A method and system for analyzing OFDM signals to compute and compensate for inter-carrier interference

Abstract

A method and system for analyzing OFDM signal calculation and compensating inter-carrier interference, comprising: performing channel equalization according to ZF criterion, obtaining frequency domain estimation result of received signal, performing constellation demodulation according to ML principle, constructing intra-symbol noise model, CPE estimation and compensation, and ICI estimation and compensation; the method and system for analyzing OFDM signal calculation and compensating inter-carrier interference, the CPE estimation and compensation and the ICI estimation and compensation two modules can be opened and closed respectively, when the DUT performs radio frequency energy test, the opening and closing of the CPE estimation and compensation and the ICI estimation and compensation can reach the mode of evaluating the influence of phase noise on performance, so as to assist chip design. The difference before and after symbol demodulation is used to estimate the convolution characteristic, and compensation is performed according to the convolution characteristic. After eliminating ICI, the EVM of the signal becomes smaller, and the subcarrier constellation of the symbol is closer to the ideal modulation state.

Description

TECHNICAL FIELD

[0001] The present application relates to the signal analysis and testing technology of wireless system, in particular to a method and system for analyzing OFDM signal, calculating and compensating inter-carrier interference. BACKGROUND

[0002] The IEEE 802.11 series (a / g / n / ac / ax / be) standard uses orthogonal frequency division multiplexing (OFDM) as the modulation technology, which has the advantages of realizing parallel transmission of high-speed serial data through frequency division multiplexing to improve transmission capacity, effectively resisting frequency selective fading, interference, and supporting multi-user access. Compared with the advantages, the disadvantages are also obvious. OFDM has strict requirements for the orthogonality of subcarriers. When the orthogonality is destroyed, inter-carrier interference (ICI) will be introduced. When the phases of multiple subcarriers are consistent, the instantaneous power of the superimposed signal will be much higher than the average power, resulting in a large peak-to-average ratio (PARA) and a high power backoff requirement for the amplifier.

[0003] A spectrum analyzer is a device used to analyze the radio frequency performance of a device under test (DUT). It is a highly accurate testing device. The power, spectrum, frequency offset, and other indicators of the DUT signal can be analyzed using conventional methods, but it is difficult to evaluate the phase noise caused by the non-ideal factors of the crystal oscillator and frequency synthesizer. Moreover, the influence of phase noise on signal quality has two categories: common phase noise and inter-carrier interference, and it is unknown which category plays a major role. Therefore, it is required that the spectrum analyzer has the ability to analyze the influence of common phase noise and inter-carrier interference and suppress their influence.

[0004] The existing phase noise solution is a linear interpolation algorithm based on the cyclic prefix. The repeatability of the OFDM signal cyclic prefix and the signal part is used to reconstruct the phase noise by linear fitting to complete the corresponding compensation. However, the cyclic prefix is mainly used to eliminate the inter-symbol interference of OFDM, and is easily affected by multipath delay. Therefore, in some scenarios, it may have the opposite effect when used for estimating and compensating phase noise. SUMMARY

[0005] Therefore, it is necessary to provide a method for analyzing OFDM signal, calculating and compensating inter-carrier interference.

[0006] At the same time, a system for analyzing OFDM signal, calculating and compensating inter-carrier interference is provided.

[0007] A method for analyzing OFDM signal, calculating and compensating inter-carrier interference, comprising:

[0008] Obtaining frequency domain estimation: obtaining frequency domain estimation result of received signal according to ZF criterion, based on frequency domain receiving model Y(k)=H(k)*X(k)+W(k), wherein Y(k) is frequency domain representation of received symbol, H(k) is channel response, X(k) is frequency domain representation of transmitted symbol, W(k) is noise interference representation, and estimation of X(k) is obtained Channel response is estimated using training sequence, and estimation result is represented by H(k), and channel response H(k) of received symbol is used to estimate X(k) by ignoring noise term Wherein l p ∈1,…,K, K is total number of subcarriers, and

[0009] Constellation modulation: constellation modulation is performed according to ML principle, according to Position of constellation point and modulation mode, on the constellation corresponding to the modulation mode, the ideal constellation point Z(k) is obtained based on the characteristics of Gray mapping and the maximum likelihood principle, and the objective function is Wherein p∈S, S is a position set related to the modulation mode

[0010] Model construction: a noise-in-symbol model is constructed, and the phase noise model of the current symbol ignoring white noise is

[0011]

[0012] Let Then

[0013] J(0) is the representation of J(r-k) at r-k, is phase noise Mathematical way of affecting time domain signal, Z(k)H(k)J(0) is common phase error CPE, is inter-carrier interference ICI, is convolution operation symbol;

