CN103685088A - Pilot frequency optimization method of sparse channel, device and channel estimation method - Google Patents

Abstract

The embodiment of the invention discloses a sparse channel pilot optimization method, and the embodiment of the invention also discloses a sparse channel pilot optimization device and a sparse channel estimation method. The method includes: randomly selecting elements from a subcarrier set to generate an initial pilot arrangement, generating a candidate set according to the subcarrier set and the initial pilot arrangement, and selecting the initial pilot from the candidate set The optimal element at each element position in the frequency arrangement to generate an optimal pilot arrangement, wherein the optimal pilot arrangement is formed by the best element at each element position in the initial pilot arrangement; multiple times The above steps are repeatedly executed, and the optimal pilot arrangement with the best first objective function value generated during the multiple repetitions is determined as the optimal pilot arrangement. The invention has the advantages of fast convergence speed, good performance of mean square error and bit error rate, and the like.

Description

Pilot optimization method, device and channel estimation method for sparse channel

technical field

The invention relates to the communication field, in particular to a sparse channel pilot optimization method, device and channel estimation method.

Background technique

Orthogonal Frequency Division Multiplexing (OFDM), as the core technology of the fourth-generation mobile communication and future wireless communication, can effectively combat multipath effects in wireless propagation, simplify equalizer design, reduce receiver complexity and power consumption and improve spectrum utilization. As one of the key links in OFDM system, channel estimation estimates the time delay, attenuation, multipath and other parameters of the channel experienced by the signal transmission. The accuracy of channel estimation has a direct impact on channel equalization, demodulation and channel decoding. Therefore, channel estimation technology has always attracted the attention of researchers.

Channel estimation can be divided into two categories: blind estimation and pilot-assisted channel estimation (Pilot Assisted Channel Estimation, PACE). Among them, the blind estimation utilizes some characteristics of the transmitted data itself to estimate the channel. Due to its high computational complexity and poor real-time performance, it is rarely used in actual communication systems. Before sending data, PACE pre-inserts some known symbols (pilots) at the receiving end, and the receiver calculates the channel based on least squares (Least Squares, LS) or minimum mean square error (Minimum Mean Square Error, MMSE) and other criteria. Make an estimate. The recently emerging Sparse Channel Estimation (Sparse Channel Estimation), or Compressed Channel Sensing (Compressed Channel Sensing), also belongs to the PACE category. It uses the sparseness of wireless channels and uses Compressed Sensing (CS) technology for channel estimation. . Compared with traditional LS or MMSE channel estimation, compressed sensing technology can greatly reduce pilot overhead, improve spectrum utilization and channel estimation accuracy. Considering the delay spread of the wireless channel and the high sampling rate of the receiver front-end, the channel multipath components are scattered in this delay spread, and the channel impulse response (Channel Impulse Response, CIR) sequence after sampling usually presents most of Zero and a few non-zero sparsity, especially for OFDM systems that generally use oversampling technology, this sparsity characteristic is more obvious.

In prior art 1, the long-term evolution (Long-Term Evolution, LTE) system mainly adopts a uniform pilot distribution method, and the positions of the pilots are evenly distributed in the frequency domain and the time domain. Many research literatures show that this kind of uniformly distributed pilot design method is not optimal for compressive sensing based channel estimation schemes. Therefore, it is necessary to carry out special pilot design for the scheme based on compressed sensing channel estimation.

In the second prior art, a compressed sensing-based discontinuous OFDM channel estimation method is disclosed. The method includes: designing a channel estimation pilot pattern; selecting the pilot pattern; and estimating the frequency domain response of the channel. The pilot pattern selection uses the following two schemes. Option 1: Retain the traditional uniform pilot pattern, disable the pilots at the subcarriers and disable them naturally, so that the available pilots show natural inhomogeneity; Option 2: Fix the number of pilots, based on the criterion of minimizing the cross-correlation of the recovery matrix , with fewer pilots to obtain better channel estimation performance and system bit error rate performance than other current methods. The method can obtain better channel estimation performance and system bit error rate performance than other current methods with fewer pilots in a variety of forbidden subcarrier scenarios. This method mainly considers how to design pilots in the case of discontinuous OFDM, and how to optimize the design of pilots when the optimized pilot position coincides with the forbidden position, and does not have any consideration for the continuous subcarriers. The optimal pilot design scheme is discussed.

