CN100493061C - A Signal Detection Method Based on Orthogonal Packet MC-CDMA Downlink Combined with Frequency Offset Compensation - Google Patents

Abstract

The method includes steps: sending the received signal to CP removal module to remove circle prefix; next, sending signal to be processed to S/P modular converter to convert the signal to parallel digital signals, which are sent to module of multiply operation and module of estimating frequency deviation; carrying out compensation for frequency deviation by module of multiply operation, then sent to FFT module to carry out demodulation, and sent to module of demultiplexing group to carry out de-grouping, sent to time-frequency transformation module and de-spread module; signal from time-frequency transformation module is sent to module of estimating frequency deviation so as to obtain estimated value, then sent to compensation module to obtain frequency deviation compensation matrix, which is sent to module of multiply operation; after de-spread in de-spread module, merging and detecting module detects out user signal. The invention raises detecting performance evidently.

Description

A Signal Detection Method Based on Orthogonal Packet MC-CDMA Downlink Combined with Frequency Offset Compensation

technical field

The invention belongs to the technical field of multi-carrier modulation adopted by a code division multiple access CDMA mobile communication system.

Background technique

CDMA mobile communication technology shows great development potential due to its simple frequency planning, large system capacity, strong anti-multipath ability, good communication quality, and low electromagnetic interference. It is one of the mainstream technologies of future mobile communication. Broadband transmission will be used in mobile communication in the future, but single-carrier CDMA technology is difficult to be extended to wideband transmission directly, and a multicarrier (MC) parallel transmission system must be adopted. There are three main forms of combining multi-carrier technology with CDMA technology: multi-carrier CDMA (MC-CDMA), multi-carrier direct-sequence spread spectrum CDMA (multicarrier DS-CDMA) and multi-tone modulation CDMA (MT-CDMA). These three schemes can realize transmission and reception conveniently through the Fourier transform pair (inverse Fourier transform/Fourier transform, IFFT/FFF) without increasing the complexity of the transmitter and receiver, and have High spectrum utilization. Among them, the MC-CDMA scheme is considered to be the most promising scheme among the three schemes because it can adopt frequency domain diversity and excellent system performance, and it is also one of the most competitive schemes in the future mobile communication system.

Orthogonal packet MC-CDMA scheme is an optimized scheme of traditional MC-CDMA scheme. Orthogonal packet MC-CDMA divides the subcarriers into mutually orthogonal subcarrier groups, and then assigns the users at the transmitting end to the corresponding subcarrier groups. Reasonable selection of the number of carriers in each group can achieve ideal frequency diversity for each user.

However, MC-CDMA schemes including orthogonal grouping MC-CDMA schemes are very sensitive to carrier frequency offset (hereinafter referred to as carrier frequency offset or frequency offset). The orthogonality among them is destroyed, resulting in inter-carrier interference (ICI), which makes the system performance drop sharply. Therefore, effectively estimating and compensating the carrier frequency offset of MC-CDMA schemes is one of the key technologies to effectively implement MC-CDMA schemes.

In recent years, some carrier frequency offset estimation and compensation technologies that can be used in MC-CDMA schemes have been proposed. In summary, there are mainly data-aided or pilot-based estimation methods and non-data-aided ( non-data-aided) estimation, that is, blind estimation method. Based on the data-assisted method, the frequency resources are wasted due to the insertion of pilot symbols. Blind estimation methods mainly include cyclic prefix (cyclic prefix, CP) estimation method and virtual subcarrier (virtual subcarrier, VSC) estimation method. The estimation method using the cyclic prefix is low in estimation accuracy because the cyclic prefix is used to carry interference and protect. The virtual subcarrier estimation method uses the orthogonality between the virtual subcarrier and the carrier used to transmit data to estimate the carrier frequency offset. This estimation method needs to allocate additional subcarriers that do not transmit any data, thus greatly reducing the system performance. Frequency Band Utilization.

Contents of the invention

The technical problem to be solved by the present invention is to propose a signal detection method combining orthogonal packet MC-CDMA downlink with frequency offset compensation. This method can effectively estimate the carrier frequency offset of the orthogonal packet MC-CDMA downlink at first, and effectively compensate the frequency offset by forming a frequency offset compensation matrix based on the estimated value of the frequency offset, and then based on the frequency offset compensation The signal detects the user signal to improve the detection performance.

