CN102291340B - Channel estimation methods and device in a kind of ofdm system - Google Patents

Abstract

The invention discloses channel estimation methods in a kind of OFDM (OFDM) system, comprise: the pilot signal received carried out eliminating the female code of pilot tone successively, afterbody is connected, inverse discrete Fourier transform (DFT) (IDFT), be separated the process of multi-user, discrete Fourier transform (DFT) (DFT) and frequency domain truncation, obtain the channel estimation Signal of each user.The present invention discloses channel estimating apparatus in a kind of ofdm system, adopt method of the present invention and device, so, effectively can isolate the channel estimation Signal of multiple user, and can channel estimation quality be improved.

Description

Channel estimation method and device in orthogonal frequency division multiplexing system

Technical Field

The present invention relates to Orthogonal Frequency Division Multiplexing (OFDM) technologies, and in particular, to a method and an apparatus for channel estimation in an OFDM system.

Background

In an Orthogonal Frequency Division Multiplexing (OFDM) system, the available frequency band is divided into a number of small frequency bands, also referred to as subcarriers, onto which data symbols are modulated for transmission. In order to enable the receiver to demodulate data accurately, known pilot signals must be placed on some subcarriers, and the receiver processes the pilot signals accordingly to obtain channel estimates at the pilot positions and obtain channel estimates at other positions by means of interpolation or the like.

After the user receives the pilot signal, the general processing procedure is as follows: and firstly transforming to a time domain for noise reduction treatment, and then transforming to a frequency domain to finally obtain channel estimation at the pilot signal. In the process, since the received pilot signal is a segment of signal intercepted from a continuous channel response, and both ends of the frequency domain are discontinuous, the Gibbs phenomenon may occur when performing Discrete Fourier Transform (DFT)/Inverse Discrete Fourier Transform (IDFT) processing, thereby deteriorating the effect of channel estimation. The Gibbs phenomenon means that jitter ripples appear at two ends of the finally obtained frequency domain channel estimation.

Chinese patent application No. 200780039909.6 discloses a channel estimator, which adds spread signals at the front and rear of a pilot signal, respectively, and then performs IDFT and DFT processing, thereby suppressing the Gibbs phenomenon to a certain extent and improving the channel estimation effect.

However, the above patent is only suitable for processing the pilot signal transmitted by a single user, and when a plurality of users transmit the pilot signal simultaneously, the method disclosed in the above patent cannot be used for processing the channel estimation of the plurality of users.

Disclosure of Invention

In view of the above, the main objective of the present invention is to provide a method and an apparatus for channel estimation in an OFDM system, which can improve the channel estimation quality and is suitable for multi-user channel estimation.

In order to achieve the purpose, the technical scheme of the invention is realized as follows:

the invention provides a channel estimation method in an OFDM system, which comprises the following steps:

and sequentially carrying out pilot frequency mother code elimination, tail connection, IDFT, multi-user separation, DFT and frequency domain truncation on the received pilot frequency signal to obtain a channel estimation signal of each user.

In the above scheme, before performing DFT processing, the method further includes:

and carrying out noise elimination processing on the time domain signals obtained after the multi-user separation processing to obtain signals required by the DFT processing.

In the foregoing aspect, before receiving the pilot signal, the method further includes:

the OFDM system restrains the pilot signal transmitted by each user terminal;

the constraining specifically includes:

and multiplying the same pilot mother code by different phase rotations respectively to obtain the pilot signal of each user terminal.

In the above scheme, the phase rotation specifically includes:

if K user terminals transmit pilot signals simultaneously, the phase rotations of the K user terminals are respectively

In the above scheme, the performing tail-joining processing on the received pilot signal specifically includes:

and constructing K-section curve signals with continuity and differentiability by adopting the frequency domain signals with the pilot frequency mother codes eliminated, and adding the K-section curve signals to the tail part of the frequency domain signals with the pilot frequency mother codes eliminated, so that the two ends of the frequency domain of the mixed channel response signals are kept continuous, and the K value is the number of user terminals simultaneously transmitting the pilot frequency signals.

In the above scheme, the constructing and adding the K-segment curve signal to the tail of the frequency domain signal specifically includes:

respectively taking out signals from the head position and the tail position of the frequency domain signal processed by eliminating the pilot mother code, constructing a section of continuous and differentiable curve, and respectively constructing K sections of continuously variable curves by K users;

the signals of the K section curves are placed in a staggered mode and combined into the required tail connection signal;

and connecting the tail connecting signal behind the frequency domain signal with the pilot mother code eliminated to obtain a tail connecting processed signal.