In prior art 3 (C.Qi and L.Wu, "A Study of Deterministic Pilot Allocation for Sparse Channel Estimation in OFDM Systems," IEEE Comm.Lett., Vol.16, No.5, pp.742-744, May 2012.), a pilot optimization method based on OFDM was disclosed. This method can use the random approximation method to obtain the approximate optimal pilot distribution when the number of subcarriers is irregular, and can obtain better convergence speed and channel estimation gain. However, the convergence speed of this method is still insufficient.

In prior art 4, an exhaustive method is used to generate an optimized pilot arrangement, that is, a pilot arrangement is randomly generated each time, and an objective function is selected from all randomly generated pilot arrangements within the specified program running time The pilot arrangement with the smallest value. The convergence performance, mean square error (Mean Square Error, MSE) performance and bit error rate (Bit Error Rate, BER) performance of this method are all lacking.

Contents of the invention

The technical problem to be solved by the embodiments of the present invention is to provide a sparse channel pilot optimization method, device and channel estimation method. Special pilot design can be carried out for the channel estimation scheme based on compressed sensing technology, the pilot arrangement can be optimized, and the mean square error and bit error rate performance of OFDM sparse channel estimation can be improved.

In order to solve the above technical problems, an embodiment of the present invention provides a pilot optimization method for sparse channels, including:

Randomly select elements from the subcarrier set to generate an initial pilot arrangement, generate a candidate set according to the subcarrier set and the initial pilot arrangement, and select each element in the initial pilot arrangement from the candidate set The best element in position to generate a preferred pilot arrangement, wherein the preferred pilot arrangement is composed of the best elements in the position of each element in the initial pilot arrangement;

The above steps are repeated multiple times, and the optimal pilot arrangement with the best first objective function value generated during the repeated repetitions is determined as the optimal pilot arrangement.

Correspondingly, an embodiment of the present invention also provides a sparse channel pilot optimization device, which includes:

The initialization unit is used to randomly select elements from the subcarrier set to generate an initial pilot arrangement;

A preferred pilot arrangement generating unit, configured to generate a candidate set according to the subcarrier set and the initial pilot arrangement and select the best position of each element in the initial pilot arrangement from the candidate set elements to generate a preferred pilot arrangement, wherein the preferred pilot arrangement is formed by the best element at each element position in the initial pilot arrangement;

The optimized pilot arrangement generation unit is used to repeatedly call the initialization unit and the optimal pilot arrangement generation unit, and determine the optimal pilot arrangement with the best first objective function value generated during the multiple repetitions To optimize the pilot arrangement.

The embodiment of the present invention also provides a channel estimation method, the method comprising:

The transmitting end determines the optimized pilot arrangement according to the method of claim 1 to insert the optimized pilot;

The receiver performs channel estimation based on compressed sensing technology.

Implementing the embodiment of the present invention has the following beneficial effects:

1) Determine the best elements in the initial pilot arrangement according to the cross-correlation to generate the optimal pilot arrangement, and select the optimal pilot arrangement with the smallest cross-correlation among various optimal pilot arrangements, which can minimize Optimize the cross-correlation of the measurement matrix and have a faster convergence speed;

2) According to the optimized pilot arrangement and channel estimation based on compressed sensing technology, the mean square error and bit error rate performance of OFDM sparse channel estimation can be improved.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained according to these drawings without any creative effort.

Fig. 1 is a schematic flow chart of the first embodiment of a pilot optimization method for a sparse channel of an OFDM system according to the present invention;

Fig. 2 is a schematic flowchart of a method for determining the best element at an element position in the initial pilot arrangement according to the present invention;

FIG. 3 is a schematic flowchart of a method for generating a preferred pilot arrangement according to the present invention;

Fig. 4 is a schematic flow chart of a second embodiment of a pilot optimization method for a sparse channel of an OFDM system according to the present invention;

FIG. 5 is a schematic structural diagram of a first embodiment of a pilot optimization device for a sparse channel of an OFDM system according to the present invention;

