><msubsup><mi>f</mi><mi>d</mi><mn>2</mn></msubsup><mo>·</mo><msub><mi>t</mi><mrow><mi>chip</mi><mn>1</mn></mrow></msub></mrow></msup><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><msub><mi>cof</mi><mrow><mi>fadmg</mi><mo>_</mo><mi>L</mi></mrow></msub><mo>·</mo><msup><mi>e</mi><mrow><mo>-</mo><mi>j</mi><mn>2</mn><mi>π</mi><msubsup><mi>f</mi><mi>d</mi><mi>L</mi></msubsup><mo>·</mo><msub><mi>t</mi><mrow><mi>chip</mi><mn>1</mn></mrow></msub></mrow></msup><mo>]</mo></mtd></mtr><mtr><mtd><mo>.</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>.</mo></mtd></mtr><mtr><mtd><mi>h</mi><mrow><mo>(</mo><mi>chip</mi><mo>_</mo><mn>864</mn><mo>)</mo></mrow><mo>=</mo><mo>[</mo><msub><mi>cof</mi><mrow><mi>fadmg</mi><mo>_</mo><mn>1</mn></mrow></msub><mo>·</mo><msup><mi>e</mi><mrow><mo>-</mo><mi>j</mi><mn>2</mn><mi>π</mi><msubsup><mi>f</mi><mi>d</mi><mn>1</mn></msubsup><mo>·</mo><msub><mi>t</mi><mrow><mi>chip</mi><mn>864</mn></mrow></msub></mrow></msup><mo>,</mo><msub><mi>cof</mi><mrow><mi>fadmg</mi><mo>_</mo><mn>2</mn></mrow></msub><mo>·</mo><msup><mi>e</mi><mrow><mo>-</mo><mi>j</mi><mn>2</mn><mi>π</mi><msubsup><mi>f</mi><mi>d</mi><mn>2</mn></msubsup><mo>·</mo><msub><mi>t</mi><mrow><mi>chip</mi><mn>864</mn></mrow></msub></mrow></msup><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><msub><mi>cof</mi><mrow><mi>fadmg</mi><mo>_</mo><mi>L</mi></mrow></msub><mo>·</mo><msup><mi>e</mi><mrow><mo>-</mo><mi>j</mi><mn>2</mn><mi>π</mi><msubsup><mi>f</mi><mi>d</mi><mi>L</mi></msubsup><mo>·</mo><msub><mi>t</mi><mrow><mi>chip</mi><mn>864</mn></mrow></msub></mrow></msup><mo>]</mo></mtd></mtr></mtable></mfenced><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mtext>1</mtext><mo>)</mo></mrow></mrow></math>

wherein h is(chip _ i) represents the channel impulse response on the ith chip (i ranges from 1 to 864), cof_{fading_k}Represents the fading coefficient of the k-th path,

indicating a shift in frequency by Doppler on the k-th path

The resulting phase offset on the ith chip.

The phase shift amount per chip of the received signal caused by the doppler shift will be referred to as the doppler shift value hereinafter.

Before the following steps are executed, the following settings are required to be made for the terminal: the terminal stores n preset range intervals (n is more than or equal to 1) of absolute values of the Doppler frequency offset values, and the n range intervals cover all possible absolute values of the Doppler frequency offset values section by section; in addition, a comparison slot [ i ] (i takes a value of 1 to n) corresponding to each range interval is stored thereon, that is, the slot [ i ] corresponds to the ith range interval, and the corresponding relationship may be: as the range interval gradually approaches to infinity, the comparison time slot corresponding to each range interval gradually approaches to the current time slot, that is, it can be understood that the larger the absolute value of the doppler frequency offset value falling within a certain range interval is, the closer the corresponding comparison time slot is to the current time slot. The comparison time slot is a time slot before the current time slot, and may be a time slot in which the time slot number in the subframe where the current time slot is located is smaller than the current time slot, or may be a time slot in a subframe before the subframe where the current time slot is located.