[0014] CPE estimation and compensation: CPE is common phase noise, and the influence on each subcarrier is the same, and the model relationship is

[0015] And the equalization relationship is After simplification, the phase part is considered, and Under the result of constellation modulation, the current common phase is The compensation mode is

[0016] ICI estimation and compensation: considering the influence of u subcarriers before and after the subcarriers of the OFDM system, after compensating the CPE part, the ICI model is simplified as Considering the convolution model of 2u+1 subcarriers, after expansion, it is Solving all subcarriers, we have

[0017]

[0018] Simplify as X CPE = Z*J, where l p ∈u,…,K-u, K is the total number of subcarriers,

[0019] The objective function for solving J is

[0020] That is, J = (Z * *Z) -1 ZX CPE ,

[0021] Then the phase noise compensation is carried out, and the compensation method is

[0022]

[0023] Where k ∈u,…,K-u.

[0024] In the preferred embodiment, in the ICI estimation and compensation, u = 3.

[0025] In the preferred embodiment, in the constellation adjustment step, when the modulation mode is QPSK, S = [±1±j], and when the modulation mode is 16QAM, S = [±1±j, ±3±3j, ±3±j, ±1±3j], j is the imaginary unit.

[0026] In the preferred embodiment, in the ICI estimation and compensation step, for subcarriers [1,…,u-1,K-u+1,…,K], u is less than the number of Wi-Fi system guard subcarriers, and no processing is performed.

[0027] In the preferred embodiment, if there are multiple symbols in the signal frame, each symbol is traversed to obtain the frequency domain estimation, constellation adjustment, model construction, CPE estimation and compensation, and ICI estimation and compensation steps.

[0028] In the preferred embodiment, it further comprises: verification analysis: inter-carrier interference compensation Based on the output of the common phase compensation Under the premise of using a comprehensive tester to analyze the DUT signal, the constellation shape and the analyzed EVM result under different modes are compared without starting compensation, starting common phase compensation Starting inter-carrier interference compensation .

[0029] In a preferred embodiment, the verification analysis step comprises starting or closing the CPE estimation and compensation and the ICI estimation and compensation respectively to evaluate which kind of phase noise has more impact on the DUT, or starting the CPE estimation and compensation and the ICI estimation and compensation simultaneously to evaluate the effect of the DUT performance after compensating the phase noise.

[0030] A system for analyzing OFDM signals to calculate and compensate inter-carrier interference, comprising:

[0031] An acquisition frequency estimation module: according to the ZF criterion, complete channel equalization, and obtain the frequency domain estimation result of the received signal, based on the frequency domain receiving model Y(k) = H(k) * X(k) + W(k), wherein Y(k) is the frequency domain representation of the received symbol, H(k) is the channel response, X(k) is the frequency domain representation of the transmitted symbol, W(k) is the noise interference representation, and the estimation of X(k) is obtained The channel response is estimated using a training sequence, and the estimation result is represented by H(k). The channel response H(k) of the received symbol is used to estimate X(k) by ignoring the noise term Wherein l p ∈1,…,K, K is the total number of subcarriers, and

[0032] A constellation diagram adjustment module: constellation diagram demodulation is performed according to the ML principle, according to The constellation point position and the modulation mode, on the constellation diagram corresponding to the modulation mode, the ideal constellation point Z(k) is obtained based on the characteristics of Gray mapping and the maximum likelihood principle, and the objective function is Where p ∈ S, S is a position set related to the modulation mode;

[0033] A model construction module: construct an intra-symbol noise model. The phase noise model of the current symbol ignoring white noise is

[0034]

[0035] Let

[0036] Then

[0037] J(0) is the representation of J(r-k) at r-k, is the phase noise The mathematical way of affecting the time domain signal is Z(k)H(k)J(0), which is the common phase error CPE, is the inter-carrier interference ICI, is the convolution operation symbol;

[0038] CPE estimation and compensation module: CPE is common phase noise, the same effect on each subcarrier, by the model relationship

[0039] and equalization relationship After simplification, consider the phase part, then Under the results of constellation adjustment, the current common phase is The compensation method is

[0040] ICI estimation and compensation module: considering the influence of u subcarriers before and after the subcarrier of OFDM system, after compensating the CPE part, the ICI model is simplified as Considering the convolution model of 2u+1 subcarriers, it is expanded as Solve all subcarriers, then

[0041]

[0042] X is written as CPE = Z*J, where l p ∈u,…,K-u, K is the total number of subcarriers,

[0043] The objective function of solving J is

[0044] That is, J=(Z * *Z) -1 ZX CPE ,

[0045] Then the phase noise compensation is carried out, and the compensation method is

[0046]

[0047] Where k∈u,…,K-u.

