CN104780130A - Improved pilot-frequency-based SIM-OFDM channel estimation method - Google Patents

Abstract

The invention discloses an improved pilot-based SIM-OFDM channel estimation method, which belongs to the technical field of communication anti-interference. The invention aims at the technical problem that the channel estimation of the existing SIM-OFDM communication method is difficult to implement, and proposes a method in The method of inserting pilot sequences on silent subcarriers to complete channel estimation: the transmitting end inserts pilot sequences of arbitrary energy on silent subcarriers; the receiving end performs unified processing on multiple received signals, according to the formula Get the average channel frequency domain response ,based on Obtain preliminary channel estimates of multiple received signals, then perform equalization processing on the received signals based on the current channel estimates, and then perform pilot position detection, so as to receive signals based on the pilot position, pilot data, and frequency domain at the pilot position The channel estimate is updated. The invention is used for channel estimation of SIM-OFDM system and data demodulation of received signal in frequency domain, not only retains the advantages of SIM-OFDM system itself, but also can use pilot frequency for channel estimation, providing basis for final detection algorithm.

Description

An Improved Channel Estimation Method Based on Pilot for SIM-OFDM

technical field

The invention belongs to the technical field of communication anti-interference, and in particular relates to a pilot-based SIM-OFDM channel estimation method.

Background technique

OFDM (Orthogonal Frequency Division Multiplexing) technology is a high-speed transmission technology for wireless communication. Its basic principle is to decompose the high-speed data stream into many low-rate sub-data streams, that is, divide the signal into many orthogonal sub-carriers, and use these mutual Orthogonal subcarriers are transmitted simultaneously. This technology uses subcarriers to modulate data, expands the pulse width of symbols, can effectively resist Inter-Symbol Interference (ISI), and improves the performance against multipath fading. Compared with traditional frequency division multiplexing (FDM), OFDM does not require a dedicated guard band. Although there will be overlap between frequency spectrums, each carrier is orthogonal to each other. According to the principle of orthogonality, it can be known that there is no interference between the various carriers, thus greatly improving the utilization rate of the frequency spectrum.

Recently, a new multi-carrier communication method——OFDM system based on Subcarrier Index Modulation (SIM) was proposed. For the SIM-OFDM system, except for adding the SIM modulation module, other steps are not different from the traditional OFDM system. Among them, the core SIM modulation module adopts the idea of subcarrier division. First, the entire multi-carrier is continuously divided into multiple sub-blocks of the same size, and in each sub-block, several sub-carriers (called active sub-carriers) are selected to transmit data through index bits, while the rest of the sub-carriers do not transmit data (called active sub-carriers). are silent subcarriers). Since the index bit itself is not sent, but is implied in the location information of the activated subcarrier, the index bit does not occupy spectrum resources. At the receiving end, the information of the index bit can be obtained by activating the position of the subcarrier.

Compared with the traditional OFDM communication method, the existing SIM-OFDM has many advantages, such as the smaller peak-to-average power ratio of the SIM-OFDM system, better performance against inter-subcarrier interference, and lower bit error rate. The power allocation strategy can also save transmitter energy. The disadvantage is that the SIM-OFDM system cannot simply insert pilots at fixed positions to complete channel estimation like in the OFDM system due to the block method. In the SIM-OFDM system, if a pilot is inserted at a fixed position in each block, it will cause a waste of spectrum resources and lose its own advantages.

Contents of the invention

Aiming at the technical problem that channel estimation is difficult to implement in the existing SIM-OFDM communication method, the present invention proposes a method for inserting arbitrary energy pilot sequences on silent subcarriers to complete channel estimation. Therefore, not only the advantages of the SIM-OFDM system are retained, but also the pilot frequency can be used for channel estimation.

