CN102098266B - Synchronization sequence construction method for multi-input multi-output orthogonal frequency division multiplexing (OFDM) system - Google Patents

Abstract

The invention relates to a method for constructing a synchronous sequence of a MIMOOFDM system, the method comprising the following steps: constructing a chu sequence s with a length of P, wherein, P<N, and (N) _P =0 , (g) _P is a modulo P operator, (N) _P means doing a modulo P operation on N, and N is the length of the synchronization sequence; perform a cyclic shift operation on the chu sequence s to obtain a shift sequence s ^(μM) , where , g ^(g) is the cyclic shift operator, s ^(μM) represents the vector obtained by cyclically shifting the chu sequence s down by μM bits, M is the shift parameter, μ represents the index of the transmitting antenna, is the rounding operator, Represents the rounding of P/N _I , the design parameter N _I makes N _t ≤ N _I <P, and N _t represents the number of transmitting antennas. The synchronous sequence construction method proposed by the invention is simple to generate and has low system load.

Description

Synchronization sequence construction method for MIMO system

technical field

The invention relates to a method for constructing a synchronous sequence applicable to a multiple-input multiple-output orthogonal frequency division multiplexing system, belonging to the technical field of synchronization in mobile communication.

Background technique

Orthogonal frequency-division multiplexing (OFDM, orthogonal frequency-division multiplexing) is very suitable for wireless communication channel transmission due to its advantages of high spectral efficiency, anti-frequency selective fading and narrow-band interference, and is also used in the current new generation of mobile communication systems. mainstream technology. The combination of multiple-input multiple-output technology and OFDM technology can provide greater spectral efficiency and higher data rates; in the current situation where spectrum resources are increasingly tight and data rates are increasingly demanding, multiple-input multiple-output OFDM technology has a broader application prospect.

However, systems using OFDM technology have a common shortcoming, which is very sensitive to synchronization errors, especially carrier frequency offset. Since the spectrums of the sub-channels of the OFDM system cover each other, strict requirements are placed on their orthogonality. However, due to the time-varying nature of the wireless channel, there will be a frequency offset of the wireless signal during transmission, such as a Doppler shift, or a frequency offset between the carrier frequency of the transmitter and the local oscillator of the receiver. The carrier frequency offset which only accounts for a small part of the subcarrier spacing will destroy the orthogonality between the subcarriers of the OFDM system, and then generate inter-subcarrier interference, which will lead to a significant decline in the performance of the OFDM receiver. Therefore, for OFDM receivers, the estimation of the carrier frequency offset and the design of the training sequence are very important. Designing a suitable training sequence can obviously improve the estimation performance and reduce the complexity of the algorithm appropriately.

Contents of the invention

Technical problem: The purpose of the present invention is to provide a synchronization sequence construction method that can effectively estimate the frequency offset in the MIMO system. The proposed synchronization sequence construction method is simple to generate and the system load is low; based on the proposed synchronization sequence, environment adaptive low-complexity frequency offset estimation can be realized, and a variety of simplified frequency offset estimation algorithms can be realized; through the design of sequence parameters, the maximum Optimize the range of frequency offset estimation, optimize the performance of frequency offset estimation, and ensure the consistency of frequency offset estimation.

Technical solution: In order to solve the above-mentioned technical problems, the present invention provides a method for constructing a synchronous sequence of a MIMOOFDM system, the method comprising the following steps:

1) Construct a chu sequence s of length P, where P<N, and (N) _P = 0, (N) _P means modulo P operation on N, and N is the length of the synchronization sequence;

2) Perform a cyclic shift operation on the chu sequence s to obtain a shift sequence s ^(μM) , where · ^(·) is a cyclic shift operator, and s ^(μM) means that the chu sequence s is cyclically shifted down by μM bits The resulting vector, M is the shift parameter, μ represents the index of the transmitting antenna, is the rounding operator, Represents the rounding of P/N _I , the design parameter N _I makes N _t ≤ N _I <P, and N _t represents the number of transmitting antennas;

3) Perform discrete Fourier transform on the shift sequence s ^(μM) to obtain the frequency domain sequence Among them, F _P is a normalized discrete Fourier transform matrix, Q=N/P, and Q>N _t , Q represents the distance between adjacent non-zero pilots corresponding to the training sequence on the transmitting antenna, and P represents The length of the chu sequence s, N is the length of the synchronization sequence, and N _t represents the number of transmitting antennas;