The technical scheme adopted by the present invention to solve the above-mentioned technical problems is first to remove the cyclic prefix of the digital signal of the down-converted baseband received by the antenna after the analog-to-digital A/D conversion, and then perform serial-to-parallel conversion on the signal from which the cyclic prefix has been removed, The serial digital signal is converted into a parallel digital signal corresponding to the parallel-serial conversion of the sending end, and then multiplied by the frequency offset compensation matrix to realize the compensation of the frequency offset, and the obtained signal is subjected to FFT operation to complete the multiplication of the received data Carrier demodulation, ungrouping the demodulated data through the ungrouping matrix, eliminating the signals of other groups of users, and obtaining the signals of the desired group of users, and then despreading the ungrouped input signals using the spreading sequence of the desired users The processing is to recover the signal of the desired user on each sub-carrier, and finally use the combination coefficient matrix of the desired user to combine the signals carried by each sub-carrier to detect the signal of the desired user.

A kind of orthogonal grouping MC-CDMA downlink combines the signal detection method of frequency offset compensation, comprises the following steps:

a. First, the digital signal received by the antenna is down-converted to the baseband after analog-to-digital A/D conversion and sent to the CP removal module (A101), the cyclic prefix is removed, and the signal with the cyclic prefix removed is sent to the S/P conversion module ( In A102), the serial digital signal is converted into a parallel digital signal r(i) corresponding to the P/S conversion at the sending end, and then r(i) is sent to the multiplication module (A103) and the frequency offset estimation module (C101) respectively In the multiplication module (A103), r(i) is multiplied with the frequency offset compensation matrix χ ^H output by the frequency offset compensation module (C102) to realize the compensation of the frequency offset, and the obtained signal is sent to the FFT In the processing module (A104), the FFT operation is carried out to the input signal through the FFT operation matrix F _M to complete the multi-carrier demodulation of the received data, and the obtained signal F _M r (i) is provided to the unpacking multiplexing module (B101);

Among them, the frequency offset compensation matrix ${\mathrm{\χ}}^h=\mathrm{diag}{(1,e^{j2\mathrm{\π}\widehat{\mathrm{\ϵ}}/m},\·\&Center\; Dot;\·,e^{j2\mathrm{\π}\widehat{\mathrm{\ϵ}}(m-1)/m})}^h,$ M × M-dimensional FFT operation matrix [F _M ] _{m, n} = M ^-1/2 exp(-j2π(m-1)(n-1)/M), (·) ^H represents the conjugate transpose operation, M is the total number of subcarriers, M subcarriers are divided into G groups, the number of subcarriers in each group is N=M/G, and M, G and N are all integers;

b. In the unpacking and multiplexing module (B101), the demodulated data is unpacked through the unpacking matrix q ¹ , and the signals of other groups of users are eliminated, and the signals of the desired group of users q ¹ F _M r(i ), then the signals after degrouping and multiplexing are sent into the time-frequency conversion module (B102) and the despreading module (D101) respectively;

Among them, q ¹ is the ungrouping matrix of the first group, that is, the superscript 1 indicates the first group,

将输入的频域信号变换为时域信号

之后将时频变换后的信号送到载波频偏估计模块(C101)中；c. In the time-frequency transformation module (B102), through the time-frequency transformation matrix

Transform the input frequency domain signal into a time domain signal

Afterwards, the signal after the time-frequency transformation is sent to the carrier frequency offset estimation module (C101);

与解组矩阵q¹以及FFT运算矩阵F_M之间满足 $q^1\left(F_M{F}_1^{*}\right)=I_N$ 的约束关系，I_N是N×N维的单位矩阵，(·)^*表示共轭运算，上、下标1表示第一组；Among them, the time-frequency transformation matrix

Satisfies with the ungrouping matrix q ¹ and the FFT operation matrix F _M $q^1\left({\mathrm{<figure-callout\; id="1"\; label="f\; m\; f"\; filenames="C200510011823E00141.png"\; state="\{\{state\}\}">f</figure-callout>}}_m{f}_{}^{}{}_1^{*}\right)=I_N$ The constraint relation of , I _N is the identity matrix of N×N dimensions, ( ) ^* represents the conjugate operation, and the superscript and subscript 1 represent the first group;