Fig. 6 is a schematic structural diagram of a preferred pilot arrangement generating unit of the present invention;

Fig. 7 is a schematic structural diagram of a preferred pilot arrangement generating unit of the present invention;

Fig. 8 is the comparison result figure of the convergence performance of the present invention and several prior art;

Fig. 9 is the MSE performance comparison result figure of the present invention and several prior art;

Fig. 10 is a graph showing the comparison results of BER performance between the present invention and several prior art.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Fig. 1 is a schematic flow chart of the first embodiment of a pilot optimization method for a sparse channel of an OFDM system according to the present invention. With reference to Fig. 1, the method includes:

S100: Randomly select elements from the subcarrier set to generate an initial pilot arrangement.

In an implementation of this embodiment, the number of subcarriers is N, the number of pilots is M, and the set subcarrier set {1, 2, 3...N} composed of N integers from 1 to N is randomly selected M elements form a one-dimensional column vector P with a length of M, and the vector P is the initial pilot arrangement.

S102: Calculate the candidate set and determine the best element at the position of each element in the initial pilot arrangement to generate an optimal pilot arrangement; wherein, the optimal pilot arrangement is determined by the best element at the position of each element in the initial pilot arrangement Element composition.

In an implementation of this embodiment, step S102 is implemented by determining the best element at each element position in the initial pilot arrangement one by one, wherein, each time the best element at an element position is determined, the element The element at the position is replaced with the best element, so as to determine and replace the best element at other element positions. Please refer to FIG. 2 specifically. FIG. 2 is a schematic flowchart of a method for determining the best element at an element position in the initial pilot arrangement according to the present invention. The method includes:

S200: Define a variable element and a fixed element, specifically, define the element at the current element position that needs to determine the best element as a variable element, keep the element at the element position other than the current element position unchanged and define it as a fixed element;

S202: Calculate a candidate set, specifically, calculate a difference set between a subcarrier set and a set composed of fixed elements to generate a candidate set;

S204: Determine the best element, specifically, replace the variable element with each element in the candidate set, and calculate the second objective function value (ie g(p) value of the initial pilot arrangement after each replacement occurs, for g For the description of (p), please refer to the corresponding description below), and determine the element with the best value of the second objective function as the best element at the current element position.

After determining the best element at each element position in the initial pilot arrangement according to the above method, the elements at each element position in the initial pilot arrangement are actually updated to the best element, and the initial pilot arrangement at this time can be As an optimized pilot arrangement.

In another implementation manner of this embodiment, the optimal pilot arrangement may be generated iteratively. For example, referring to FIG. 3, FIG. 3 is a schematic flowchart of a method for generating an optimal pilot arrangement according to the present invention, the method including:

S300: Determine the best element at each element position in the initial pilot arrangement one by one, wherein, each time the best element at an element position is determined, replace the element at the element position with the best element (determine the initial pilot Please refer to the implementation shown in Figure 2 for the method of the best element at an element position in the frequency arrangement);

S302: Repeat step S300 for a preset number of times, or repeat step S300 until the best element in the position of each element in the initial pilot arrangement does not change, wherein, each time step S300 is executed, the initial pilot arrangement The elements at the positions of each element in are the newly determined best elements, that is, each time the best element is determined, the initial pilot arrangement is updated at the same time, and after the end of the first cycle, the updated initial pilot arrangement participates in the next cycle.

In this implementation manner, after step S302 is completed, the element at each element position in the initial pilot arrangement is the final determined optimal element, and at this time, the initial pilot arrangement is the optimal pilot arrangement.

S104: Repeat step S100 and step S102 multiple times, and determine the optimal pilot arrangement with the best value of the first objective function (that is, g(p), see later) generated in the multiple repetition process as the optimization guide frequency arrangement.