A kind of channel estimation method based on the SIM-OFDM of pilot frequency improvement of the present invention, comprises the following steps:

sender:

Step 1-1: Sub-carriers of the system are divided into blocks to obtain g=N/n sub-blocks, where N represents the total number of sub-carriers of the system, n represents the number of sub-carriers of each sub-block, and determine the system activate subcarriers;

Step 1-2: For each sub-block of the system, the active sub-carrier carries the transmission data, the silent sub-carrier carries the pilot data, and the pilot data in each sub-block is the same;

Receiving end:

Step 2-1: Initial channel estimation:

Average the m received signals Y _r to get the average signal Where r=1,2,...,m, m≤1/f _doppler , f _doppler represents the Doppler frequency;

According to the formula to preprocess the received signal Where k represents the number of activated subcarriers of each sub-block of the system, and the pilot data P=[P ₁ P ₂ ... P _g ] ^T is extended to the preprocessed received signal The dimensions are the same, and the extended pilot matrix is obtained Where i=1,2,...,g, j=1,2,...,n, use P _i to represent each element in the pilot data P, then the value of each element P _i,j in the matrix is P _{i , j} = P _i ;

According to the formula Get the average channel frequency domain response Based on the principle that channel changes are slow in the time domain, the average channel frequency domain response Get channel estimates for m received signals in r=1,2,...,m.

Step 2-2: Pilot location judgment:

Based on channel estimates Perform equalization processing on the received signal Y=[Y ₁ Y ₂ ... Y _m ] to obtain an equalized signal Equalization methods include, but are not limited to, zero-breaking equalization (ZF), linear minimum mean square error equalization (LMMSE), and the like. At the same time, using the same block method as the sending end, each equalized signal Divided into g sub-blocks, each sub-block contains n subcarriers, where r=1,2,...,m;

Compute the equalized signal separately The Euclidean distance between the data on the n subcarriers in each subblock and the pilot data corresponding to the current subblock, find the nk subcarrier positions with the smallest Euclidean distance and record it as the pilot position of the current subblock l _i (subscript i is the sub-block identifier), the pilot position set L _r is formed by g pilot positions l _i , and then the pilot position set L={L _r } _1≤r≤m is obtained from m pilot position sets L _r , based on the pilot position set L to obtain the frequency-domain received signal Y _p at the pilot position = [Y _p,1 Y _p,2 ... Y _p,m ];

Step 2-3: Update channel estimates Channel estimation is performed based on the frequency-domain received signal Y _P at the pilot position, the pilot position set L and the pilot data P (channel estimation methods include but not limited to linear least squares estimation (LS), linear minimum mean square error estimation ( LMMSE) etc.), to get the updated channel estimate

In summary, due to the adoption of the above technical solution, the beneficial effect of the present invention is: on the basis of the existing SIM-OFDM system, inserting a pilot sequence of arbitrary energy on the silent subcarrier to complete the channel estimation method, namely The transmitting end inserts arbitrary energy pilot sequences on silent subcarriers, and the receiving end performs unified processing on multiple received signals, according to the formula Get the average channel frequency domain response Then obtain the preliminary channel estimation value, and then perform equalization processing on the received signal based on the current channel estimation value, and then perform pilot position detection, so as to perform channel estimation based on the pilot position, pilot data and the frequency domain received signal at the pilot position. renew. The invention not only retains the advantages of the SIM-OFDM system itself, but also can use the pilot frequency to perform channel estimation and provide a basis for the final detection algorithm.

Detailed ways

In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the embodiments.

The present invention is based on the existing SIM-OFDM system, and the difference from the existing SIM-OFDM system is that the transmitting end uses the pilot sequence on the silent subcarrier, and then the receiving end uses a special detection method to detect the The position is detected to complete the channel estimation, and the SIM modulation, sending and receiving processes involved are the same as the existing SIM-OFDM system.