送入频偏补偿模块(C102)中；d. In the frequency offset estimation module (C101), use the signal r(i) input by the module (A102) and the signal F ₁ ^* q ¹ F _M r(i) input by the time-frequency transformation module (B102), by calculating the cost function ${\widehat{\mathrm{\&epsiv;}}}_{\mathrm{opt}}=\arg \underset{\widehat{\mathrm{\&epsiv;}}}{\min }{\left|q^1{f}_m{\mathrm{\&chi;}}^h(r\left(i\right)-{\mathrm{<figure-callout\; id="1"\; label="f"\; filenames="C200510011823E00141.png"\; state="\{\{state\}\}">f</figure-callout>}}_1^{*}{q}^1{f}_{m}r\left(i\right))\right|}^2$ The optimal calculation of the carrier frequency offset is obtained Then the estimated value of the carrier frequency offset

Send it into the frequency offset compensation module (C102);

Among them, (·) ^* represents a conjugate operation, (·) ^H represents a conjugate transpose operation, and ‖·‖ ₂ is a 2-norm operation;

代入频偏补偿矩阵 ${\mathrm{\&chi;}}^H=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/M},L,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(M-1)/M})}^H$ 中，得到频偏补偿矩阵，然后将频偏补偿矩阵送入乘法运算模块(A103)，使频偏补偿矩阵与模块(A102)的输入信号r(i)相乘，实现对载波频偏的补偿；e. In the frequency offset compensation module (C102), the estimated value of the carrier frequency offset

Substitute into the frequency offset compensation matrix ${\mathrm{\&chi;}}^h=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/m},L,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(m-1)/m})}^h$ , obtain the frequency offset compensation matrix, and then send the frequency offset compensation matrix to the multiplication module (A103), so that the frequency offset compensation matrix is multiplied with the input signal r(i) of the module (A102) to realize the compensation of the carrier frequency offset ;

Among them, (·) ^H represents the conjugate transpose operation;

，之后送入合并检测模块(D102)中；f. In the despreading module (D101), despread the signal q ¹ F _M r(i) input by the degrouping and multiplexing module (B101) using the spreading sequence c ₁ of the desired user, and perform despreading processing on each subcarrier Recover the desired user's signal

, then sent into the combined detection module (D102);

Wherein, (·) ^T represents transpose operation;

；g. In the combination detection module (D102), use the combination coefficient matrix α ₁ =diag[α _1,1 , α _1,2 L, α _1,N ] of the desired user to combine the signals carried by each subcarrier, and detect Expect user signals

;

Among them, (·) ^T represents the transpose operation.

Beneficial effects of the present invention:

By exploiting the orthogonal grouping property of the orthogonal grouping MC-CDMA scheme, that is, the subcarriers between each grouping are mutually orthogonal when there is no frequency offset, and by minimizing the value of the proposed cost function in the presence of frequency offset, the Efficient estimation of frequency offset. This method does not need auxiliary means such as pilot symbols or virtual subcarriers to estimate the frequency offset, thus improving the frequency band utilization of the system; the estimation of the frequency offset does not use the cyclic prefix used to carry interference and protect, ensuring The accuracy of the estimation is improved; the frequency offset estimation process only uses the received signal and the signal generated by the ungrouping and multiplexing during the normal detection process of the user to the data, so the estimation method is simple and practical.

The proposed frequency offset compensation method only uses the estimation result of the frequency offset to form the corresponding compensation matrix, and then multiplies the received signal through the multiplication module to realize the effective compensation of the carrier frequency offset without changing the mobile terminal The design of the voltage-controlled oscillator (VCO), and the usual frequency offset compensation is mostly realized by changing the carrier oscillation frequency of the VCO.

The detection of the desired user signal is realized based on the frequency offset compensated signal, thus obviously improving the performance of user signal detection.

Description of drawings

Fig. 1 is the structural block diagram of the transmitting unit of the orthogonal grouping MC-CDMA base station;

Fig. 2 is a structural block diagram of a mobile terminal with carrier frequency offset estimation and compensation functions;

Figure 3 shows the simulation results of MSE estimation performance versus E _b /N ₀ when the number of users is different;

Fig. 4 is the simulation result of MSE estimation performance to E _b /N ₀ when the number of groups is different;

Fig. 5 is the result of the user's signal detection bit error rate (BER) versus E _b /N ₀ after the system adopts the frequency offset compensation method proposed by the present invention.