In step S104, the objective function is the cross-correlation g(p) of the measurement matrix A, specifically:

Let the number of subcarriers of the OFDM system be N, the number of pilots be M, and the index numbers of the pilot subcarriers be k ₁ , k ₂ ,...k _M , satisfying 1≤k ₁ <k ₂ <... <k _M ≤ N. The transmitted pilot symbols are denoted as X(k ₁ ), X(k ₂ ), ..., X(k _M ), and the received pilot symbols are denoted as Y(k ₁ ), Y(k ₂ ), ..., Y (k _M ). Then the frequency domain channel estimation problem of OFDM system can be written as:

Where h=[h(1), h(2),...,h(L)] ^T is an equivalent discrete channel impulse response function, and its length is L. The superscript "T" indicates vector transpose. η=[η(1), η(2),..., η(M)] ^T is a noise vector, each element of which is independently and identically distributed, and satisfies a complex Gaussian distribution with a mean value of 0 and a variance of ^σ2 . For a standard N-dimensional DFT square matrix F, the row numbers of F are k ₁ , k ₂ , ...k _M rows of M and the first L columns of F to form an M×L-dimensional DFT sub-matrix F _M×L . Let y=[Y(k ₁ ), Y(k ₂ ), . . . , Y(k _M )] ^T , X=diag{X(k ₁ ), X(k ₂ ), . . . , X(k _M )} is a diagonal matrix composed of transmitted pilot symbols X(k ₁ ), X(k ₂ ), ..., X(k _M ), and the product of the square matrix X and F _M×L is A=X·F _{M ×L} , formula (1) can be further written as:

y=A h+η(1)

Among them, h=[h(1), h(2), ..., h(L)] ^T is sparse, that is, among the L elements, most of them are zero and only a few non-zero, but non-zero elements The number, location, and value are all unknown. In this case, compressed sensing technology can be used to reconstruct h, and the reconstruction performance is related to the reconstruction algorithm used on the one hand, and is closely related to the matrix A on the other hand. In essence, the channel estimation problem is to estimate h from the known y and A when the noise item is unknown, and make full use of the prior information that h is sparse. The matrix A is called the measurement matrix. If the cross-correlation of A can be minimized, the sparse reconstruction performance will be improved. This involves the optimization of the pilot. Let the pilot arrangement be p=[k ₁ , k ₂ ,...k _M ]. Once p is determined, F _M×L is also determined, and the position of the pilot subcarrier is also determined.

The cross-correlation g(p) of matrix A is defined as:

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="0"\; label="max"\; filenames="HDA00002166014400061.png"\; state="\{\{state\}\}">max</figure-callout></span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; A</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; A</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ></span><span\; class="notranslate">\; |</span>$

$<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="0"\; label="max"\; filenames="HDA00002166014400061.png"\; state="\{\{state\}\}">max</figure-callout></span>}}<span\; class="notranslate">\; |</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; k</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{{\mathrm{<span\; class="notranslate">\; k</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; |</span>$

Then the optimal pilot frequency is:

${\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; opt</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\mathrm{<span\; class="notranslate">\; p</span>}}{\mathrm{<span\; class="notranslate">\; min</span>}}\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>$

That is, the pilot arrangement that minimizes the cross-correlation of matrix A. In fact, the calculation of g(p) is equivalent to finding the absolute value of the element with the largest absolute value among all off-diagonal upper triangular elements of A ^H A, where the superscript "H" denotes the conjugate transpose of the matrix.

In an implementation of this embodiment, the number of repetitions of step S100 and step S102 is preset so as to obtain multiple preferred pilot arrangements and calculate the objective function value (first objective function value) of each preferred pilot arrangement , select the optimal pilot arrangement with the best objective function value as the optimized pilot arrangement. If there are at least two optimal pilot arrangements with the best objective function value, one of them is randomly selected as the optimal pilot arrangement.

Fig. 4 is a schematic flow chart of the second embodiment of a pilot optimization method for a sparse channel of an OFDM system according to the present invention. In this embodiment, it is assumed that the number of subcarriers of the OFDM system is N=256, where the number of pilots is M=16; the guard interval of the OFDM system is 64, and the length of the channel impulse response after sampling is L=60. In formula (2), the length of the vector y is 16, the length of the vector h is 60, and the matrix A is 16 by 64 dimensions.

In this embodiment, when data is initialized (S400), the number of outer loops is preset to be S=100, and the number of second-tier loops T=20, so that 100 initial pilot arrangements are randomly selected in 100 outer loops, and for each An initial pilot arrangement is selected, and an optimal pilot arrangement of the initial pilot arrangement is determined in a maximum of 20 Layer 2 cycles. in,

In step S402, M (16) elements are randomly selected from the subcarrier set {1, 2, ..., 256} composed of 256 integers from 1 to N (256) to form a one-dimensional column vector p with a length of 16 . Let p^ be a zero vector of length 16.