[0048] In the preferred embodiment, it further comprises: verification analysis module: inter-carrier interference compensation Based on the output of common phase compensation Under the premise, compare the constellation shape and the analyzed EVM result under the condition of not starting compensation, starting common phase compensation Starting inter-carrier interference compensation

[0049] ​The method and system for calculating and compensating inter-carrier interference of the OFDM signal, including CPE estimation and compensation, ICI estimation and compensation, the two modules of CPE estimation and compensation and ICI estimation and compensation can be opened and closed by a switch, when the DUT performs the radio frequency energy test, the opening and closing of CPE estimation and compensation and ICI estimation and compensation can be used to evaluate the influence of phase noise on performance, so as to assist the chip design.

[0050] Since the phase noise is random phase interference in the time domain, in the frequency domain of the OFDM system, it is a nonlinear convolution effect, and the ICI between the carriers conforms to the convolution characteristics, so the difference between the symbol demodulation before and after is used to estimate the convolution characteristics, and compensation is performed according to the convolution characteristics.

[0051] From the signal analysis result, the main indicators are the EVM (Error Vector Magnitude) change and the constellation diagram change, when the ICI is eliminated, the EVM of the signal becomes smaller, and the subcarrier constellation diagram of the symbol is closer to the ideal modulation state. In some cases where the frequency selective fading difference between the subcarriers is very large, there is indeed deviation and compensation error, and when the performance is found to be poor after compensation, it can be deduced that the frequency selective fading difference between the subcarriers of this signal is very large, that is, another problem is located. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 The constellation diagram for CPE influence;

[0053] Figure 2 The constellation diagram for ICI influence;

[0054] Figure 3 The flowchart of the method for calculating and compensating inter-carrier interference of the OFDM signal of the application;

[0055] Figure 4 The flowchart of the signal analysis process of the measuring instrument;

[0056] Figure 5 The signal analysis result diagram without compensation;

[0057] Figure 6 The signal analysis result diagram after compensating CPE;

[0058] Figure 7 The signal analysis result diagram after compensating ICI. DETAILED DESCRIPTION

[0059] The following examples facilitate a better understanding of the application, but do not limit the application.

[0060] After the synthesizer is calibrated, its phase noise is negligible compared to the phase noise of the DUT (Device Under Test). Different phase noise models are used for different oscillators. The phase noise of the DUTs used in the IEEE 802.11 series (a / g / n / ac / ax / be) standards is usually modeled using the winner model.

[0061] The present application comprises: building a model: building a model of the phase noise within one OFDM (Orthogonal Frequency Division Multiplexing) symbol, the transmitter is denoted as x(n), the receiver is denoted as y(n), the channel response is denoted as h(n), and the phase noise is denoted as White noise is denoted as ξ(n), n = 0, …, N-1, N is the number of time-domain symbol points. X(k) is the frequency-domain representation of x(n), Y(k) is the frequency-domain representation of y(n), H(k) is the frequency-domain representation of h(n), k = 0, …, N-1, N is the number of frequency-domain subcarriers, Fourier transform is used for time-domain to frequency-domain conversion, inverse Fourier transform is used for frequency-domain to time-domain conversion, and the number of time-domain symbol points N is equivalent to the number of frequency-domain subcarriers N.

[0062] The phase noise model is built in the time domain

[0063]

[0064] Then in the frequency domain, we have

[0065]

[0066]

[0067] Let

[0068]

[0069] Then

[0070]

[0071]

[0072] Generally, X(k)H(k)J(0) is the common phase error CPE, is the inter-carrier interference ICI, and W(k) is the noise part, is the convolution operation symbol.

[0073] Since the influence between carriers is considered, not only the current subcarrier k is considered, but also the affected subcarriers r. is the phase noise Mathematical way of the impact on time domain signal. W(k) is the frequency domain representation of white noise ξ(n). H(r) is the frequency domain representation of channel response of subcarrier r. X(r) is the frequency domain representation of the originating of subcarrier r.

[0074] The CPE part estimation target function is

[0075]

[0076] Take partial derivative of J(0) and let the partial derivative equal to 0, then

[0077]

[0078] Let Combine the phase noise and interference term as The relationship without considering phase noise in frequency domain is Using ZF (Zero-Force) criterion, then the solution is

[0079]

[0080] ICI part estimation, first establish the convolution model of 2u+1 related subcarriers, that is

[0081] Y(l) = X(l+u)J(-u) + … + X(l-u)J(u)

[0082] Solve all subcarriers together, then

[0083]

[0084] Simplify as Y = X*J, where P = K-2u-1, l p ∈1…,P.