Detailed ways

The method of the present invention will be discussed in detail below in conjunction with the accompanying drawings.

The method of the invention is applicable to any mobile communication system using orthogonal grouping MC-CDMA technology.

1. Orthogonal packet MC-CDMA system

Orthogonal packet MC-CDMA systems are considered in this section. Fig. 1 is a structural block diagram of a base station transmitting unit of an orthogonal grouping MC-CDMA system, and Fig. 2 is a structural block diagram of a mobile terminal of an orthogonal grouping MC-CDMA system with carrier frequency offset estimation and compensation functions.

In this system, it is assumed that there are M subcarriers in total, the system bandwidth is W, and the chip period of the spreading sequence is T _c =1/W, that is, the symbol period T=MT _c . The system bandwidth W is shared by M subcarriers at equal intervals, and the interval between each adjacent subcarrier is 1/T. In the design, M subcarriers are divided into G groups, the number of subcarriers in each group is N=M/G, and M, G and N are all integers. The system divides the sub-carriers into mutually orthogonal sub-carrier groups, and then assigns the users at the transmitting end to the corresponding sub-carrier groups in turn, and the users assigned to the same group are distinguished by mutually orthogonal Walsh spreading codes.

2. The base station transmits the signal

表示第n个数字子载波，其中，(·)^*表示共轭运算。定义M×N维的时频变换矩阵的列由第g组的N个数字子载波组成，这样

就可以表示为：As shown in Fig. 1, it is assumed that d _g,k (i) is the i-th transmitted symbol of the k-th user in the g-th group. The N×1-dimensional spreading sequence c _k =[c _{k, 1} , K, c _{k, N} ] ^T of the kth user is used to spread d _{g, k} (i) to N subcarriers. The definition matrix [F _M ] _{m, n} = M ^-1/2 exp(-j2π(m-1)(n-1)/M) represents an M×M dimensional FFT operation matrix. If f _n denotes the nth column of F _M , then

Indicates the nth digital subcarrier, where (·) ^* represents the conjugate operation. Define an M×N-dimensional time-frequency transformation matrix The column of is composed of N digital subcarriers of group g, such that

It can be expressed as:

$f_g^{*}=[f_g^{*},f_{G+g}^{*},f_{2G+g}^{*},...,f_{(N-1)G+g}^{*}]$ [Formula 1]

Without loss of generality, it can be assumed that each group has K active users, so the i-th M×1-dimensional transmission signal of the gth group can be expressed as:

${\mathrm{the\; s}}_g\left(i\right)=f_g^{*}{C}_g{d}_g\left(i\right)=f_g^{*}{a}_g\left(i\right)$ [Formula 2]

In the formula, d _g (i)=[d _{g, 1} (i), d _{g, 2} (i), L, d _{g, K} (i)] ^T is the K×1 dimension of the i-th symbol of the g-th group Data vector, C _g =[c ₁ , K, c _K ] is a matrix of N×K dimensions, each column of which is composed of N spreading codes of group g, a _g (i)=C _g d _g (i) represents the N×1-dimensional transmission data sequence after group g spreading, where (·) ^T represents a transpose operation.

In this way, parallel-to-serial P/S conversion is performed on s _g (i), and after CP is added to each data block, the signal is transmitted to the wireless channel for transmission.

3. Channel

For the considered downlink transmission, without loss of generality, the channel is assumed to be a frequency selective fading channel. Assuming that the symbol period of the transmission signal is greater than the delay spread of the channel, that is, the transmission channel of each subcarrier can be considered as an independent flat fading channel. In this way, the channel response of the nth subcarrier of the gth group in the frequency domain can be expressed as:

$h_g\left(\mathrm{no}\right)={\mathrm{\&beta;}}_{g,\mathrm{no}}{e}^{{\mathrm{j\&phi;}}_{g,\mathrm{no}}},$ g=1, L, G; n=1, L, N [Formula 3]

Among them, β _{g, n} and φ _{g, n} represent the magnitude and phase response of the fading channel respectively, and can be assumed to be independent and identically distributed random variables and uniformly independent and identically distributed random variables on [0, 2π), respectively.