The following takes (n,k)=(2,1), the total number of subcarriers N=1024, the cyclic prefix CP=64, the modulation symbol uses QPSK (4-QAM), ZF equalization and LMMSE channel estimation are used as an example to introduce Specific embodiments of the present invention.

sender:

Step 1-1: Determine the parameters of the system to be selected, that is, determine the number of sub-carriers N=1024, the number of sub-carriers in each sub-block n=2, how many sub-carriers are activated in the sub-block k=1, divided into The number of sub-blocks g=N/n=512, and the modulation order M=4. Then calculate the number of bits in a frame according to the formula. For any one of the sub-blocks, the index bit length is: Represents rounding down, then for the index OFDM system of one frame, there are m ₁ =p ₁ g=512 index bits; send 4-QAM modulation symbols to the activated subcarriers, for (n,k)=( 2,1), the number of modulation bits that can be sent in one frame is: m ₂ =gp ₂ =gklog ₂ M=1024, then the total number of bits in one frame m=m ₁ +m ₂ =1536. The frame data is divided into two groups, one group is index bits, which are used to select subcarriers to transmit data, and one group is modulation bits, which are used to modulate into activated subcarriers and send them out.

Step 1-2: Perform conventional SIM modulation, determine the corresponding activated subcarriers (the number of activated subcarriers in each subblock is denoted by k), and then assign SIM modulation symbols to the corresponding activated subcarriers.

The specific process of SIM modulation is: k subcarriers carry p ₂ -bit bit information, each subcarrier carries 1 symbol, and each symbol is an M-order signal modulation symbol mapped from log ₂ M bits. In each sub-block, it is determined which k active subcarriers are used to transmit the k M-order signal modulation symbols according to the values of _p1 bits.

Receiving end:

Perform unified processing on six SIM-OFDM received signals Y _r , r=1,2,...,6, the dimension of any Y _r is 1024×1, then the received signal Y=[Y ₁ Y ₂ Y ₃ Y ₄ Y ₅ Y ₆ ] has a dimension of 1024×6.

Step 1: Initial channel estimation. Calculate the average of 6 SIM-OFDM received signals Y _r , r=1,2,...,6 In the transmitted signal, the probability of pilot frequency appearing on each subcarrier is Using the mean value of the received signal Divide by this probability to get the preprocessed received signal Because the pilots in each sub-block are the same, the pilot data P=[P ₁ P ₂ ... P ₅₁₂ ] ^T is extended so that the dimension of Y is the same, and the extended pilot matrix is obtained $\tilde{P}=\left[\begin{array}{ccccccc}{P}_1& P_1& P_2& P_2& ...& P_{512}& P_{512}\end{array}\right],$ Receive signal by preprocessing pilot data According to the formula Get the average channel frequency domain response In the time domain, the change of the channel is considered to be slow, and the average channel frequency domain response is used Indicates the channel frequency domain response of 6 consecutive symbols, and obtains the channel estimation value $\hat{h}=\left[\begin{array}{cccccc}\tilde{h}& \tilde{h}& \tilde{h}& \tilde{h}& \tilde{h}& \tilde{h}\end{array}\right].$

Step 2: Judging the pilot position. Using the channel estimate obtained in step 1 Equalize the received signal Y to get Similarly, for balanced signals Each column of (ie ) is divided into blocks in the same manner as the sending end to obtain g=512 sub-blocks, each block containing 2 sub-carriers. After chunking, check each chunk one by one. The detection method is: the data on the 2 subcarriers in each block ( Blocked data) is compared with the pilot data P _i corresponding to the block by Euclidean distance d _j , where Where i represents the i-th sub-block, j represents the j-th sub-carrier in the i-th sub-block, the position of the sub-carrier with the smallest Euclidean distance d _j is considered to be the sub-carrier position where the pilot data P _i of the current sub-block is located, that is, the i-th sub-carrier In the sub-blocks, the sub-carrier position corresponding to the minimum Euclidean distance d _j is recorded as the pilot position l _i , and the pilot position l _i of 512 sub-blocks constitutes a pilot position set L _r corresponding to each received signal, and finally we get Pilot position set L={L ₁ L ₂ ... L ₆ }, from the set L of pilot positions, the frequency-domain received signal Y _p at the pilot position can be obtained = [Y _p,1 Y _p,2 ... Y _{p, 6} ].