In steps S404 and S406, it is judged whether the vector p and p^ are the same at the beginning of each second-layer cycle, if they are the same, it means that the best element in the position of each element in the initial pilot arrangement will no longer change, and at this time execute Step S410, if different, assign each element in the vector p to p^, and determine the best element at the position of each element in the vector p again.

In step S408, the best element at each element position in the vector p is sequentially determined through M (16) cycles. For example: Assuming m=3, the current random pilot arrangement

p = [4, 5, 18, 39, 55, 72, 92, 111, 130, 153, 177, 192, 211, 218, 237, 241]. First, fix the remaining 15 elements except the third element among all the elements of the vector p, that is, fix 4, 5, 39, 55, 72, 92, 111, 130, 153, 177, 192, 211, 218, 237 , 241 unchanged. Then, let the set c be the set of the remaining 15 elements except the third element in the vector p, that is, c=[4, 5, 39, 55, 72, 92, 111, 130, 153, 177, 192, 211, 218, 237, 241], let the candidate set a be the difference set between the subcarrier set and the set c, that is, a={1, 2,...,256}\c, then the number of elements in the set a is 241. Each time a different element is taken from the set a and placed in the third element position of p, there are a total of 241 placement methods; calculate the objective function corresponding to the pilot arrangement formed by each placement method, from 241 target Take the smallest one in the function, and use its corresponding pilot arrangement as the new p for subsequent iterations.

In step S410, a zero vector b with a length of S (100) and a zero matrix Z of dimension S×M (100×16) can be preset, and each row of Z is used to store a pilot arrangement. After jumping out of the second-layer loop each time, save the vector p in the i-th row of the matrix Z, calculate the objective function value of the vector p and save the objective function value as the i-th element of the vector b.

In step S412, the smallest element is selected from the S (100) elements of the vector b, assuming that the smallest element is the kth element, then the kth row of Z is output, and the kth row is the optimized pilot arrangement.

Of course, it is also possible to compare the objective function values in step S410, and directly output the optimized pilot arrangement in step S412.

FIG. 5 is a schematic structural diagram of a first embodiment of a pilot optimization device for a sparse channel of an OFDM system according to the present invention. The pilot optimization device 50 includes:

An initialization unit 502, configured to randomly select elements from the subcarrier set to generate an initial pilot arrangement;

The preferred pilot arrangement generating unit 504 is configured to generate a candidate set according to the subcarrier set and the initial pilot arrangement, and select the best element at each element position in the initial pilot arrangement from the candidate set to generate the preferred pilot arrangement cloth, wherein the preferred pilot arrangement is formed by the best element at each element position in the initial pilot arrangement;

The optimized pilot arrangement generation unit 506 is used to repeatedly call the initialization unit 502 and the preferred pilot arrangement generation unit 504, and arrange the optimal pilot arrangement with the optimal first objective function value generated during the multiple repetitions Determined to optimize the pilot arrangement.

In an implementation of this embodiment, referring to FIG. 6, the preferred pilot arrangement generation unit 504 includes:

The optimal element determination module 60 is configured to determine the optimal element at each element position in the initial pilot arrangement one by one, wherein, each time the optimal element at an element position is determined, the element at the element position is replaced is the corresponding best element. The optimal element determination module 60 may include:

The definition sub-module 602 is used to define the element at the current element position that needs to determine the best element as a variable element, keep the element at the element position other than the current element position unchanged and define it as a fixed element;

The candidate set generation submodule 604 is used to calculate the difference set of the subcarrier set and the set composed of fixed elements to generate a candidate set;

The determination sub-module 606 is used to replace the variable elements with each element in the candidate set, calculate the second objective function value of the initial pilot arrangement after each replacement, and make the element with the optimal second objective function value Determine the best element at the current element position.