In the foregoing scheme, the performing processing for separating multiple users specifically includes:

let the position coordinate in the time domain signal g (n) obtained after the IDFT processing by the kth user be:

$p_k=k\frac{M+\mathrm{KL}}{K},k=\mathrm{0,1},...,K-1;$

when k is not equal to 0, constructing an equivalent time domain shock response signal g of each user_kThe method of (n) is:

slave signal g (p)_k) Front face of (2) taking out N_FA signal, taking out N from the rear_BAnd zero-adding signals of other positions, and performing time delay processing to obtain:

g_k(n)＝[g(p_k)，g(p_k+1)，...，g(p_k+N_B)，0，0，...，0，g(p_k-N_F)，...，g(p_k-1)]；

when k is 0, construct g₀The method of (n) is:

g₀(n)＝[g(0)，g(1)，...，g(N_B)，0，0，...0，g(M+KL-N_F)，...，g(M+KL-1)]。

in the foregoing scheme, the processing for eliminating noise specifically includes:

setting a noise threshold;

and sequentially comparing the power of the non-zero signal in the signal after the multi-user processing is separated with the noise threshold, if the power is greater than the noise threshold, the signal is reserved, otherwise, the signal is set to be zero.

The invention also provides a channel estimation device in the OFDM system, which comprises: the system comprises a pilot frequency processing unit, a tail connection unit, an IDFT unit, a multi-user separation unit, a DFT unit and a frequency domain tail cutting unit; wherein,

the pilot frequency processing unit is used for eliminating pilot frequency mother codes of the received pilot frequency signals and sending the processed signals to the tail connecting unit;

the tail connecting unit is used for carrying out tail connecting processing on the received signal after receiving the signal sent by the pilot frequency processing unit and sending the processed signal to the IDFT unit;

the IDFT unit is used for carrying out IDFT processing on the received signals after receiving the signals sent by the tail connection unit and sending the processed signals to the multi-user separation unit;

the multi-user separation unit is used for separating multi-users from the received signal after receiving the signal sent by the IDFT unit and sending the processed signal to the DFT unit;

the DFT unit is used for performing DFT processing on the received signals after receiving the signals sent by the multi-user separation unit and sending the processed signals to the frequency domain truncation unit;

and the frequency domain truncation unit is used for performing frequency domain truncation processing on the received signal after receiving the signal sent by the DFT unit to obtain the channel estimation signal of each user terminal.

In the above scheme, the apparatus further comprises: the noise elimination unit is used for eliminating noise of the received signals after receiving the signals sent by the multi-user separation unit and sending the processed signals to the DFT unit;

the multi-user separation unit is used for separating multi-user from the received signal after receiving the signal sent by the IDFT unit and sending the processed signal to the noise elimination unit;

and the DFT unit is used for performing DFT processing on the received signal after receiving the signal sent by the noise separation unit and sending the processed signal to the frequency domain truncation unit.

In the above scheme, the apparatus further comprises: and the configuration unit is used for multiplying different phase rotations by the same pilot mother code respectively to obtain pilot signals of a plurality of user terminals.

In the above scheme, the tail connecting unit is specifically configured to:

and constructing K-section curve signals with continuity and differentiability by adopting the frequency domain signals with the pilot frequency mother codes eliminated, and adding the K-section curve signals to the tail part of the frequency domain signals with the pilot frequency mother codes eliminated, so that the two ends of the frequency domain of the mixed channel response signals are kept continuous, and the K value is the number of user terminals simultaneously transmitting the pilot frequency signals.

In the foregoing solution, the noise cancellation unit further includes:

the statistical module is used for counting the average power of the residual signals of the signals processed by the IDFT unit after being processed by the multi-user separation unit and sending the obtained average power to the setting module;

the setting module is used for setting a noise threshold value after receiving the average power sent by the counting module;

and the processing module is used for comparing the power of a non-zero signal in the received signal with the noise threshold, if the power is greater than the noise threshold, the signal is reserved, otherwise, the signal is set to be zero, and the processed signal is sent to the DFT unit.