Step 3: Update the channel estimate. Channel estimation is obtained from the frequency domain received signal Y _p at the pilot position, the pilot data P and the pilot position set L obtained in step 2

If it is considered that the channel estimate obtained in step 3 If the performance requirements of the system cannot be met, the number of times to update the channel estimation can be increased, that is, steps 2 and 3 are repeated, and when the number of repetitions is equal to the preset threshold T (T ranges from 1 to 3), the repetition is terminated to obtain the final channel estimate

The channel estimation value obtained based on the present invention It is used for signal detection of received signals Y _r , r=1, 2, ..., 6, to obtain demodulated data, and the signal detection method may adopt conventional ML detection or the like.

The above is only a specific embodiment of the present invention. Any feature disclosed in this specification, unless specifically stated, can be replaced by other equivalent or alternative features with similar purposes; all the disclosed features, or All method or process steps may be combined in any way, except for mutually exclusive features and/or steps.

Claims (5)

1. the channel estimation methods of the SIM-OFDM based on pilot tone improved, is characterized in that, comprise the following steps:

Transmitting terminal:

Step 1-1: carry out piecemeal process to the subcarrier of system, obtains g=N/n sub-block, and wherein N represents the total number of the subcarrier of system, and n represents the subcarrier number of each sub-block, and the activation subcarrier of certainty annuity;

Step 1-2: to each sub-block of system, carry transmission data by activation subcarrier, silent subcarriers carries pilot data, and pilot data in each sub-block is all identical;

Receiving terminal:

Step 2-1: initial channel estimation:

To m Received signal strength Y _rbe averaging and obtain average signal wherein r=1,2 ..., m, m≤1/f _doppler, f _dopplerrepresent Doppler frequency;

According to formula obtain preliminary treatment Received signal strength wherein k represents the activation subcarrier number of each sub-block of system, by pilot data P=[P ₁p ₂p _g] ^textend to and preliminary treatment Received signal strength dimension identical, be expanded pilot matrix wherein i=1,2 ..., g, j=1,2 ..., n, uses P _irepresent each element in pilot data P, then matrix in each element value be

According to formula obtain average channel frequency domain response by average channel frequency domain response obtain the channel estimation value of m Received signal strength wherein r=1,2 ..., m;

Step 2-2: pilot frequency locations judges:

Based on channel estimation value y=[Y to received signal ₁y ₂y _m] carry out equilibrium treatment and obtain equalizing signal

Adopt the partitioned mode identical with transmitting terminal, by each equalizing signal be divided into g sub-block, each sub-block comprises n subcarrier, wherein r=1, and 2 ..., m;

Calculate equalizing signal respectively each sub-block in Euclidean distance between data on n the subcarrier pilot data corresponding with current sub-block, search the pilot frequency locations of the minimum n-k of an Euclidean distance sub-carrier positions as current sub-block, form pilot frequency locations set L by the pilot frequency locations of g sub-block _r, by m pilot frequency locations set L _robtain pilot frequency locations set L, obtain the frequency-domain received signal Y at pilot frequency locations place based on pilot frequency locations set L _p=[Y _{p, 1}y _{p, 2}y _p,m];

Step 2-3: upgrade channel estimation value based on the frequency-domain received signal Y at pilot frequency locations place _p, pilot frequency locations set L and pilot data P carries out channel estimating, obtains the channel estimation value after upgrading

2. the method for claim 1, is characterized in that, repeated execution of steps 2-2 and 2-3, until number of repetition is more than or equal to predetermined threshold value T.

3. method as claimed in claim 2, it is characterized in that, the span of described predetermined threshold value T is 1 ~ 3.

4. method as claimed in claim 1 or 2, it is characterized in that, the equalization processing method described in step 2-2 is: broken zero equalization methods or linear minimum mean-squared error equalization methods.

5. method as claimed in claim 1 or 2, it is characterized in that, channel estimation methods described in step 2-3 is: Linear least square estimation method or Linear Minimum Mean-Square Error Estimation method.