4) For frequency domain sequences Do matrix operations to get the training sequence corresponding to the μth root transmit antenna ${\tilde{t}}_{\mathrm{\&mu;}}={\mathrm{\&Theta;}}_{i_{\mathrm{\&mu;}}}{\widetilde{\mathrm{the\; s}}}_{\mathrm{\&mu;}},$ in, ${\mathrm{\&Theta;}}_{i_{\mathrm{\&mu;}}}=[e_N^{i_{\mathrm{\&mu;}}},e_N^{i_{\mathrm{\&mu;}}+Q},\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;,e_N^{i_{\mathrm{\&mu;}}+(P-1)Q}],$ Represents the i _μth column vector of the unit matrix I _N , i _μ indicates the index of the first non-zero element of the training sequence corresponding to the μth root transmitting antenna, μ represents the index of the transmitting antenna, Q represents the index corresponding to the transmitting antenna The spacing between adjacent non-zero pilots in the training sequence, P represents the length of the chu sequence s, N is the length of the synchronization sequence, $0\&le;i_0<i_1<\&CenterDot;\&CenterDot;\&CenterDot;<i_{\mathrm{\&mu;}}<\&CenterDot;\&CenterDot;\&CenterDot;<i_{N_t-1}<Q.$

Obtain the training sequence corresponding to the μth transmit antenna The conditions that should be met are:

a) Criteria for uniform energy distribution of transmitting antennas: frequency domain sequence N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, and ||·|| represents the 2-norm of the corresponding vector;

b) Frequency offset estimation consistency criterion: (NN _t P)≥N _t P, P≥L, (1 _Q -l) ^T l ^(q) >0, Among them, L represents the maximum multipath delay of the channel, and 1 _Q represents a vector of all 1s of Q×1, Represents the non-zero pilot vector of Q×1; N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, P represents the length of the chu sequence s, and Q represents the adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna , q represents the index corresponding to the zero pilot between adjacent non-zero pilots in the training sequence on the transmitting antenna, ^T represents the transpose operation, and i _μ represents the training sequence corresponding to the μth transmitting antenna The index of the first non-zero element of , μ represents the index of the transmitting antenna;

c) Integer frequency offset estimation reliability criterion: $\max \underset{\mathrm{\&mu;}\&NotEqual;{\mathrm{\&mu;}}^{\&prime;}}{\min }\left\{{(i_{{\mathrm{\&mu;}}^{\&prime;}}-i_{\mathrm{\&mu;}})}_Q\right\},$ $\max \underset{1\&le;q\&le;Q-1}{\min }\left\{{(1_Q-l)}^T{l}^{\left(q\right)}\right\},$ i _μ represents the index of the first non-zero element of the training sequence corresponding to the μ-th transmitting antenna, i _μ′ represents the index of the first non-zero element of the training sequence corresponding to the μ-th transmitting antenna, and ^T represents Transpose operation, q represents the index corresponding to the zero pilot between adjacent non-zero pilots in the training sequence on the transmitting antenna, and Q represents the index corresponding to the adjacent non-zero pilot in the training sequence on the transmitting antenna spacing,

d) Simplified criterion for fractional frequency offset estimation: N _r P > Q, P represents the length of the chu sequence s, Q represents the distance between adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna, and N _r represents the receiving antenna Number of.

Beneficial effects: the synchronization sequence construction method proposed by the present invention only needs to perform cyclic shift operation, discrete Fourier transform and zero insertion operation on the chu base sequence to generate training sequences corresponding to each transmitting antenna, and the sequence construction method is simple and systematic The load is small.

The synchronization sequence construction method proposed by the present invention, based on its frequency offset estimation consistency criterion, can ensure that the frequency offset estimation of the receiver is within a sufficiently large frequency offset range are consistent and identifiable;

The synchronization sequence construction method proposed by the present invention, based on its integer frequency offset estimation reliability criterion, can realize the simplified calculation of the integer frequency offset, especially, in a friendly channel environment, the integer frequency offset can be realized by simple binary logic operation Low-complexity implementation of partial estimation;

The synchronous sequence construction method proposed by the present invention, based on the simplification criterion of the fractional frequency offset estimation and the period attribute of the constructed sequence, can flexibly adopt various methods according to actual needs to realize low-complexity and high-precision fractional frequency offset estimation;

The synchronous sequence construction method proposed by the present invention is suitable for centralized multi-antenna systems, distributed multi-antenna systems, cooperative multi-relay systems, cooperative multi-user systems and other communication systems conforming to the basic characteristics of multiple input and multiple outputs.