[0085] The target function is

[0086]

[0087] That is

[0088] J = (X * *X) -1 XY

[0089] Phase-locked loop brings phase noise In time domain, it is approximated to rectangular window, in frequency domain, power spectral density (PSD) is approximated to sinc function, the frequency interval of 3dB bandwidth is recorded as B PLL , the subcarrier interval of OFDM system is recorded as Δf.

[0090] When B PLL < Δf / 2, CPE is dominant at this time, constellation diagram presents rotation form, such as Figure 1as shown.

[0091] When B PLL >Δf / 2, at this time, ICI is dominant, multiple independent subcarriers interfere with each other, according to the central limit theorem, the interference is normally distributed, and the constellation point is similar to white noise, and the influence presents a divergent form, as Figure 2 as shown.

[0092] As Figure 3 shown, the method for analyzing OFDM signals, calculating and compensating for inter-carrier interference according to the embodiments of the application further comprises:

[0093] Obtaining a frequency domain estimate: completing channel equalization according to a ZF (Zero-Force) criterion to obtain a frequency domain estimate result of a received signal, based on a frequency domain receiving model Y(k) = H(k) * X(k) + W(k), wherein Y(k) is a frequency domain representation of a received symbol, H(k) is a channel response, X(k) is a frequency domain representation of a transmitted symbol, and W(k) is a noise interference representation, and an estimate of X(k) is obtained The channel response is estimated using a training sequence, and the estimation result is represented by H(k). When the channel is not time-varying, the channel response of the received symbol is approximated by H(k), and the noise term is ignored to estimate X(k) wherein l p ∈1,…,K, K is the total number of subcarriers, and

[0094] Constellation diagram demodulation: performing constellation diagram demodulation according to an ML (maximum likelihood) principle, which is the inverse process of modulation, according to the constellation point position and the modulation mode, on the constellation diagram corresponding to the modulation mode, the ideal constellation point Z(k) is obtained based on the maximum likelihood principle and the characteristics of Gray mapping, and the objective function is wherein p∈S, S is a position set related to the modulation mode; when the modulation mode is QPSK (Quadrature Phase Shift Keying), S = [±1±j], and when the modulation mode is 16QAM, S = [±1±j, ±3±3j, ±3±j, ±1±3j]

[0095] j is an imaginary unit, and 16QAM is a QAM (Quadrature Amplitude Modulation) modulation mode containing 16 symbols;

[0096] Model construction: constructing an intra-symbol noise model, based on a phase noise frequency domain model derivation, the phase noise model ignoring white noise in the current symbol is

[0097] Let Then Z(k)H(k)J(0) is common phase error CPE, is inter carrier interference ICI, is convolution operator symbol, J(0) is J(r-k) at r-k, is phase noise Mathematical way of affecting time domain signal;

[0098] CPE estimation and compensation: CPE is common phase noise, the same effect on each subcarrier, by the model relationship and equalization relationship After simplification, consider the phase part, then Under the results of constellation adjustment, the current common phase is The compensation method is

[0099]

[0100] ICI estimation and compensation: the subcarrier of OFDM system is affected by the nearby subcarrier, generally, considering the influence of the front and back u subcarriers can, in actual operation, preferably, u=3 is a good choice, the calculation amount is not large and can estimate the influence as much as possible, based on CPE estimation and compensation, after compensating the CPE part, the ICI model is simplified as When only considering the convolution model of 2u+1 subcarriers, it is expanded as Together with all subcarriers, then

[0101]

[0102] X CPE =Z*J, where l p ∈u,…,K-u, K is the total number of subcarriers, the objective function for solving J is

[0103]

[0104] That is, J=(Z * *Z) -1 ZX CPE ,

[0105] Then the phase noise compensation is carried out, and the compensation method is

[0106]

[0107] Where k∈u,…,K-u.

[0108] The above frequency domain estimation, constellation adjustment, model construction, CPE estimation and compensation, ICI estimation and compensation steps are a symbol processing process, and when there are multiple symbols in a signal frame, each symbol traverses the frequency domain estimation, constellation adjustment, model construction, CPE estimation and compensation, and ICI estimation and compensation steps.

[0109] Further, in the ICI estimation and compensation step of the embodiment, for subcarriers [1,...,u-1,K-u+1,...,K], u is less than the number of WiFi system guard subcarriers, and no processing is performed.