Preferably, n is set to 2, that is, the range intervals of the absolute values of the doppler frequency offset values are respectively set as: the value of the interval a ═ 0, X), the interval B ═ X, + ∞, X can be determined by engineering practice. The comparison time slot corresponding to the interval A is the time slot with the same time slot number as the current time slot in the previous sub-frame; the comparison time slot corresponding to the interval B is the TS0 time slot in the subframe where the current time slot is located. The two time slots are used as comparison time slots because the training sequences are always in the training code segments of the two time slots, so that the channel estimation can be carried out. Of course, the comparing time slot may also be any other time slot before the current time slot, such as the time slot before the current time slot, but it is necessary to ensure that there is a training sequence in the training code segment of the comparing time slot.

The invention eliminates the Doppler shift factor by several steps

The influence on the correct reception and demodulation of the symbols by the mobile terminal includes:

(1) pre-judging which range interval the absolute value of the Doppler frequency offset value on the current time slot is in;

(2) estimating the phase offset of the received signal caused by Doppler frequency shift in each chip length by comparing the phase difference of the current time slot and the comparison time slot corresponding to the range interval on the corresponding path;

(3) according to the phase shift amount, the phase shift of the received symbol caused by Doppler frequency shift is eliminated by correcting an A matrix in the joint detection by combining the traditional joint detection technology.

(4) When demodulating the received symbol, fine adjustment is performed on the previously estimated doppler frequency offset while demodulating.

After the steps (1) and (2) are executed, the Doppler frequency offset value of the mobile terminal can be preliminarily estimated; and the deviation caused by the doppler shift can be compensated by performing step (3) subsequently. And (4) after the step (4) is executed, the Doppler frequency offset value of the mobile terminal on the time slot is more accurately estimated.

In the step (1), the prejudged value of the doppler frequency offset is a doppler frequency offset value estimated by the mobile terminal in a last processing time slot. Preferably, the estimated doppler bias value is processed for the time slot having the same time slot number as the current time slot in the previous subframe.

In the step (2), comparing the phase difference between the current time slot and the time slot on the corresponding path means: firstly, obtaining channel estimation sequences of the two time slots; then, matching the two sequences according to paths; and after acquiring the phase difference on the corresponding path, dividing the phase difference by the corresponding chip interval to obtain the phase offset of the mobile terminal on each path and each chip length. In order to obtain a channel estimation sequence, several actions such as intercepting a training code, performing FFT transformation, IFFT inverse transformation, etc. need to be performed on a corresponding time slot, and an improved steiner method is adopted for implementing channel estimation, which belongs to the field of the prior art and is not described herein again.

In addition, after the channel estimation is performed on the current time slot, the estimation sequence can be saved. In this way, when the subsequent time slot is used as the comparison time slot, the terminal stores the channel estimation sequence of the time slot, so that the channel estimation process for the time slot can be omitted, and the channel estimation sequence is directly used as the channel estimation sequence of the comparison time slot.

As can be seen from step (3), the elimination of the doppler shift bias is achieved in conjunction with the prior art-joint detection of TD. However, in the present invention, the generated joint detection matrix a is modified, which specifically includes: the generated b vector of the A matrix is formed by convolution of a corrected channel estimation sequence and a composite spread spectrum code, and the corrected channel estimation sequence is formed by modulating an upper Doppler frequency offset value through an estimated average channel estimation sequence, so that the corrected channel estimation sequence is not fixed in a time slot any more.

Preferably, in a typical TD mobile terminal high-speed motion scene, only the phase offset estimation and the channel estimation of the strongest path of the mobile terminal need to be performed, and the combined detection a matrix needs to be corrected, so as to eliminate the doppler shift deviation.

In the step (4), in order to improve the accuracy of estimation, the doppler frequency offset estimated before is finely adjusted during demodulation, the doppler frequency offset value of the strongest path after fine adjustment or the maximum value of the doppler frequency offset values of all paths is used as the doppler frequency offset value of the time slot, and the value is stored in the mobile terminal as the doppler frequency offset prejudgment value of the next processing time slot.