Description of drawings

Fig. 1 is a block diagram of a multiple-input multiple-output OFDM system transmitter;

Fig. 2 is a schematic diagram of a synchronous sequence structure of a multiple-input multiple-output OFDM system;

Fig. 3 is a schematic diagram of the realization structure of the synchronous sequence construction method of the MIMOOFDM system.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings.

The present invention is based on a multiple-input multiple-output Orthogonal Frequency-Division Multiplexing (OFDM, orthogonal frequency-division multiplexing) system, and is proposed for the problem of frequency offset existing in the system.

A method for constructing a synchronous sequence of a multiple-input multiple-output OFDM system, the method comprising the steps of:

1) Construct a chu sequence s of length P, where P<N, and (N) _P = 0, (N) _P means modulo P operation on N, and N is the length of the synchronization sequence;

2) Perform a cyclic shift operation on the chu sequence s to obtain a shift sequence s ^(μM) , where · ^(·) is a cyclic shift operator, and s ^(μM) means that the chu sequence s is cyclically shifted down by μM bits The resulting vector, M is the shift parameter, μ represents the index of the transmitting antenna, is the rounding operator, Represents the rounding of P/N _I , the design parameter N _I makes N _t ≤ N _I <P, and N _t represents the number of transmitting antennas;

3) Perform discrete Fourier transform on the shift sequence s ^(μM) to obtain the frequency domain sequence Among them, F _P is a normalized discrete Fourier transform matrix, Q=N/P, and Q>N _t , Q represents the distance between adjacent non-zero pilots corresponding to the training sequence on the transmitting antenna, and P represents The length of the chu sequence s, N is the length of the synchronization sequence, and N _t represents the number of transmitting antennas;

4) For frequency domain sequences Do matrix operations to get the training sequence corresponding to the μth root transmit antenna ${\tilde{t}}_{\mathrm{\&mu;}}={\mathrm{\&Theta;}}_{i_{\mathrm{\&mu;}}}{\widetilde{\mathrm{the\; s}}}_{\mathrm{\&mu;}},$ in, ${\mathrm{\&Theta;}}_{i_{\mathrm{\&mu;}}}=[e_N^{i_{\mathrm{\&mu;}}},e_N^{i_{\mathrm{\&mu;}}+Q},\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;,e_N^{i_{\mathrm{\&mu;}}+(P-1)Q}],$ Represents the i _μth column vector of the unit matrix I _N , i _μ indicates the index of the first non-zero element of the training sequence corresponding to the μth root transmitting antenna, μ represents the index of the transmitting antenna, Q represents the index corresponding to the transmitting antenna The spacing between adjacent non-zero pilots in the training sequence, P represents the length of the chu sequence s, N is the length of the synchronization sequence, $0\&le;i_0<i_1<\&Center\; Dot;\&CenterDot;\&Center\; Dot;<i_{\mathrm{\&mu;}}<\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;<i_{N_t-1}<Q.$

a) Criteria for uniform energy distribution of transmitting antennas: frequency domain sequence N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, and ||·|| represents the 2-norm of the corresponding vector;

b) Frequency offset estimation consistency criterion: (NN _t P)≥N _t P, P≥L, (1 _Q -l) ^T l ^(q) >0, Among them, L represents the maximum multipath delay of the channel, and 1 _Q represents a vector of all 1s of Q×1, Represents the non-zero pilot vector of Q×1; N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, P represents the length of the chu sequence s, and Q represents the adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna , q represents the index corresponding to the zero pilot between adjacent non-zero pilots in the training sequence on the transmitting antenna, ^T represents the transpose operation, and i _μ represents the training sequence corresponding to the μth transmitting antenna The index of the first non-zero element of , μ represents the index of the transmitting antenna;

c) Integer frequency offset estimation reliability criterion: $\max \underset{\mathrm{\&mu;}\&NotEqual;{\mathrm{\&mu;}}^{\&prime;}}{\min }\left\{{(i_{{\mathrm{\&mu;}}^{\&prime;}}-i_{\mathrm{\&mu;}})}_Q\right\},$ $\max \underset{1\&le;q\&le;Q-1}{\min }\left\{{(1_Q-l)}^T{l}^{\left(q\right)}\right\},$ i _μ represents the index of the first non-zero element of the training sequence corresponding to the μ-th transmitting antenna, i _μ′ represents the index of the first non-zero element of the training sequence corresponding to the μ-th transmitting antenna, and ^T represents Transpose operation, q represents the index corresponding to the zero pilot between adjacent non-zero pilots in the training sequence on the transmitting antenna, and Q represents the index corresponding to the adjacent non-zero pilot in the training sequence on the transmitting antenna spacing,

d) Simplified criterion for fractional frequency offset estimation: N _r P > Q, P represents the length of the chu sequence s, Q represents the distance between adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna, and N _r represents the receiving antenna Number of.