The channel estimation method and the device in the OFDM system sequentially carry out the processing of eliminating pilot mother codes, tail connection, IDFT, multi-user separation, DFT and frequency domain truncation on received pilot signals to obtain the channel estimation signal of each user, thus effectively separating the channel estimation signals of a plurality of users, inhibiting the Gibbs phenomenon and improving the channel estimation quality;

in addition, the phase rotation of K user terminals is defined asThe pilot signal of the kth user is:thus, the separation effect of the time domain shock response signals of a plurality of user terminals can be optimized.

Drawings

FIG. 1 is a flow chart of a method for channel estimation in an OFDM system according to the present invention;

FIG. 2 is a flow chart of a method of channel estimation according to an embodiment;

FIG. 3 is a schematic diagram of an embodiment of a signal after pilot signal cancellation pilot mother code processing;

FIG. 4 is a schematic diagram of signals after tail-joining processing in one embodiment;

FIG. 5 is a diagram illustrating IDFT-processed signals in an embodiment;

FIG. 6 is a schematic time domain signal diagram of a first UE in an embodiment;

FIG. 7 is a plot of actual channel estimates versus real part of ideal channel estimates for a first user terminal;

fig. 8 is a schematic structural diagram of a channel estimation apparatus in an OFDM system according to the present invention.

Detailed Description

The basic idea of the invention is: and sequentially carrying out pilot frequency mother code elimination, tail connection, IDFT, multi-user separation, DFT and frequency domain truncation on the received pilot frequency signal to obtain a channel estimation signal of each user.

The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

The method for estimating channels in the OFDM system of the present invention, as shown in fig. 1, includes the following steps:

step 101: after receiving the pilot signal, processing to eliminate the pilot mother code to obtain a first signal;

here, it is assumed that a pilot signal occupies M subcarriers, and accordingly, the sequence length of a pilot mother code is M, and the pilot mother code is r (n), where n is 0.

Assuming that the received signal at the pilot position is y (n), the processing for removing the pilot mother code specifically includes: dividing the received signal at the pilot position by the pilot mother code r (n), and expressing the first signal as x (n) if there is any

x(n)＝y(n)/r(n)；

Wherein, the signal y (n) comprises the product of pilot signals transmitted by a plurality of user terminals and corresponding channel transfer functions, and noise signals; signal x (n) comprises a mixed channel response of a plurality of users, the mixed channel response containing a noise signal;

the OFDM system restricts the pilot signal transmitted by each user terminal, specifically:

using a same pilot mother code r (n) to respectively multiply different phase rotations to obtain a pilot signal of each user terminal; specifically, assuming that there are K user terminals transmitting pilot signals simultaneously, the phase rotations of the K user terminals are defined asThe pilot signal of the kth user is:

<math> <mrow> <msub> <mi>r</mi> <mi>k</mi> </msub> <mrow> <mo>(</mo> <mi>n</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>r</mi> <mrow> <mo>(</mo> <mi>n</mi> <mo>)</mo> </mrow> <mo>&CenterDot;</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mi>j</mi> <mn>2</mn> <mi>&pi;</mi> <mfrac> <mi>k</mi> <mi>K</mi> </mfrac> <mi>n</mi> <mo>)</mo> </mrow> <mo>,</mo> <mi>n</mi> <mo>=</mo> <mn>0</mn> <mo>,</mo> <mo>.</mo> <mo>.</mo> <mo>.</mo> <mo>,</mo> <mi>M</mi> <mo>-</mo> <mn>1</mn> <mo>,</mo> <mi>k</mi> <mo>=</mo> <mn>0</mn> <mo>,</mo> <mo>.</mo> <mo>.</mo> <mo>.</mo> <mo>,</mo> <mi>K</mi> <mo>-</mo> <mn>1</mn> <mo>;</mo> </mrow> </math>

the OFDM system restricts the pilot signal transmitted by each user terminal and adopts the pilot mother code which is the same as the pilot mother code adopted by the receiving terminal to eliminate the pilot mother code, and informs the receiving terminal of the pilot mother code after restricting the pilot signal transmitted by each user terminal;

it should be noted that: other phase rotation schemes specified in the protocol of the OFDM system may be used, but only with the above method, the separation of the time domain impulse response signals of multiple user terminals can be optimized. The K user terminals are not limited to the number of terminals in the physical sense. Such as: a physical terminal simultaneously transmits pilot signals by using 2 antennas, and the physical terminal is regarded as 2 equivalent user terminals; such situations may arise, among others, in future 4G standards.