4. Receive signal

When the signal transmitted by the base station goes through the downlink fading channel described by formula 3, it reaches the mobile terminal. In the mobile terminal, first remove the digital signal cyclic prefix CP from the down-converted baseband received by the antenna after analog-to-digital A/D conversion, and then perform serial-to-parallel S/P conversion corresponding to the P/S conversion at the transmitting end, and convert the serial The digital signal is converted into a parallel digital signal, which can be expressed as:

                           [公式4]

[Formula 4]

Wherein, η(i) is an M×1-dimensional noise vector, and it is assumed that its one-sided power spectral density is N ₀ .

The other matrices in Equation 4 can be expressed as:

是由频偏引起的频偏矩阵，ε(-0.5≤ε≤0.5)为相对频偏，被定义为CFO/Δf，CFO为载波频偏的绝对值，Δf为相邻载波的频率间隔。in,

It is the frequency offset matrix caused by frequency offset, ε(-0.5≤ε≤0.5) is the relative frequency offset, which is defined as CFO/Δf, CFO is the absolute value of carrier frequency offset, and Δf is the frequency interval of adjacent carriers.

5. Signal detection method combined with frequency offset compensation

A. Orthogonal Packet MC-CDMA Downlink Signal Detection Method Combined with Frequency Offset Compensation

The block diagram of the mobile terminal structure of the orthogonal packet MC-CDMA mobile communication system with carrier frequency offset estimation and compensation functions is shown in Figure 2, including: demodulator A, demultiplexer and time-frequency converter B, carrier frequency offset estimation and compensator C. Merge and detector D.

Demodulator A includes: CP removal module A101, S/P serial-to-parallel conversion module A102, multiplication operation module A103 and FFT processing module A104;

Degroup multiplexing and time-frequency converter B include: degroup multiplexing module B101 and time-frequency conversion module B102;

The carrier frequency offset estimation and compensator C includes: a frequency offset estimation module C101 and a frequency offset compensation module C102;

The combination and detector D includes: a despreading module D101 and a combination and detection module D102.

A kind of orthogonal packet MC-CDMA downlink combines the signal detection method of frequency offset compensation comprising the following steps:

a. First, the digital signal of the down-converted baseband received by the antenna and converted by analog-to-digital A/D is sent to the CP removal module A101, and the cyclic prefix is removed, and the signal with the cyclic prefix removed is sent to the S/P conversion module (A102) The serial digital signal is converted into a parallel digital signal r(i) corresponding to the P/S conversion of the sending end, and then r(i) is sent to the multiplication module A103 and the frequency offset estimation module C101 respectively. In the multiplication module A103, r (i) is multiplied with the frequency offset compensation matrix χ ^H input by the frequency offset compensation module C102 to realize the compensation of the frequency offset, and the obtained signal is sent into the FFT processing module A104, through The FFT operation matrix F _M performs FFT operation on the input signal, completes the multi-carrier demodulation of the received data, and provides the obtained signal F _M r(i) to the degrouping and multiplexing module B101;

b. In the degrouping and multiplexing module B101, the demodulated data is degrouped through the degrouping matrix q ¹ to eliminate the signals of other groups of users, and obtain the signals q ¹ F _M r(i) of the desired group of users, Then the degrouped and multiplexed signals are sent to the time-frequency conversion module B102 and the despreading module D101 respectively;

将输入的频域信号变换为时域信号

之后将时频变换后的信号送到载波频偏估计模块C101中；c. In the time-frequency transformation module B102, through the time-frequency transformation matrix

Transform the input frequency domain signal into a time domain signal

Afterwards, the signal after the time-frequency transformation is sent to the carrier frequency offset estimation module C101;

通过公式11对代价函数的寻优运算，得到载波频偏的估计值

之后将载波频偏的估计值

送入频偏补偿模块C102中；d. In the frequency offset estimation module C101, use the signal r(i) input by the module A102 and the signal input by the time-frequency conversion module B102

The estimated value of the carrier frequency offset is obtained through the optimization operation of the cost function in formula 11

Then the estimated value of the carrier frequency offset

Send it into the frequency offset compensation module C102;

代入频偏补偿矩阵 ${\mathrm{\&chi;}}^H=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/M},\&CenterDot;\&CenterDot;\&CenterDot;,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(M-1)/M})}^H$ 中，得到频偏补偿矩阵，然后将频偏补偿矩阵送入乘法运算模块A103，使频偏补偿矩阵与模块A102的输入信号r(i)相乘，实现对载波频偏的补偿；e. In the frequency offset compensation module C102, the estimated value of the carrier frequency offset