In the MIMO OFDM system considered in the present invention, the numbers of transmitting antennas and receiving antennas are respectively set as N _t and N _r . Let the length of one OFDM symbol or synchronous training sequence be N, let μ represent the index of the transmitting antenna, 0≤μ<N _t .

Based on the above-mentioned definition about system and training sequence, the method for synchronous sequence construction of MIMO OFDM system of the present invention is:

1) Construct a chu sequence s with a length of P, wherein, P<N, and (N) _P = 0, (·) _P is a modulo P operator; the chu sequence is a training sequence widely used in the field of wireless communication, and The corresponding Chinese literature is also directly called the chu sequence;

2) Perform a cyclic shift operation on s to obtain a sequence s ^(μM) , where ^{( )} is a cyclic shift operator, M is a shift parameter, is a rounding operator, design parameter N _I so that N _t ≤ N _I <P;

3) Do a discrete Fourier transform on s ^(μM) to get the sequence Wherein, F _P is a normalized discrete Fourier transform matrix, Q=N/P, and Q>N _t ;

4) yes Do matrix operations to get the training sequence corresponding to the μth root transmit antenna ${\tilde{t}}_{\mathrm{\&mu;}}={\mathrm{\&Theta;}}_{i_{\mathrm{\&mu;}}}{\widetilde{\mathrm{the\; s}}}_{\mathrm{\&mu;}},$ in, ${\mathrm{\&Theta;}}_{i_{\mathrm{\&mu;}}}=[e_N^{i_{\mathrm{\&mu;}}},e_N^{i_{\mathrm{\&mu;}}+Q},\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;,e_N^{i_{\mathrm{\&mu;}}+(P-1)Q}],$ Represents the i _μth column vector of the identity matrix I _N , i _μ represents the index of the first non-zero element of the training sequence corresponding to the μth root transmit antenna, $0\&le;i_0<i_1<\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;<i_{\mathrm{\&mu;}}<\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;<i_{N_t-1}<Q.$

The synchronization sequence design method is as follows:

a) Criteria for uniform energy distribution of transmitting antennas: frequency domain sequence N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, and ||·|| represents the 2-norm of the corresponding vector;

b) Frequency offset estimation consistency criterion: (NN _t P)≥N _t P, P≥L, (1 _Q -l) ^T l ^(q) >0, Among them, L represents the maximum multipath delay of the channel, and 1 _Q represents a vector of all 1s of Q×1, Represents the non-zero pilot vector of Q×1; N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, P represents the length of the chu sequence s, and Q represents the adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna , q represents the index corresponding to the zero pilot between adjacent non-zero pilots in the training sequence on the transmitting antenna, ^T represents the transpose operation, and i _μ represents the training sequence corresponding to the μth transmitting antenna The index of the first non-zero element of , μ represents the index of the transmitting antenna;

c) Integer frequency offset estimation reliability criterion: $\max \underset{\mathrm{\&mu;}\&NotEqual;{\mathrm{\&mu;}}^{\&prime;}}{\min }\left\{{(i_{{\mathrm{\&mu;}}^{\&prime;}}-i_{\mathrm{\&mu;}})}_Q\right\},$ $\max \underset{1\&le;q\&le;Q-1}{\min }\left\{{(1_Q-l)}^T{l}^{\left(q\right)}\right\},$ i _μ represents the index of the first non-zero element of the training sequence corresponding to the μ-th transmitting antenna, i _μ′ represents the index of the first non-zero element of the training sequence corresponding to the μ-th transmitting antenna, and ^T represents Transpose operation, q represents the index corresponding to the zero pilot between adjacent non-zero pilots in the training sequence on the transmitting antenna, and Q represents the index corresponding to the adjacent non-zero pilot in the training sequence on the transmitting antenna spacing,

d) Simplified criterion for fractional frequency offset estimation: N _r P > Q, P represents the length of the chu sequence s, Q represents the distance between adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna, and N _r represents the receiving antenna Number of.