In summary, as shown in fig. 3, the mobile terminal repeatedly executes the following steps for each received timeslot data containing the data of the mobile terminal:

step 301: the mobile terminal takes the stored Doppler frequency offset estimation value of the last processing time slot as a pre-judgment value (for the first receiving time slot, the pre-judgment value is 0 or a pre-specified value), and determines the range interval in which the absolute value of the pre-judgment value falls;

step 302: selecting a corresponding comparison time slot according to the range interval, intercepting a training sequence of the comparison time slot, and performing first channel estimation by using an improved Steiner method to obtain the average channel impact response of each mobile terminal in the time period of the comparison time slot;

step 303: intercepting the training sequence of the time slot, performing second channel estimation, and estimating the average channel impulse response of each mobile terminal in the time slot period;

of course, the order of channel estimation for the time slot and the comparison time slot is not sequential, so long as the average channel impulse response in the two time slots is obtained.

Step 304: matching the channel estimation values of the current time slot and the comparison time slot according to the mobile terminal and the corresponding path, carrying out phase subtraction on the two matched channel sequences according to the path, and then dividing the phase subtraction by the chip interval length between the current time slot and the comparison time slot to estimate the phase offset of the received signal of each mobile terminal on each path and each chip caused by Doppler frequency shift, namely the Doppler frequency offset value of each mobile terminal on each path;

step 305: multiplying the phase offset of each mobile terminal on each path and each chip obtained in the step 304 by a spreading factor SF to obtain a phase offset factor of each mobile terminal on each path and each symbol length duration, correcting a channel estimation sequence of the current time slot by using the factor, performing convolution operation by using the corrected estimated channel sequence and a composite spreading code, and constructing a corrected joint detection A matrix;

step 306: and performing joint detection on the received signal of the current time slot by using the corrected A matrix, and acquiring the demodulation symbols of each mobile terminal at one time.

Step 307: comparing the phase of the demodulated symbols with the closest standard modulation symbols (the standard modulation symbols are the standard complex modulation symbols sent after modulation by the transmitter, for example for QPSK modulation, the standard modulation symbols are + i, 1, -1, -i), smoothing the obtained phase difference, and obtaining the phase offset; then, the phase offset is used to fine-tune the doppler frequency offset value of the strongest channel path of the mobile terminal in the time slot estimated in step 304;

step 308: and determining the Doppler frequency offset value of the time slot and storing the Doppler frequency offset value to the next receiving time slot of the mobile terminal for reference comparison. The doppler frequency offset value of the time slot is the doppler frequency offset value on the strongest path of the mobile terminal after being fine-tuned in step 307, or the maximum doppler frequency offset value in each path of the mobile terminal, or the weighted sum of the two.

When the mobile terminal is in a high-speed motion scene, in the above steps 305 to 308, only the strongest path of the mobile terminal may be subjected to the estimation of the phase offset and the correction of the channel estimation of the strongest path, and the joint detection a matrix may be corrected to eliminate the doppler shift deviation.

The invention is further illustrated below by means of a specific example.

Assuming that a TD mobile terminal is engaged in 12.2kbps voice service, the service is characterized in that the mobile terminal occupies a fixed time slot in each subframe, and the downlink spreading factor SF is 16.

Carrying out initial configuration on the terminal: presetting 2-segment DupupuRange interval of absolute value of doppler frequency offset: interval a ═ 0, 1 ═ 10^-3Radian/chip), interval B ═ 1 x 10^-3Radian/chip, + ∞). (the boundary between the two selected intervals is equal to 1 x 10^-3Radian/chip, corresponding mobile terminal speed of 120km/h, maximum Doppler shift f_{d max}Is 222 Hz).

Setting comparison time slots of two corresponding range intervals: corresponding to the interval A, the comparison time slot is the time slot with the same time slot number as the current time slot in the previous sub-frame; corresponding to the interval B, the comparison time slot is the TS0 time slot of the present subframe.