Substitute into the frequency offset compensation matrix ${\mathrm{\&chi;}}^h=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/m},\&Center\; Dot;\&CenterDot;\&CenterDot;,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(m-1)/m})}^h$ , obtain the frequency offset compensation matrix, and then send the frequency offset compensation matrix to the multiplication module A103, so that the frequency offset compensation matrix is multiplied with the input signal r (i) of the module A102 to realize the compensation of the carrier frequency offset;

，之后送入合并检测模块D102中；f. In the despreading module D101, despread the signal q ¹ F _M r(i) input by the degrouping and multiplexing module B101 using the spreading sequence c ₁ of the desired user, and restore the desired user on each subcarrier signal of

, then sent to the merger detection module D102;

g. In the combination detection module D102, use the combination coefficient matrix α ₁ =diag[α _1,1 , α _1,2 ..., α _1,N ] of the desired user to combine the signals carried by each subcarrier to detect the desired user signal of

The frequency offset estimation and compensation method involved in the above steps will be described below.

In the above steps, without loss of generality, it is assumed that the signal of the first user in the first group is the desired signal, and the i-th bit of the decision variable v _1,1 can be expressed as:

${\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C200510011823E00141.png"\; state="\{\{state\}\}">v</figure-callout>}}_{\mathrm{1,1}}\left(i\right)=c_1^T{\mathrm{\&alpha;}}_1{q}^1{f}_{m}r\left(i\right)$ [Formula 5]

Wherein, the combination coefficient matrix is α ₁ =diag[α _1,1 , α _1,2 . . . , α _1,N ].

^q1 is the ungrouping matrix of group 1 defined as,

             [公式6]

[Formula 6]

In the case of no frequency offset (ε=0), since different groups of carriers are orthogonal, the following formula holds,

$q^1\left({\mathrm{<figure-callout\; id="1"\; label="f\; m\; f"\; filenames="C200510011823E00141.png"\; state="\{\{state\}\}">f</figure-callout>}}_m{f}_{}^{}{}_1^{*}\right)=I_N$ [Formula 7]

$q^1\left(f_m{f}_g^{*}\right)=0_N,g=2,\&CenterDot;\&CenterDot;\&CenterDot;,G$ [Formula 8]

Wherein, I _N is an N×N-dimensional identity matrix, and 0 _N is an N×N-dimensional zero matrix.

In the case of frequency offset (ε≠0), since the orthogonality between different groups of carriers is destroyed, formula 8 will no longer hold, that is,

               [公式9]

[Formula 9]

But if the carrier frequency offset can be estimated effectively, according to formula 8, there will be

              [公式10]

[Formula 10]

In the formula ${\mathrm{\&chi;}}^h=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/m},\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(m-1)/m})}^h,$ is the estimated value of the carrier frequency offset, x ^H is defined as the frequency offset compensation matrix, where (·) ^H represents the conjugate transpose operation.

Equation 8 and Equation 10 show that in the case that there is no carrier frequency offset or carrier frequency offset can be effectively estimated and compensated, the ungrouped interference signal matrix will be a 0 _N matrix, so far the orthogonality is derived from the mathematical relationship This remarkable characteristic of the downlink of packet MC-CDMA system, and based on this, an estimation method of carrier frequency offset is proposed.

B. Carrier Frequency Offset Estimation

The interference signal can be obtained by subtracting the signal of the desired group from r(i). After that, FFT processing is performed on the interference signal, and then q ¹ is used to ungroup the interference signal. As mentioned above, in the case of no frequency offset (ε=0), the matrix after ungrouping the interference signal will be 0 _N . In the case of frequency offset (ε=0), the estimated value of carrier frequency offset can be obtained by solving the following optimization problem of the cost function

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}}}_{\mathrm{<span\; class="notranslate">\; opt</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}}}{\mathrm{<span\; class="notranslate">\; min</span>}}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; q</span>}}^{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&chi;</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="f"\; filenames="C200510011823E00141.png"\; state="\{\{state\}\}">f</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; *</span>}{\mathrm{<span\; class="notranslate">\; q</span>}}^{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 11</span>}<span\; class="notranslate">\; )</span>$

In the formula, ‖·‖ ₂ is a 2-norm operation.