Based on the above-mentioned construction method of the synchronization sequence, according to the schematic diagram of the synchronization sequence structure shown in Figure 2 and the realization structure diagram of the synchronization sequence construction method shown in Figure 3, the specific implementation steps of the proposed synchronization sequence are given as follows:

1) Construct a chu sequence s of length P;

2) Perform a cyclic shift operation on s to obtain the sequence s ^(μM) ;

3) Do a discrete Fourier transform on s ^(μM) to get the sequence

4) yes Do zero-insertion operation to get the training sequence corresponding to the μ-th transmitting antenna

The above is only a preferred embodiment of the utility model, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the utility model, some improvements and modifications can also be made. Retouching should also be regarded as the protection scope of the present invention.

Claims (1)

1. a synchronous sequence construction method of MIMO system, is characterized in that: the method comprises the steps: 1) Constructing a chu sequence s whose length is P, P<N, and (N) _P =0, (N) _P means doing a modulo P operation on N, and N is the length of the synchronization sequence; 2) Perform a cyclic shift operation on the chu sequence s to obtain a shift sequence s ^(μM) , where, ( ⁾ is a cyclic shift operator, and s ^(μM) means that the chu sequence s is cyclically shifted down by μM bits The resulting vector, M is the shift parameter, μ is related to the index value of the transmitting antenna, μ+1 represents the index value of the transmitting antenna, is the rounding operator, Represents the rounding of P/N _I , the design parameter N _I makes N _t ≤ N _I <P, and N _t represents the number of transmitting antennas; 3) Perform discrete Fourier transform on the shift sequence s ^(μM) to obtain the frequency domain sequence Among them, F _P is a normalized discrete Fourier transform matrix, Q=N/P, and Q>N _t , Q represents the distance between adjacent non-zero pilots corresponding to the training sequence on the transmitting antenna, and P represents The length of the chu sequence s, N is the length of the synchronization sequence, and N _t represents the number of transmitting antennas; 4) For frequency domain sequences Do matrix operations to get the training sequence corresponding to the μ+1th transmitting antenna in, Represents the i _μth column vector of the unit matrix I _N , i _μ represents the index of the first non-zero element of the training sequence corresponding to the μ+1th transmitting antenna, μ is related to the index value of the transmitting antenna, μ+1 Represents the index value of the transmitting antenna, Q represents the distance between adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna, P represents the length of the chu sequence s, N is the length of the synchronization sequence, N _t represents the number of transmit antennas; According to the above synchronous sequence construction method, the training sequence corresponding to the μ+1th transmitting antenna The conditions that should be met are: a) Criteria for uniform distribution of transmitting antenna energy: frequency domain sequence N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, and ||·|| ² represents the 2-norm of the corresponding vector; b) Frequency offset estimation consistency criterion: (NN _t P)≥N _t P, P≥L, (1 _Q -l) ^T l ^(q) >0, Among them, L represents the maximum multipath delay of the channel, and 1 _Q represents a vector of all 1s of Q×1, Represents the non-zero pilot vector of Q×1; N is the length of the synchronization sequence, N _t represents the number of transmitting antennas, P represents the length of the chu sequence s, and Q represents the adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna , q represents the index corresponding to the zero pilot between adjacent non-zero pilots in the training sequence on the transmitting antenna, T represents the transpose operation, and i _μ represents the index corresponding to the μ+1th transmitting antenna The index of the first non-zero element of the training sequence, μ is related to the index value of the transmitting antenna, and μ+1 represents the index value of the transmitting antenna; c) Integer frequency offset estimation reliability criterion: i _μ represents the index of the first non-zero element of the training sequence corresponding to the μ+1th transmitting antenna, and i _μ′ represents the index of the first non-zero element of the training sequence corresponding to the μ’+1th transmitting antenna Index, T represents the transpose operation, q represents the index corresponding to the zero pilot between the adjacent non-zero pilots in the training sequence on the transmitting antenna, and Q represents the index corresponding to the adjacent non-zero pilots in the training sequence on the transmitting antenna the spacing between frequencies, d) Simplified criterion for fractional frequency offset estimation: N _r P > Q, P represents the length of the chu sequence s, Q represents the distance between adjacent non-zero pilots in the training sequence corresponding to the transmitting antenna, and N _r represents the receiving antenna Number of.