[0110] Further, the method for analyzing an OFDM signal and compensating for inter-carrier interference of the embodiment further includes a verification analysis: inter-carrier interference compensation Output based on common phase compensation is performed, DUT signals are analyzed using a spectrum analyzer, and the constellation shapes and analyzed EVM (Error Vector Magnitude) results in three modes of no compensation, common phase compensation Inter-carrier interference compensation is started

[0111] Further, the verification analysis step of the embodiment analyzes the starting of CPE estimation and compensation and ICI estimation and compensation respectively to evaluate which type of phase noise has a greater impact on the DUT; or the starting of CPE estimation and compensation and ICI estimation and compensation at the same time to evaluate the effect of the performance of the DUT after compensating for the phase noise.

[0112] The error vector magnitude EVM (Error Vector Magnitude) is an important indicator for evaluating the quality of signal reception, and is the average vector distance between the equalization information and the demodulation information Z(k), which can be referred to as EVM RMS The present application provides two compensation contents, which are output based on common phase compensation and inter-carrier interference compensation The average vector distance between them and the demodulation information Z(k) is respectively denoted as EVM CPE and EVM ICI .

[0113] The present application is applied in the scenario of analyzing DUT signals using a spectrum analyzer, and the constellation shapes and analyzed EVM (Error Vector Magnitude) differences in three modes of no compensation, common phase compensation Inter-carrier interference compensation is started RMS , EVM CPE , and EVM ICI ​​), to evaluate the size of CPE and ICI effects. For example, when the constellation is like Figure 1 , and converges to a circle when CPE compensation is turned on, it is determined that the CPE of the DUT signal affects the reception performance, and the EVM CPE is less than the EVM RMS , and the difference is the scale of performance improvement. If the constellation is like Figure 2 , and converges to a circle when ICI compensation is turned on, it is determined that the ICI of the DUT signal affects the reception performance, and the EVM ICI is less than the EVM RMS , and the difference is the scale of performance improvement.

[0114] The present application can also be directly used for other signals based on OFDM communication, and also belongs to the protection scope of the present application. The complete process of analyzing the DUT signal is shown in Figure 4 , wherein the phase noise estimation and compensation can be newly introduced on the conventional process, and the phase noise estimation and compensation process is shown in Figure 3 . The present application takes the 11a signal as an example, and can be easily extended to 11g / n / ac / ax / be, and multiple input multiple output (MIMO) mode. The present application does not specifically describe the process other than the phase noise estimation and compensation, such as synchronization, frequency offset estimation and compensation, channel estimation and equalization, etc.

[0115] The present application provides the estimation and compensation test of the phase noise in the field of DUT radio frequency performance test, and can assist the chip development to adjust the direction of the phase noise effect.

[0116] The initial value of the carrier according to the 802.11a protocol is 312.5KHz, and the initial frequency point can be any public ISM (Industrial Scientific Medical) frequency band. The ISM (Industrial Scientific Medical) frequency band is defined by ITU-R (ITU Radiocommunication Sector, International Telecommunication Union Radiocommunication Sector). The Chinese meaning of ISM is industrial (Industrial), scientific (Scientific) and medical (Medical), so the ISM frequency band is a certain frequency band moved by each country and mainly opened to industrial, scientific and medical institutions.

[0117] The principle of the present application is that the phase noise is random phase interference in the time domain, and is a nonlinear convolution effect in the frequency domain of the OFDM system. The ICI between the carriers conforms to the convolution characteristics, so that the difference between the symbol demodulation before and after is used to estimate the convolution characteristics, and compensation is performed according to the convolution characteristics.

[0118] The verification analysis of the application is mainly from the signal analysis result, and the main indicators of the result are EVM (Error Vector Magnitude) change and constellation change. When the ICI is eliminated, the EVM of the signal becomes smaller, and the subcarrier constellation of the symbol is closer to the ideal modulation state. In some cases where the inter-subcarrier frequency selective fading difference is very large, there are indeed deviation and compensation errors. When the performance is found to be poor after compensation, it can be deduced that the inter-subcarrier frequency selective fading difference of this signal is very large, which is another problem of auxiliary positioning.

[0119] The method for analyzing OFDM signals to calculate and compensate inter-carrier interference is preferably a test method for evaluating the influence of inter-carrier interference on signal quality when analyzing 802.11a / g / n / ac / ax / be signal radio frequency quality using a spectrum analyzer.