It can be seen from formula 11 that the proposed estimation algorithm is realized only by using the FFT transform characteristic of the orthogonal packet MC-CDMA system. In addition, according to formula 8 and formula 10, under the condition of no noise, formula 11 is an unbiased estimate. Therefore, the proposed estimation method has the significant advantage of high estimation accuracy.

C. Carrier frequency offset compensation

后，将载波频偏的估计值

送入前面所定义的频偏补偿矩阵 ${\mathrm{\&chi;}}^H=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/M},L,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(M-1)/M})}^H$ 中，得到频偏补偿矩阵，然后将频偏补偿矩阵送入如图2所示的乘法模块，使补偿矩阵与接收信号相乘，实现对载波频偏的补偿。After estimating the carrier frequency offset

After that, the estimated value of the carrier frequency offset

Input the frequency offset compensation matrix defined above ${\mathrm{\&chi;}}^h=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/m},L,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(m-1)/m})}^h$ In , the frequency offset compensation matrix is obtained, and then the frequency offset compensation matrix is sent to the multiplication module shown in Figure 2, so that the compensation matrix is multiplied by the received signal to realize the compensation of the carrier frequency offset.

It can be seen from the above steps that the effective estimation and compensation of the frequency offset of the downlink carrier of the orthogonal packet MC-CDMA system and the detection of the expected user signal are realized by using the method proposed by the present invention.

Fig. 3-Fig. 5 show the signal detection performance of estimating the carrier frequency offset of the orthogonal packet MC-CDMA downlink and combining the frequency offset compensation by adopting the method proposed by the present invention. The effects of different numbers of users and different groups on the performance of frequency offset estimation and the effect of using the frequency offset compensation method proposed by the present invention are discussed respectively. In the simulation, it is assumed that the orthogonal packet MC-CDMA system adopts BPSK modulation, the number of subcarriers is 128, and the spreading code uses the same Walsh Hadamard code as the number of subcarriers in each group as the spreading code, and the relative frequency offset ε of the carrier is assumed to be 0.1.

In order to evaluate the estimated performance, 100 independent simulation results were statistically averaged, and the minimum mean square error (MSE) was used as the evaluation index, which was defined as:

$\mathrm{MSE}=\frac{1}{u}{\mathrm{\&Sigma;}}_{u=1}^u{|{\widehat{\mathrm{\&epsiv;}}}_u-\mathrm{\&epsiv;}|}^2$ [Formula 12]

为每次独立估计的相对频偏值，|·|为求绝对值运算。where U is the number of independent trials,

is the relative frequency offset value estimated independently each time, |·| is the absolute value operation.

Fig. 3 shows the simulation results of the MSE estimation performance versus the normalized signal-to-noise ratio E _b /N ₀ under the conditions of N=32 and G=4. Fig. 3 shows that the frequency offset of the downlink carrier of the system is effectively estimated by using the estimation method of the present invention. Figure 3 also shows that the estimated performance is improved when E _b /N ₀ increases; when the number of users is small, the estimated performance is improved due to the reduction of multiple access interference.

Fig. 4 shows the simulation results of the MSE estimation performance on the normalized signal-to-noise ratio E _b /N ₀ under the condition that the number of users K=16 and when the number of groups is different. FIG. 4 also shows that the frequency offset of the system downlink carrier is effectively estimated by using the estimation method of the present invention. Figure 4 also shows that the performance of the method of the present invention is affected by the number of groups, that is, the larger the number of groups, the better the estimated performance.

Fig. 5 is the result of the user's signal detection bit error rate (BER) versus E _b /N ₀ after the system adopts the frequency offset compensation method proposed by the present invention. The maximum ratio combining (MRC) and equal gain combining (EGC) methods are used in the simulation respectively. The simulation conditions are: N=32, G=4 and K=16. It can be seen from Fig. 5 that after adopting the frequency offset compensation method proposed by the present invention, the user's BER performance for signal detection is greatly improved compared with when no frequency offset compensation measure is used, and is close to the performance when fully compensated , to achieve the effect of compensation.