Step 102: carrying out tail connection processing on the first signal to obtain a second signal;

here, since the received pilot signal is a segment of signal intercepted from a continuous channel response, two ends of the frequency domain are discontinuous, and performing tail-joining processing on the first signal means: constructing K-segment curve signals with continuity and differentiability according to the frequency domain signals with the pilot frequency mother codes eliminated, and adding the K-segment curve signals to the tail of the first signal, so that the head and tail signals of the obtained second signal are continuous, namely: keeping two ends of the frequency domain of the mixed channel response signal continuous;

the tail joining processing is carried out on the first signal, and the tail joining processing specifically comprises the following steps:

respectively taking out signals from the head position and the tail position of the frequency domain signal processed by eliminating the pilot mother code, constructing a section of curve with continuity and differentiability, and respectively constructing K sections of continuously changing curves by K users;

the signals of the K section curves are placed in a staggered mode and combined into the required tail connection signal;

connecting the tail connecting signal behind the frequency domain signal with the pilot mother code eliminated to obtain a tail connecting processed signal;

specifically, the second signal is denoted as f (n), and the length is set to M + KL; the ligation signal is denoted as z (l) and is KL in length; wherein z (l) is represented by a curve z of K_k(l) The interleaving is formed, K is 0,1, and K-1, the length of each curve is L, and the size of the L value is set according to the implementable principle of the DFT/IDFT function, that is: the signal sequence with the length of M + KL is easy to realize by DFT/IDFT;

at the construction curve z_k(l) It is necessary to select some signals to construct at the head and tail positions of x (n), and it is required to: curve z_k(l) Continuity and micromability are provided, but continuity and micromability of the splicing signal z (l) are not guaranteed; in particular, a variety of design methods may be employed to construct the curve z_k(l) Only the constructed curve z needs to be guaranteed_k(l) The signal sequence itself is smooth, continuously varying, without any break points present, i.e.: constructed curve z_k(l) Continuity and micromability;

a simple design is to select the curve z_k(l) Constructed for a straight line using a head signal having position coordinates x (K) and a tail signal having position coordinates x (M-K + K), thereby producing a sequence which varies linearly in amplitude and phase; some combination of x (K) and x (M-K + K) may also be selected to construct z_k(l) Alternatively, z is constructed by selecting arbitrary head and tail signals_k(l)；

z_k(l) After construction is complete, use z_k(l) To construct a splicing signal z_k(l) In particular, the signal z (l) is derived from z_k(l) The staggered formation is as follows:

z(l)＝[z₀(1)，z₁(1)，...，z_K-1(1)，z₀(2)，z₁(2)，...，z_K-1(2)，...，z₀(L)，z₁(L)，...，z_K-1(L)]；

the signal x (n) and the linking signal z (l) form a second signal f (n), which has

f(n)＝[x(n}，z(l)]；

Step 103: performing IDFT processing on the second signal to obtain a third signal;

here, the purpose of performing IDFT processing on the second signal f (n) is to obtain a time domain signal g (n), that is: a third signal; the IDFT processing performed on the second signal may specifically be Inverse Fast Fourier Transform (IFFT) processing if the length of the second signal sequence is an integer power of 2; the IDFT processing of the second signal f (n) refers to the operation of the IDFT function on the second signal f (n);

the time domain signal g (n) is:

g(n)＝[g(0)，g(1)，....，g(M+KL-1)]；

the sequence length of the time domain signal g (n) is: m + KL.

Step 104: processing the third signal to separate multiple users to obtain a fourth signal;

here, the fourth signal includes an equivalent time domain impulse response signal g for each user_k(n) wherein g_k(n) is composed of samples extracted from the time domain signal g (n);

let the position coordinates of a certain user in g (n) be:

$p_k=k\frac{M+\mathrm{KL}}{K},k=\mathrm{0,1},...,K-1;$

when k ≠ 0, construct g_kThe method of (n) is:

slave signal g (p)_k) Front face of (2) taking out N_FA signal, taking out N from the rear_BZero-adding the signals of other positions and carrying out time delay processing to obtain the signal

g_k(n)＝[g(p_k)，g(p_k+1)，...，g(p_k+N_B)，0，0，...，0，g(p_k-N_F)，...，g(p_k-1)]；

Wherein, g_k(N) has a sequence length of M + KL and a sequence length of the head signal of N_B+1, sequence length of tail signal N_F(ii) a Wherein, according to the channel characteristic parameter, the length of the channel time domain shock response signal, N, is preset_FAnd N_BIs chosen such that g is constructed_k(n) the length of the non-zero signal is greater than a preset length;