[0120] The system for analyzing OFDM signals to calculate and compensate inter-carrier interference according to the embodiments of the application comprises:

[0121] The frequency domain estimation module: according to the ZF (Zero-Force) criterion, the channel equalization is completed, the frequency domain estimation result of the received signal is obtained, and based on the frequency domain reception model Y(k) = H(k) * X(k) + W(k), wherein Y(k) is the frequency domain representation of the received symbol, H(k) is the channel response, X(k) is the frequency domain representation of the transmitted symbol, W(k) is the noise interference, and the estimation of X(k) is obtained The channel response is estimated using a training sequence, and the estimation result is represented by H(k). When the channel time variation is not large, the channel response of the received symbol is approximated by H(k), and the noise term is ignored to estimate X(k) Wherein l p ∈1,…,K, K is the total number of subcarriers, and

[0122] The constellation modulation module: according to the ML (maximum likelihood) principle, the constellation demodulation is performed, and the demodulation is the inverse process of modulation. According to The constellation point position and the modulation mode, on the constellation graph corresponding to the modulation mode, the characteristics of Gray mapping are used, and the ideal constellation point Z(k) is obtained based on the maximum likelihood principle, and the objective function is where p∈S, S is a set of positions related to modulation mode; when the modulation mode is QPSK (Quadrature Phase Shift Keying), S=[±1±j], when the modulation mode is 16QAM, S=[±1±j,±3±3j,±3±j,±1±3j], j is an imaginary unit, and 16QAM is a QAM (Quadrature Amplitude Modulation) modulation mode containing 16 symbols;

[0123] The model construction module: a model of in-symbol noise is constructed, and a phase noise model in a frequency domain is derived, and a phase noise model ignoring white noise in a current symbol is

[0124]

[0125] Let

[0126] Then J(0) is a representation of J(r-k) at r-k, is phase noise a mathematical way of affecting a time domain signal, Z(k)H(k)J(0) is a common phase error CPE, is inter-carrier interference ICI, is a convolution operation symbol;

[0127] The CPE estimation and compensation module: CPE is a common phase noise, and the influence on each subcarrier is the same, and the model relationship is and the equalization relationship is After simplification, the phase part is considered, and there is Under the constellation adjustment result, the current common phase is The compensation method is

[0128] The ICI estimation and compensation module: the subcarriers of the OFDM system are most affected by the nearby subcarriers, generally, the influence of the front and rear u subcarriers is considered, and in actual operation, preferably, u=3 is a better choice, the calculation amount is not large, and the influence can be estimated as much as possible, based on the CPE estimation and compensation, after the CPE part is compensated, the ICI model is simplified as When only the convolution model of 2u+1 subcarriers is considered, it is expanded as All subcarriers are solved together, and there is

[0129]

[0130] XCPE=Z*J, where lp∈u,…,K-u, K is the total number of subcarriers, and the objective function of J is

[0131]

[0132] J = (Z * *Z) -1 ZX CPE ,

[0133] Then phase noise compensation is carried out, and the compensation mode is

[0134]

[0135] Wherein k belongs to u,…,K-u.

[0136] The above-mentioned frequency domain estimation module, constellation diagram mediation module, model construction module, CPE estimation and compensation module and ICI estimation and compensation module are a symbol processing process, and when there are multiple symbols in a signal frame, each symbol is connected to the frequency domain estimation module, constellation diagram mediation module, model construction module, CPE estimation and compensation module and ICI estimation and compensation module.

[0137] Further, in the ICI estimation and compensation module of the embodiment, for subcarriers [1,…,u-1,K-u+1,…,K], u is less than the number of WiFi system guard subcarriers, and the subcarriers can not be processed.

[0138] Further, the system for analyzing OFDM signals and compensating inter-carrier interference of the embodiment further comprises a verification analysis module. Output based on common phase compensation Under the premise, the DUT signal is analyzed by using a comprehensive tester, and the compensation is not started, the common phase compensation is started Start of inter-carrier interference compensation The constellation shapes and the analyzed EVM (Error Vector Magnitude) results in different modes.

[0139] Further, the verification analysis module of the embodiment further comprises: analyzing starting CPE estimation and compensation and ICI estimation and compensation respectively, so as to evaluate which type of phase noise has greater influence on the DUT; or simultaneously starting CPE estimation and compensation and ICI estimation and compensation to evaluate the effect of the DUT performance after compensating the phase noise.

[0140] The initial value of the carrier is 312.5KHz according to the 802.11a protocol, and the initial frequency point can be any public ISM frequency band.

[0141] The complete flow of analyzing the DUT signal by using the comprehensive tester is as follows Figure 4As shown in the figure, the phase noise estimation and compensation module can be newly introduced on the conventional flow. The flow of the method for analyzing OFDM signals to calculate and compensate inter-carrier interference according to the application is shown in the figure. Figure 3 The application is described by taking the 11a signal as an example, and can be easily extended to 11g / n / ac / ax / be and multiple-input multiple-output (MIMO) modes.