Claims (1)

1. a kind of orthogonal packet MC-CDMA downlink combines the signal detection method of frequency offset compensation, it is characterized in that comprising the following steps: a. First, the digital signal received by the antenna is down-converted to the baseband after analog-to-digital A/D conversion and sent to the CP removal module (A101), the cyclic prefix is removed, and the signal with the cyclic prefix removed is sent to the S/P conversion module ( In A102), the serial digital signal is converted into a parallel digital signal r(i) corresponding to the P/S conversion at the sending end, and then r(i) is sent to the multiplication module (A103) and the frequency offset estimation module (C101) respectively In the multiplication module (A103), r(i) is multiplied with the frequency offset compensation matrix χ ^H output by the frequency offset compensation module (C102) to realize the compensation of the frequency offset, and the obtained signal is sent to the FFT In the processing module (A104), the FFT operation is carried out to the input signal through the FFT operation matrix F _M to complete the multi-carrier demodulation of the received data, and the obtained signal F _M r (i) is provided to the unpacking multiplexing module (B101); Among them, the frequency offset compensation matrix ${\mathrm{\&chi;}}^h=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/m},\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(m-1)/m})}^h,$ M × M-dimensional FFT operation matrix [F _M ] _{m, n} = M ^-1/2 exp(-j2π(m-1)(n-1)/M), (·) ^H represents the conjugate transpose operation, M is the total number of subcarriers, M subcarriers are divided into G groups, the number of subcarriers in each group is N=M/G, M, G and N are all integers,

is the estimated value of carrier frequency offset; b. In the unpacking and multiplexing module (B101), the demodulated data is unpacked through the unpacking matrix q ¹ , and the signals of other groups of users are eliminated, and the signals of the desired group of users q ¹ F _M r(i ), then the signals after degrouping and multiplexing are sent into the time-frequency conversion module (B102) and the despreading module (D101) respectively; Among them, q ¹ is the ungrouping matrix of the first group, that is, the superscript 1 indicates the first group, c. In the time-frequency transformation module (B102), through the time-frequency transformation matrix

Transform the input frequency domain signal into a time domain signal Afterwards, the signal after the time-frequency transformation is sent to the carrier frequency offset estimation module (C101); Among them, the time-frequency transformation matrix

Satisfies with the ungrouping matrix q ¹ and the FFT operation matrix F _M $q^1\left(f_m{f}_1^{*}\right)=I_N$ The constraint relation of , I _N is the identity matrix of N×N dimensions, ( ) ^* represents the conjugate operation, and the superscript and subscript 1 represent the first group; d. In the frequency offset estimation module (C101), use the signal r(i) output by the module (A102) and the signal output by the time-frequency transformation module (B102)

through the cost function ${\widehat{\mathrm{\&epsiv;}}}_{\mathrm{opt}}=\arg \underset{\widehat{\mathrm{\&epsiv;}}}{\min }{\left|\right|q^1{f}_m{\mathrm{\&chi;}}^h(r\left(i\right)-f_1^{*}{q}^1{f}_{m}r\left(i\right))\left|\right|}_2$ The optimal calculation of the carrier frequency offset is obtained

Then the estimated value of the carrier frequency offset Send it into the frequency offset compensation module (C102); Among them, (·) ^* represents a conjugate operation, (·) ^H represents a conjugate transpose operation, and ‖·‖ ₂ is a 2-norm operation; e. In the frequency offset compensation module (C102), the estimated value of the carrier frequency offset

Substitute into the frequency offset compensation matrix ${\mathrm{\&chi;}}^h=\mathrm{diag}{(1,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}/m},\&CenterDot;\&CenterDot;\&CenterDot;,e^{j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}(m-1)/m})}^h$ , obtain the frequency offset compensation matrix, and then send the frequency offset compensation matrix to the multiplication module (A103), so that the frequency offset compensation matrix is multiplied by the output signal r(i) of the module (A102) to realize the compensation of the carrier frequency offset ; Among them, (·) ^H represents the conjugate transpose operation; f. In the despreading module (D101), despread the signal q ¹ F _M r(i) output by the degrouping and multiplexing module (B101) using the spreading sequence c ₁ of the desired user, and perform despreading processing on each subcarrier Recover the desired user's signal

Send in the merge detection module (D102) afterwards; Wherein, (·) ^T represents transpose operation; g. In the combination detection module (D102), combine the signals carried by each subcarrier by using the combination coefficient matrix α ₁ =diag[α _1,1 , α _1,2 ..., α _1,N ] of the desired user, and detect Expect user signals

Among them, (·) ^T represents the transpose operation, α _{1, N} where N is the number of subcarriers in each group.

Images