The predetermined value of the initial doppler frequency offset value is set to 0.

The mobile terminal repeats the following steps for the received time slot containing the data of the mobile terminal in each sub-frame:

step 1: selecting a comparison time slot according to which range interval the absolute value of the Doppler frequency offset value of the time slot is in (for the first sub-frame, the predetermined value is 0, for the following sub-frames, the predetermined value is the Doppler frequency offset value of the time slot with the same time slot number as the current time slot in the previous sub-frame)

When the pre-judgment value falls into the interval A, the mobile terminal is considered to be in a small Doppler frequency offset state, and a time slot with the same time slot number as the current time slot in the previous subframe is selected as a comparison time slot;

when the pre-determined value falls into the interval B, the mobile terminal is considered to be in a large Doppler frequency offset state, so that more accurate Doppler frequency offset estimation needs to be carried out, and the TS0 time slot in the subframe is selected as a comparison time slot.

Step 2: and intercepting the training sequence of the comparison time slot to carry out channel estimation, and acquiring the average channel response ch _ est1 of each mobile terminal on the comparison time slot. The specific method of channel estimation adopts the improved Steiner method, and can be divided into the following two substeps:

step 2.1, intercepting training sequence receiver _ mid of comparison time slot₁The FFT conversion is carried out, and the FFT conversion is carried out,then dividing the result by FFT transformation of the basic training sequence mid _ basic, and performing IFFT transformation on the result to obtain a Channel Estimation sequence (Channel _ Estimation):

Channel_Estimation＝IFFT(FFT(receiver_mid₁)/FFT(mid_basic)) (2)

wherein the basic training sequence is initially assigned to each cell by the system and signaled to the mobile terminal.

Step 2.2, as shown in fig. 4, the path with power greater than epsilon in the channel estimation sequence corresponding to each mobile terminal obtained in step 2.1 is taken as a noise path to be removed, and the remaining sequence ch _ est1 is the average channel response of each mobile terminal in the comparison time slot. (wherein ∈ r ═ r²·σ²，r²Representing the threshold signal-to-noise ratio, σ²Representing the noise power, the values of r and sigma can be selected according to the actual engineering, and r is less than 1).

In concrete implementation, when the comparison time slot is the time slot with the same time slot number as the current time slot in the previous subframe, because the channel estimation is already carried out on the time slot in the processing of the previous subframe, only the storage of the channel estimation needs to be added in the mobile terminal, and the storage is read out in the step, and the steps 2.1 and 2.2 do not need to be carried out;

and step 3: and intercepting the training sequence of the current time slot for channel estimation to obtain the average channel response ch _ est2 of each mobile terminal on the current time slot. The specific method of channel estimation adopts the modified steiner method, which can be divided into two sub-steps (as described in steps 2.1 and 2.2), and will not be described herein again.

After the channel estimation sequence on the current time slot is obtained, the sequence is stored.

Step 4, respectively matching the estimation sequences of the average channel responses of the current time slot and the comparison time slot according to the mobile terminal and the path, carrying out phase subtraction on the average channel response of the strongest path of each mobile terminal, and then dividing the phase subtraction by the number of interval chips between the current time slot and the comparison time slot to obtain the phase offset of each chip of the received signal of the mobile terminal on the strongest path caused by Doppler frequency shift, namely the Doppler frequency offset value of the mobile terminal on the strongest path:

{phase_estimation}_{per  chip}^{strongest  tap} = \frac{{phase}_{strongest  tap}^{ch_est 2} - {phase}_{strongest  tap}^{ch_est 1}}{N_{chip}} - - - (3)

this value can be considered as a coarse estimate of the doppler frequency offset (which also needs to be fine-tuned in step 6). Wherein phase-represents the phase-taking operation, and string tap represents the strongest path, i.e. the path with the maximum power in the channel response of each path, N_chipNumber of chips representing the separation between the current slot and the comparison slot:

for TS0 time slot selected from the sub-frame in step 1 as comparison time slot, N_chipThe current timeslot number 864+352, where 352 is the total chip length of the dl synchronization code, the guard interval, and the ul synchronization code.