[0142] When analyzing OFDM-based signals using a spectrum analyzer, in order to support the evaluation of the influence of DUT phase noise, the application designs a method and system for analyzing OFDM signals to calculate and compensate inter-carrier interference, including two subsystems, CPE estimation and compensation module system and ICI estimation and compensation module system. When measuring the DUT, the subsystem switches can be opened to evaluate which type of phase noise has a greater impact on the DUT, or both subsystem switches can be opened to evaluate the performance effect of the DUT after compensating for the phase noise.

[0143] The application provides phase noise estimation and compensation testing in the field of DUT radio frequency performance testing, which can assist in adjusting the direction of phase noise influence during chip development.

[0144] The application is applied in a spectrum analyzer, in order to support the evaluation of the influence of DUT phase noise, the method and system for analyzing OFDM signals to calculate and compensate inter-carrier interference according to the application includes a common phase noise CPE estimation and compensation subsystem and an inter-carrier interference ICI estimation and compensation subsystem, and the two subsystems can be opened and closed by switches. When performing radio frequency performance testing on the DUT, the opening and closing of the subsystems can be used to evaluate the influence of phase noise on performance, thereby assisting chip design, and a radio frequency performance testing method and system for compensating phase noise are provided.

[0145] The application is applied in a spectrum analyzer to analyze DUT signals, observe constellation diagrams and EVM values. The spectrum analyzer is preferably a WT series spectrum analyzer of Shenzhen Jizhihui Instrument Technology Co., Ltd. The spectrum analyzer analyzes 802.11a / g / n / ac / ax / be type models, as shown in the embodiment shown in the figure. Figures 5 to 7 As shown in the figure, Figure 5 The analysis result obtained without compensation is shown in the figure, the constellation diagram has a certain rotation shape and divergence shape, indicating that it is affected by CPE and ICI, and the EVM value is -28.2 dB. Figure 6 The analysis result obtained by compensating CPE is shown in the figure, the rotation of the constellation diagram is controlled, and the EVM value is -33.68 dB, indicating that compensating CPE has an effect. Figure 7 The analysis result obtained by compensating ICI is shown in the figure, the divergence degree of the constellation diagram is controlled to a certain extent, and the EVM value is -34.96, indicating that compensating ICI also has an effect.

[0146] With the above ideal embodiments according to the present application as the inspiration, through the above description, relevant staff can make various changes and modifications without deviating from the scope of the technical idea of the present application. The technical scope of the present application is not limited to the content of the specification, and the technical scope must be determined according to the scope of claims.

Claims

1. A method for analyzing OFDM signals, calculating and compensating for inter-carrier interference, characterized in that, include: Obtaining Frequency Domain Estimation: Channel equalization is performed according to the ZF criterion. The received signal frequency domain estimation result is obtained. Based on the frequency domain reception model Y(k) = H(k) * X(k) + W(k), where Y(k) is the frequency domain representation of the received symbol, H(k) is the channel response, X(k) is the frequency domain representation of the transmitted symbol, and W(k) is the noise interference representation, the estimate of X(k) is obtained. The channel response is estimated using the training sequence, and the estimation result is denoted by H(k). The channel response of the received symbol is also denoted by H(k), and the noise term is ignored when estimating X(k). get Astrological chart adjustment: Astrological chart adjustment is performed based on the ML principle. Given the constellation point location and modulation scheme, the ideal constellation point Z(k) is obtained on the constellation diagram corresponding to the modulation scheme, utilizing the characteristics of the Gray mapping and based on the maximum likelihood principle. The objective function is... Where p∈S, and S is the set of positions related to the modulation method; Model Construction: Construct an intra-symbol noise model. The current intra-symbol phase noise model, ignoring white noise, is as follows: make So J(0) is the representation of J(rk) when rk. Phase noise The mathematical representation of the effect on the time-domain signal is Z(k)H(k)J(0), which represents the common phase error (CPE). Inter-carrier interference (ICI) To find the convolution operator; CPE estimation and compensation: CPE is the common phase error, which has the same impact on each subcarrier, determined by the model relationship. and equilibrium relationship After simplification, considering the phase component, we have: According to the results of the astrological chart adjustment, the current common aspect is The compensation method is ICI estimation and compensation: Considering the influence of u subcarriers before and after the subcarrier in the OFDM system, after compensating for the CPE part, the ICI model simplifies to... Consider a convolutional model with 2u+1 subcarriers, which, after unfolding, is... By combining all subcarriers, we have Abbreviated as X CPE =Z*J, where l p ∈u,…,Ku, where K is the total number of subcarriers. The objective function for solving J is: That is, J = (Z * * Z) -1 ZX CPE , Then phase noise compensation is performed, and the compensation method is as follows: Where k∈u,…,Ku.