when k is 0, g₀The construction method of (n) is as follows:

g₀(n)＝[g(0)，g(1)，...，g(N_B)，0，0，...0，g(M+KL-N_F)，...，g(M+KL-1)]。

step 105: carrying out noise elimination processing on the fourth signal to obtain a fifth signal;

here, before the fourth signal is subjected to the noise elimination process, the noise power estimation process needs to be performed on the time domain signal g (n), specifically, the samples taken in step 104 of the time domain signal g (n) are all set to zero, that is: comprising K (N)_F+N_B+1 zeros, making statistics of the average power of the remaining signals in g (n), and setting the average power to be P_n(ii) a Besides, the prior art methods for counting noise power are all applicable here;

the processing for eliminating noise on the fourth signal specifically includes:

setting a noise threshold;

specifically, the average power P_nA certain time ofThe value of the number is set as the noise threshold; in the practical application process, different multiples can be taken to carry out simulation experiments, and the multiple with the best fit between the channel estimation signal obtained by simulation and an ideal channel estimation signal is selected as the basis for setting the noise threshold;

comparing g in sequence_kN in the sequence of (N)_F+N_BThe power of +1 non-zero signals and the magnitude of the noise threshold, if the power is greater than the noise threshold, the signal is retained, otherwise, the signal is set to zero; for K g_k(n) performing the above-mentioned processing to obtain a fifth signal, respectively, and representing the fifth signal asThe fifth signal comprises a signal obtained after the equivalent time domain impulse responses of the K users are subjected to noise elimination;the length of the sequence is still M + KL;

in fact, step 105 may not be performed, and the purpose of this step is to obtain better channel estimation effect, and if this step is not performed, the obtained user signal estimation signal quality is poor.

Step 106: performing DFT processing on the fifth signal to obtain a sixth signal;

here, for the fifth signalThe purpose of the DFT processing is to obtain a frequency domain signal h_k(n) is: a sixth signal; if five signalsWhen the length of the sequence is an integer power of 2, the DFT processing performed on the fifth signal may specifically be Fast Fourier Transform (FFT), where the DFT processing performed on the fifth signal refers to performing DFT function operation on the fifth signal;

frequency domain signal h_kThe sequence length of (n) is M + KL.

Step 107: performing frequency domain truncation processing on the sixth signal to obtain a channel estimation signal of each user, and ending the current processing flow;

specifically, the frequency domain signal h_k(n) cutting off KL signals with the tail length, and only keeping M signals with the front length to obtain a channel estimation signal of a kth user;

for K h_kAnd (n) respectively carrying out the processing to obtain channel estimation signals of K users.

The above scheme is not only applicable to multi-user channel estimation, but also applicable to single-user channel estimation, and at this time, K in the above scheme is 1.

The scheme of the invention is further described below with reference to examples.

The application scenario of this embodiment is: sequence length of pilot mother code: when M is 60, there are four ues transmitting pilot signals, and the phase rotation of each of these four ues is:k＝0，1，2，3。

the method for estimating the channel of the present embodiment, as shown in fig. 2, includes the following steps:

step 201: dividing the received signal at the pilot frequency position by a pilot frequency mother code with the sequence length of 60 to obtain a signal x (n);

the signal x (n) obtained after this step is performed, as shown in fig. 3; the abscissa represents the position of each signal, and the ordinate represents the intensity of each signal.

Step 202: carrying out tail connection processing on the signals x (n) to obtain signals f (n);

here, the IDFT function is selected according to a rule that is convenient to implementSelecting L as 10, and constructing z_k(l) And (3) constructing tail splicing signals z (l) by interleaving signals of four straight lines, and adding the constructed tail splicing signals z (l) to the tail of the signals x (n) to obtain signals f (n), wherein the sequence length of the signals f (n) is 100 as shown in fig. 4, wherein the abscissa represents the position of each signal, and the ordinate represents the strength of each signal.

Step 203: carrying out IDFT processing on the signal f (n) to obtain a time domain signal g (n);

here, the time domain signal g (n) is shown in fig. 5, and the sequence length of the time domain signal g (n) is 100; wherein the abscissa represents the position of each signal, and the ordinate represents the intensity of each signal;

the time domain signal g (n) comprises equivalent time domain impulse response signals of four user terminals.