For the time slot with the same time slot number in the last sub-frame selected in the step 1 as the comparison time slot, the interval between the current time slot and the comparison time slot is one sub-frame length, and the chip interval is N_chip＝6400。

In this embodiment, it is preferable to estimate only the doppler frequency offset value of the strongest path of each terminal user, and the complexity of terminal implementation can be reduced by adopting a preferable method; the reason for this is that in a typical mobile terminal high-speed motion scene, there is generally a wireless signal with a direct path (i.e., the radio wave can reach the mobile terminal directly from the base station antenna), and the direct path is a main factor affecting the quality of the received signal, i.e., the estimated strongest path;

and 5: and multiplying the estimated phase offset of each chip of the received signal on the strongest path by a spreading factor, correcting the channel estimation sequence of the current time slot, and performing convolution operation on the corrected channel estimation sequence and the composite spreading code to form a corrected joint detection A matrix. The method is specifically divided into four substeps:

step 5.1, multiplying the phase offset of each chip of the received signal obtained in step 4 on the strongest path by the spreading factor SF to obtain the phase offset factor of the received signal on each symbol of the strongest path:

{phase_estimation}_{per symbol}^{strongest tap} = SF * phase_{estimation}_{per chip}^{strongest tap} - - - (4)

step 5.2, compensating the phase offset of the received signal on each symbol length of the strongest path to the strongest path of the current time slot average channel estimation sequence obtained in step 3, and obtaining a corrected channel estimation sequence:

<math><mrow><mi>channel</mi><mo>_</mo><msubsup><mi>estimation</mi><mrow><mi>symbol</mi><mo>_</mo><mi>n</mi></mrow><mrow><mo>′</mo><mi>tap</mi><mo>_</mo><mi>strongest</mi></mrow></msubsup><mo>=</mo><mi>channel</mi><mo>_</mo><msup><mi>estimation</mi><mi>strongest tap</mi></msup><mo>*</mo><msup><mi>e</mi><mrow><mo>-</mo><mi>j</mi><mo>*</mo><mn>2</mn><mi>π</mi><mo>*</mo><mi>n</mi><mo>*</mo><mi>phase</mi><mo>_</mo><msubsup><mi>estimation</mi><mi>per symbol</mi><mi>strongest tap</mi></msubsup></mrow></msup><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>5</mn><mo>)</mo></mrow></mrow></math>

wherein, channel _ animation^{strongest tap}Indicating the strongest path of the mobile terminal in the current slot average channel estimate obtained in step 3, and n indicating the nth modulation symbol.

The corrected channel estimation sequence is not fixed in the time slot any more, but modulated by the estimated Doppler frequency offset in a symbol-by-symbol manner.

Step 5.3, multiplying the spread spectrum code, the channel code and the scrambling code of each mobile terminal to respectively obtain the composite spread spectrum code of each mobile terminal

c_{i}^{user} = spreading_{code}_{i}^{user} * channeliasation_{code}_{i} * scrambling_{code}_{i} - - - (6)

And 5.4, constructing a joint detection matrix A by using the corrected channel estimation sequence obtained in the step 5.2 and the composite spread spectrum codes of each mobile terminal generated in the step 5.3.