In another implementation of this embodiment, referring to FIG. 7 , the preferred pilot arrangement generation unit 504 includes, in addition to the above-mentioned optimum element determination module 60, also includes:

The cyclic call module 70 is used to repeatedly call the optimal element determination module 60 according to a preset number of times, or to repeatedly call the optimal element determination module 60 until the optimal element at each element position in the initial pilot arrangement does not change, Wherein, each time the optimal element determination module 60 is called, the elements at the positions of the elements in the initial pilot arrangement are the latest determined optimal elements;

The final determination module 72 is configured to determine an optimal pilot arrangement, where the optimal pilot arrangement is formed by the best element at each element position in the initial pilot arrangement finally determined by the cyclic calling module 70 .

In yet another implementation of this embodiment, the optimized pilot arrangement generation unit 506 further includes:

The optimized pilot arrangement selection module is configured to randomly select one of them as the preferred pilot arrangement when there are at least two preferred pilot arrangements with the best first objective function value.

The pilot optimization device 50 and the units, modules and sub-modules of the device are used to realize the corresponding functions in the embodiments shown in FIG. 1 to FIG. 4 , which will not be repeated here.

In addition, the present invention also provides a sparse channel estimation method for an OFDM system, the method comprising:

1) The transmitting end determines the optimized pilot arrangement to insert the optimized pilot. For the specific determination method, please refer to the detailed description in the embodiment shown in Figure 1 to Figure 4, or the transmitting end determines the optimized pilot arrangement through the device shown in Figure 5 cloth;

2) The receiving end performs channel estimation based on compressed sensing technology. The specific compressive sensing technology can adopt the existing technology, which will not be described in detail here.

The pilot optimization method and device for sparse channels provided by the present invention can realize lower measurement matrix cross-correlation and faster convergence speed in the pilot optimization process.

The invention can improve the mean square error and bit error rate performance of OFDM sparse channel estimation.

Figures 8, 9 and 10 are schematic diagrams of comparison results comparing the performance of the present invention with several existing technologies.

In the simulation test, the number of sub-carriers of the OFDM system is 256, the number of pilots is 16, and the guard interval of the OFDM system is 64. QPSK modulation is adopted, the length of the channel impulse response is 60, and the number of non-zero taps of the channel is 6. The simulation platform is based on the Windows 7 operating system, MATLAB2011a software, the CPU is dual-core 2.5GHz, and the memory is 3G bytes. The set program running time is 348.6 seconds. The optimized pilot results obtained by using the present invention, prior art 4, and prior art 3 are [4,21,29,46,63,95,116,136,140,184,187,192,197,200,221,248], [35,38,45,47,49,71,74, 79,99,115,147,156,174,194,213,240], [11,23,42,60,105,127,148,171,175,178,182,190,205,207,217,241], corresponding to the objective functions 4.8421, 5.3535, 5.0241, respectively.

Fig. 8 is a comparison result diagram of the convergence performance between the present invention and several existing technologies.

When performing MSE and BER performance simulations, each channel generation method is as follows: select 6 taps from 60 taps, and the tap coefficients at these 6 tap positions obey the complex Gaussian distribution, and use The mainstream Orthogonal Matching Pursuit (OMP) algorithm performs sparse channel estimation, the channel is randomly generated 30,000 times, and finally the results of 30,000 times are averaged. During channel estimation, it is assumed that the number of non-zero taps of the channel is unknown, the positions of non-zero taps are unknown, and the values of coefficients of non-zero taps are unknown.

Fig. 9 is a graph showing the comparison results of MSE performance between the present invention and several prior art.

Fig. 10 is a graph showing the comparison results of BER performance between the present invention and several prior art.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented through computer programs to instruct related hardware, and the programs can be stored in a computer-readable storage medium. During execution, it may include the processes of the embodiments of the above-mentioned methods. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM), etc.

The above disclosure is only a preferred embodiment of the present invention, which certainly cannot limit the scope of rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope of the present invention.