Step 204: processing the time domain signals g (n) to separate multiple users to obtain the equivalent time domain response g of each user_k(n)；

Here, the position coordinates of the four user terminals in g (n) are: 0. 25, 50, 75; selection of N_F＝9，N_BLet p be 5_kEqual to 25, 50 and 75, respectively, from the signal g (p)_k) The front of the system takes out 9 signals, the back takes out 5 signals, the signals at other positions are all set to be zero, and the signals are subjected to time delay processing, so that g can be obtained respectively₁(n)、g₂(n) and g₃(n) in combination with g₀(n) the construction method can obtain the equivalent time domain responses g of four user terminals_k(n)。

Step 205: equivalent time domain response g for each user terminal_k(n) processing to eliminate noise to obtain signal

In particular, according to the channel characteristic parameter and the average power P_nSetting a noise threshold; comparing g in sequence_k(n)The power of the 15 non-zero signals in the sequence of (1) and the magnitude of the noise threshold, if the power is greater than the noise threshold, the signal is retained, otherwise, the signal is set to zero; for four g_k(n) the above treatments were carried out to obtain:andhas a sequence length of 100, wherein the signalNamely: the time domain signal of the first ue is shown in fig. 6, wherein the abscissa represents the position of each signal, and the ordinate represents the strength of each signal;

before setting a noise threshold, performing noise power estimation on a time domain signal g (n) to obtain an average power;

specifically, in performing step 204, K (N) has been taken from the time domain signal g (N)_F+N_B+1), i.e. 4 × (9+1+5), is 60 samples, so there are 40 samples left, and the average power P of these 40 samples is calculated_n。

Step 206: to pairDFT processing is carried out to obtain a frequency domain signal h_k(n)；

Here, the frequency domain signal h_kThe sequence length of (n) is 100.

Step 207: for frequency domain signal h_k(n) carrying out frequency domain truncation processing to obtain a channel estimation signal of each user, and ending the current processing flow;

specifically, the frequency domain signal h is removed_k(n) reserving only the first 60 signals from the tail 40 signals to obtain the channel estimation signal of the user terminal;

for four h_k(n) respectively carrying out the processing to obtain channel estimation signals of four users;

fig. 7 shows a plot of the real part of the channel estimate over 60 subcarriers of the first user terminal versus an ideal plot of the real part of the channel estimate, where the abscissa represents the position of each signal and the ordinate represents the strength of each signal; curve 1 represents the ideal real part curve of channel estimation, and curve 2 represents the real part curve of channel estimation obtained by the method of the present invention, and it can be seen from the figure that the degree of coincidence between curve 2 and curve 1 is very large, which shows that the channel estimation of each user terminal can be well obtained by the method of the present invention.

In order to implement the above method, the present invention further provides a channel estimation apparatus in an OFDM system, where the apparatus is shown in fig. 8, and the apparatus includes: a pilot processing unit 81, a tail joining unit 82, an IDFT unit 83, a multi-user separation unit 84, a DFT unit 85, and a frequency domain truncation unit 86; wherein,

a pilot processing unit 81, configured to perform pilot mother code elimination processing on the received pilot signal, and send the processed signal to a tail joining unit 82;

a tail joining unit 82, configured to, after receiving the signal sent by the pilot processing unit 81, perform tail joining processing on the received signal, and send the processed signal to the IDFT unit 83;

an IDFT unit 83, configured to, after receiving the signal sent by the tail joining unit 82, perform IDFT processing on the received signal, and send the processed signal to the multiuser separating unit 84;

a multiuser separating unit 84, configured to, after receiving the signal sent by the IDFT unit 83, perform processing of separating multiusers on the received signal, and send the processed signal to the DFT unit 85;

a DFT unit 85, configured to, after receiving the signal sent by the multiuser separating unit 84, perform DFT processing on the received signal, and send the processed signal to a frequency domain truncation unit 86;

a frequency domain truncation unit 86, configured to perform frequency domain truncation processing on the received signal after receiving the signal sent by the DFT unit 85, to obtain a channel estimation signal of each user terminal.

Wherein the apparatus may further comprise:

a noise removing unit 87 configured to, after receiving the signal sent by the multiuser separating unit 84, perform noise removal processing on the received signal, and send the processed signal to the DFT unit 85;

the multiuser separating unit 84 is further configured to send the processed signal to a noise canceling unit 87;

the DFT unit 85 is configured to, after receiving the signal sent by the noise separation unit 87, perform DFT processing on the received signal, and send the processed signal to the frequency domain truncation unit 86.