2. The method for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to claim 1, characterized in that, In the constellation diagram adjustment step, when the modulation method is QPSK, S = [±1±j], and when the modulation method is 16QAM, S = [±1±j, ±3±3j, ±3±j, ±1±3j], where j is the imaginary unit.

3. The method for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to claim 1, characterized in that, In the ICI estimation and compensation steps, u = 3.

4. The method for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to claim 1, characterized in that, In the ICI estimation and compensation step, for subcarriers [1,…,u-1,K-u+1,…,K], u is less than the number of protected subcarriers in the Wi-Fi system, so no processing is performed.

5. The method for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to claim 1, characterized in that, If there are multiple symbols in the signal frame, each symbol is traversed to obtain frequency domain estimation, constellation diagram adjustment, model construction, CPE estimation and compensation, and ICI estimation and compensation steps.

6. The method for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to any one of claims 1 to 5, characterized in that, Also includes: Verification Analysis: Inter-carrier Interference Compensation Output based on common phase compensation Under the premise of enabling common phase compensation, the comparison is made between not enabling compensation and enabling common phase compensation. Enable inter-carrier interference compensation Constellation diagram shapes and EVM results under different modes.

7. The method for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to claim 6, characterized in that, The verification and analysis steps include: enabling or disabling CPE estimation and compensation and ICI estimation and compensation respectively to evaluate which type of phase noise has a greater impact on the DUT, or enabling CPE estimation and compensation and ICI estimation and compensation simultaneously to evaluate the effect achieved by the DTU performance after compensating for phase noise.

8. The method for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to any one of claims 1 to 5, characterized in that, The initial value of the carrier is 312.5 kHz according to the 802.11a protocol, and the initial frequency point is any ISM band.

9. A system for analyzing OFDM signals and calculating and compensating for inter-carrier interference, characterized in that, include: Frequency domain estimation module: Based on the ZF criterion, channel equalization is performed, and the frequency domain estimation result of the received signal is obtained. Based on the frequency domain reception model Y(k) = H(k) * X(k) + W(k), where Y(k) is the frequency domain representation of the received symbol, H(k) is the channel response, X(k) is the frequency domain representation of the transmitted symbol, and W(k) is the noise interference set, the estimate of X(k) is obtained. The channel response is estimated using the training sequence, and the estimation result is denoted by H(k). The channel response of the received symbol is estimated using H(k), while the noise term is ignored when estimating X(k). get Constellation Chart Mediation Module: Demodulates constellation charts based on the ML principle. Given the constellation point location and modulation scheme, the ideal constellation point Z(k) is obtained on the constellation diagram corresponding to the modulation scheme, utilizing the characteristics of the Gray mapping and based on the maximum likelihood principle. The objective function is... Where p∈S, and S is the set of positions related to the modulation method; Model building module: Constructs the intra-symbol noise model. The current intra-symbol phase noise model, ignoring white noise, is... make So J(0) is the representation of J(rk) when rk. It is phase noise The mathematical representation of the effect on the time-domain signal is Z(k)H(k)J(0), which represents the common phase error (CPE). Inter-carrier interference (ICI) To find the convolution operator; CPE estimation and compensation module: CPE stands for Common Phase Error, which has the same impact on each subcarrier, determined by the model relationship. and equilibrium relationship After simplification, considering the phase component, we have: According to the results of the astrological chart adjustment, the current common aspect is The compensation method is ICI estimation and compensation module: Considering the influence of u subcarriers before and after the subcarrier in the OFDM system, after compensating for the CPE part, the ICI model is simplified to... Consider a convolutional model with 2u+1 subcarriers, which, after unfolding, is... By combining all subcarriers, we have Abbreviated as X CPE =Z*J, where l p ∈u,…,Ku, where K is the total number of subcarriers. The objective function for solving J is: That is, J = (Z * * Z) -1 ZX CPE , Then phase noise compensation is performed, and the compensation method is as follows: Where k∈u,…,Ku.

10. The system for analyzing OFDM signals and calculating and compensating for inter-carrier interference according to claim 9, characterized in that, Also includes: Verification and Analysis Module: Inter-carrier Interference Compensation Output based on common phase compensation Under the premise of enabling common phase compensation, the comparison is made between not enabling compensation and enabling common phase compensation. Enable inter-carrier interference compensation Constellation diagram shapes and EVM results under different modes.

Images