Claims (10)

1. a kind of pilot optimization method of sparse channel, is characterized in that, described method comprises: Randomly select elements from the subcarrier set to generate an initial pilot arrangement, generate a candidate set according to the subcarrier set and the initial pilot arrangement, and select each element in the initial pilot arrangement from the candidate set The best element in position to generate a preferred pilot arrangement, wherein the preferred pilot arrangement is composed of the best elements in the position of each element in the initial pilot arrangement; The above steps are repeated multiple times, and the optimal pilot arrangement with the best first objective function value generated during the repeated repetitions is determined as the optimal pilot arrangement. 2. The method according to claim 1, characterized in that, generating a candidate set according to the subcarrier set and the initial pilot arrangement and selecting each of the initial pilot arrangements from the candidate set The best elements on element position include: Step A: Determine the best element at each element position in the initial pilot arrangement one by one, wherein, each time the best element at an element position is determined, replace the element at the element position with the corresponding best element good elements. 3. The method according to claim 2, wherein said determining the best element at an element position comprises: Defining the element at the current element position that needs to determine the best element as a variable element, keeping the element at the element position other than the current element position unchanged and defining it as a fixed element; calculating a difference set between the set of subcarriers and the set of fixed elements to generate the candidate set; Respectively replace the changed elements with each element in the candidate set, calculate the second objective function value of the initial pilot arrangement after each replacement, and determine the element that makes the second objective function value optimal as the current The best element at element position. 4. The method according to claim 2 or 3, wherein said generating a candidate set according to said subcarrier set and said initial pilot arrangement and selecting said initial pilot arrangement from said candidate set The best elements at each element position in also include: Step B: Repeat step A for a preset number of times, or repeat step A until the best element in the position of each element in the initial pilot arrangement does not change, wherein, each time step A is executed, the initial The elements at the position of each element in the pilot arrangement are the latest determined best elements; At this time, the optimal pilot arrangement is composed of the best element at each element position in the initial pilot arrangement finally determined in step B. 5. The method according to claim 4, wherein said determining the optimal pilot arrangement with the optimal first objective function value generated in the multiple repetition process as an optimized pilot arrangement comprises: If there are at least two optimal pilot arrangements with the best value of the first objective function, one of them is randomly selected as the optimal pilot arrangement. 6. A pilot optimization device for sparse channels, characterized in that the device comprises: The initialization unit is used to randomly select elements from the subcarrier set to generate an initial pilot arrangement; A preferred pilot arrangement generating unit, configured to generate a candidate set according to the subcarrier set and the initial pilot arrangement and select the best position of each element in the initial pilot arrangement from the candidate set elements to generate a preferred pilot arrangement, wherein the preferred pilot arrangement is formed by the best element at each element position in the initial pilot arrangement; The optimized pilot arrangement generation unit is used to repeatedly call the initialization unit and the optimal pilot arrangement generation unit, and determine the optimal pilot arrangement with the best first objective function value generated during the multiple repetitions To optimize the pilot arrangement. 7. The device according to claim 6, wherein the preferred pilot arrangement generating unit comprises: The optimal element determination module is configured to determine the optimal element at each element position in the initial pilot arrangement one by one, wherein, each time the optimal element at an element position is determined, the element at the element position Replaced by the corresponding best element. 8. The device according to claim 7, wherein the optimal element determination module comprises: A submodule is defined, which is used to define the element at the current element position that needs to determine the best element as a variable element, keep the element at the element position other than the current element position unchanged and define it as a fixed element; A candidate set generating submodule, configured to calculate the difference between the set of subcarriers and the set of fixed elements to generate the candidate set; The determining submodule is used to replace the changed elements with each element in the candidate set, calculate the second objective function value of the initial pilot arrangement after each replacement, and maximize the second objective function value The optimal element is determined as the best element at the current element position. 9. The device according to claim 7 or 8, wherein the preferred pilot arrangement generating unit further comprises: The cyclic calling module is used to repeatedly call the optimal element determination module according to the preset number of times, or to repeatedly call the optimal element determination module until the optimal element at each element position in the initial pilot arrangement is no longer Further change, wherein, each time the optimal element determination module is called, the elements at the positions of the elements in the initial pilot arrangement are the latest determined optimal elements; Final determination module: used to determine the optimal pilot arrangement, the optimal pilot arrangement is composed of the best elements at the positions of the elements in the initial pilot arrangement finally determined by the loop calling module. 10. A sparse channel estimation method, characterized in that the method comprises: The transmitting end determines the optimized pilot arrangement according to the method of claim 1 to insert the optimized pilot; The receiver performs channel estimation based on compressed sensing technology.

Images