The apparatus may further comprise: and the configuration unit is used for multiplying different phase rotations by the same pilot mother code respectively to obtain the pilot signal of each user terminal.

The noise cancellation unit 87 may further include:

the statistical module is used for counting the average power of the residual signals of the signals processed by the IDFT unit 83 and processed by the multi-user separation unit 84 and sending the obtained average power to the setting module;

the setting module is used for setting a noise threshold value after receiving the average power sent by the counting module;

and a processing module, configured to compare the power of a non-zero signal in the received signal with a noise threshold, if the power is greater than the noise threshold, the signal is retained, otherwise, the signal is set to zero, and the processed signal is sent to the DFT unit 85.

The tail connecting unit is specifically used for:

and constructing K-section curve signals with continuity and differentiability by adopting the frequency domain signals with the pilot frequency mother codes eliminated, and adding the K-section curve signals to the tail part of the frequency domain signals with the pilot frequency mother codes eliminated, so that the two ends of the frequency domain of the mixed channel response signals are kept continuous, and the K value is the number of user terminals simultaneously transmitting the pilot frequency signals.

The above description is only exemplary of the present invention and should not be taken as limiting the scope of the present invention, and any modifications, equivalents, improvements, etc. that are within the spirit and principle of the present invention should be included in the present invention.

Claims (10)

1. A method for channel estimation in an Orthogonal Frequency Division Multiplexing (OFDM) system, the method comprising:

sequentially carrying out pilot frequency mother code elimination, tail connection, Inverse Discrete Fourier Transform (IDFT), multi-user separation, Discrete Fourier Transform (DFT) and frequency domain tail truncation on the received pilot frequency signal to obtain a channel estimation signal of each user;

the tail joining processing is performed on the received pilot signals, specifically:

constructing K-segment curve signals with continuity and differentiability by adopting the frequency domain signals with the pilot frequency mother codes eliminated, and adding the K-segment curve signals to the tail of the frequency domain signals with the pilot frequency mother codes eliminated so that two ends of a frequency domain of the mixed channel response signals are kept continuous, wherein the K value is the number of user terminals simultaneously transmitting the pilot frequency signals;

the processing for separating multiple users specifically comprises:

let the position coordinate in the time domain signal g (n) obtained after the IDFT processing by the kth user be:

k-0, 1,. K-1; wherein, M is the sequence length of the pilot mother code;

when k is not equal to 0, constructing an equivalent time domain shock response signal g of each user_kThe method of (n) is:

slave signal g (p)_k) Front face of (2) taking out N_FA signal, taking out N from the rear_BAnd zero-adding signals of other positions, and performing time delay processing to obtain:

g_k(n)＝[g(p_k),g(p_k+1),...,g(p_k+N_B),0,0,...,0,g(p_k-N_F),...,g(p_k-1)]；

when k is 0, construct g₀The method of (n) is:

g₀(n)＝[g(0),g(1),...,g(N_B),0,0,...0,g(M+KL-N_F),...,g(M+KL-1)]；

the processing of frequency domain truncation specifically comprises:

will frequency domain signal h_kAnd (n) cutting off KL signals with the tail length, and only keeping M signals with the front length to obtain a channel estimation signal of the kth user.

2. The method of claim 1, wherein prior to performing the DFT processing, the method further comprises:

and carrying out noise elimination processing on the time domain signals obtained after the multi-user separation processing to obtain signals required by the DFT processing.

3. The method of claim 1 or 2, wherein prior to receiving the pilot signal, the method further comprises:

the OFDM system restrains the pilot signal transmitted by each user terminal;

the constraining specifically includes:

and multiplying the same pilot mother code by different phase rotations respectively to obtain the pilot signal of each user terminal.

4. The method according to claim 3, characterized in that the phase rotation is in particular:

if K user terminals transmit pilot signals simultaneously, the phase rotations of the K user terminals are respectivelyk＝0,...,K-1,n＝0,....,M-1。

5. The method according to claim 1, wherein the constructing and adding the K-segment curve signal to the tail of the frequency domain signal are specifically:

respectively taking out signals from the head position and the tail position of the frequency domain signal processed by eliminating the pilot mother code, constructing a section of continuous and differentiable curve, and respectively constructing K sections of continuously variable curves by K users;

the signals of the K section curves are placed in a staggered mode and combined into the required tail connection signal;

and connecting the tail connecting signal behind the frequency domain signal with the pilot mother code eliminated to obtain a tail connecting processed signal.

6. The method according to claim 2, wherein the processing for removing noise is specifically:

setting a noise threshold;

and sequentially comparing the power of the non-zero signal in the signal after the multi-user processing is separated with the noise threshold, if the power is greater than the noise threshold, the signal is reserved, otherwise, the signal is set to be zero.

7. A channel estimation apparatus in an OFDM system, the apparatus comprising: the system comprises a pilot frequency processing unit, a tail connection unit, an IDFT unit, a multi-user separation unit, a DFT unit and a frequency domain tail cutting unit; wherein,

the pilot frequency processing unit is used for eliminating pilot frequency mother codes of the received pilot frequency signals and sending the processed signals to the tail connecting unit;

the tail connecting unit is used for carrying out tail connecting processing on the received signal after receiving the signal sent by the pilot frequency processing unit and sending the processed signal to the IDFT unit;

the IDFT unit is used for carrying out IDFT processing on the received signals after receiving the signals sent by the tail connection unit and sending the processed signals to the multi-user separation unit;

the multi-user separation unit is used for separating multi-users from the received signal after receiving the signal sent by the IDFT unit and sending the processed signal to the DFT unit;

the DFT unit is used for performing DFT processing on the received signals after receiving the signals sent by the multi-user separation unit and sending the processed signals to the frequency domain truncation unit;

the frequency domain truncation unit is used for performing frequency domain truncation processing on the received signal after receiving the signal sent by the DFT unit to obtain a channel estimation signal of each user terminal;

the tail connecting unit is specifically used for:

constructing K-segment curve signals with continuity and differentiability by adopting the frequency domain signals with the pilot frequency mother codes eliminated, and adding the K-segment curve signals to the tail of the frequency domain signals with the pilot frequency mother codes eliminated so that two ends of a frequency domain of the mixed channel response signals are kept continuous, wherein the K value is the number of user terminals simultaneously transmitting the pilot frequency signals;

the processing for separating multiple users specifically comprises:

let the position coordinate in the time domain signal g (n) obtained after the IDFT processing by the kth user be:

k-0, 1,. K-1; wherein, M is the sequence length of the pilot mother code;

when k is not equal to 0, constructing an equivalent time domain shock response signal g of each user_kThe method of (n) is:

slave signal g (p)_k) Front face of (2) taking out N_FA signal, taking out N from the rear_BAnd zero-adding signals of other positions, and performing time delay processing to obtain:

g_k(n)＝[g(p_k),g(p_k+1),...,g(p_k+N_B),0,0,...,0,g(p_k-N_F),...,g(p_k-1)]；

when k is 0, construct g₀The method of (n) is:

g₀(n)＝[g(0),g(1),...,g(N_B),0,0,...0,g(M+KL-N_F),...,g(M+KL-1)]；

the processing of frequency domain truncation specifically comprises:

will frequency domain signal h_kAnd (n) cutting off KL signals with the tail length, and only keeping M signals with the front length to obtain a channel estimation signal of the kth user.

8. The apparatus of claim 7, further comprising: the noise elimination unit is used for eliminating noise of the received signals after receiving the signals sent by the multi-user separation unit and sending the processed signals to the DFT unit;

the multi-user separation unit is used for separating multi-user from the received signal after receiving the signal sent by the IDFT unit and sending the processed signal to the noise elimination unit;

and the DFT unit is used for performing DFT processing on the received signal after receiving the signal sent by the noise separation unit and sending the processed signal to the frequency domain truncation unit.

9. The apparatus of claim 7 or 8, further comprising: and the configuration unit is used for multiplying different phase rotations by the same pilot mother code respectively to obtain pilot signals of a plurality of user terminals.

10. The apparatus of claim 8, wherein the noise cancellation unit further comprises:

the statistical module is used for counting the average power of the residual signals of the signals processed by the IDFT unit after being processed by the multi-user separation unit and sending the obtained average power to the setting module;

the setting module is used for setting a noise threshold value after receiving the average power sent by the counting module;

and the processing module is used for comparing the power of a non-zero signal in the received signal with the noise threshold, if the power is greater than the noise threshold, the signal is reserved, otherwise, the signal is set to be zero, and the processed signal is sent to the DFT unit.