The structure of the a matrix is shown in fig. 5. It is composed of N V blocks, N represents the number of modulation symbols, each V block is arranged by U b vectors, U represents the number of mobile terminals, each b vector is obtained by the convolution of the corrected channel estimation and the composite spread spectrum code:

<math><mrow><msubsup><mover><mi>b</mi><mo>&RightArrow;</mo></mover><mrow><mi>symbol</mi><mo>_</mo><mi>n</mi></mrow><mrow><mi>user</mi><mo>_</mo><mi>u</mi></mrow></msubsup><mo>=</mo><mi>Channel</mi><mo>_</mo><msubsup><msup><mi>Estimation</mi><mo>′</mo></msup><mrow><mi>symbol</mi><mo>_</mo><mi>n</mi></mrow><mrow><mi>user</mi><mo>_</mo><mi>u</mi></mrow></msubsup><mo>&CircleTimes;</mo><msup><mi>c</mi><mrow><mi>user</mi><mo>_</mo><mi>u</mi></mrow></msup><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>7</mn><mo>)</mo></mrow></mrow></math>

wherein,indicating the revised channel estimate at the nth symbol for the u-th mobile terminal, c^user_uIndicating the composite spread spectrum code of the u mobile terminal;

representing a convolution operation.

Step 6: using the corrected A matrix to make joint detection for the data segment of the current time slot to obtain the demodulation symbol of each mobile terminal

d＝(I+σ^-2A^HA)^-1·A^He (8)

I.e. doppler shift cancellation of the demodulated symbols is achieved. Wherein, I represents an identity matrix, sigma represents noise power, and e represents a received data segment chip; a. the^HHermite transformation of a representation matrix A, A^-1Representing the inverse of matrix a.

And 7: in order to obtain a more accurate doppler frequency offset estimation value, the coarse estimation value of the doppler frequency offset estimated in step 4 is finely adjusted, which specifically includes the following steps:

A. comparing the phase of the demodulation symbol obtained in the step 6 with the phase of the standard modulation symbol, smoothing the obtained phase difference, and finely adjusting the coarse estimation value of the Doppler frequency offset according to the obtained result, which specifically comprises the following steps:

a1, obtaining the modulation symbol d of the demodulation symbol identical to the standard_stdThe phases of (a) are compared, and the obtained phase difference is averaged:

Δphase′＝AVG(phase(d)-phase(d_std)) (9)

here, AVG represents an averaging operation, and specifically means averaging all (all users) of demodulated symbols obtained by joint detection.

A2, fine-tuning the rough estimation value of the Doppler frequency offset estimated in the step 4 by using the obtained delta phase':

<math><mrow><mi>Δphase</mi><mo>_</mo><mi>e</mi><msubsup><mi>stimation</mi><mi>per chip</mi><mrow><mi>tap</mi><mo>_</mo><mi>strongest</mi></mrow></msubsup><mo>=</mo><mi>phase</mi><mo>_</mo><msubsup><mi>estimation</mi><mi>per chip</mi><mrow><mi>tap</mi><mo>_</mo><mi>strongest</mi></mrow></msubsup><mo>+</mo><mfrac><msup><mi>Δphase</mi><mo>′</mo></msup><mi>SF</mi></mfrac><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>10</mn><mo>)</mo></mrow></mrow></math>

the value can be considered as

That is, the Doppler frequency offset value of the mobile terminal in the time slot can be obtained according to the value_d。

B. The phase shift amount of the adjusted strongest path obtained in step A2

The Doppler frequency offset value of the time slot is stored in the mobile terminal as the prejudgment quantity when the next sub-frame is processed.

Repeating steps 1-7 for the next subframe data.

As shown in fig. 6, it can be seen that the bit error rate of the mobile terminal is significantly reduced by using the method of the present invention.

The present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof, and it should be understood that various changes and modifications can be effected therein by one skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A method for estimating Doppler frequency offset in TD-SCDMA system of time division synchronous code division multiple access (TD-SCDMA), which is characterized in that,

2. The method of claim 1,

3. The method of claim 2,

4. The method of claim 2,

5. The method of claim 2 or 4,

6. The method of claim 5,

7. A method for compensating Doppler shift in TD-SCDMA system with time division synchronous code division multiple access (TD-SCDMA),

8. The method of claim 7,

9. The method of claim 8,

10. The method of claim 8,

11. The method of claim 8,

12. The method according to any one of claims 7 to 11,

13. The method of claim 12, further comprising: