A nonlinear amplified three-dimensional signal modulation and demodulation method and device

Abstract

The present invention discloses a nonlinear amplified three-dimensional signal modulation and demodulation method and device. The method uses a wireless transmission system to implement three-dimensional modulation with Stokes constraints. The two signals generated by the three-dimensional modulation at the transmitting end are independently orthogonalized through a power amplifier and an antenna working in nonlinear amplification for signal transmission. The receiving end obtains a received signal with nonlinear distortion through an antenna with orthogonal polarization or a single antenna. The nonlinear distortion of the amplitude and phase of the received signal is estimated and compensated by using the Stokes constraint relationship of the two-way modulation symbols. The present invention can be applied to a single-carrier frequency domain equalization wireless transmission system, and information is transmitted by using three-dimensional modulation with Stokes constraints. Even if the power amplifier works in a nonlinear region or even a saturated amplification region, the proposed method enables the wireless receiver to still achieve high-performance reception, thereby significantly improving the power efficiency of the wireless communication system.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and in particular relates to a nonlinear amplified three-dimensional signal modulation and demodulation method and device. Background Art

[0002] Pursuing the improvement of spectrum efficiency and power efficiency is the key for future wireless networks to cope with ultra-large capacity and ultra-large-scale connection applications. The efficiency of the power amplifier is the main factor affecting the power efficiency of the wireless system. The power efficiency of the power amplifier is related to the working area of ​​the power amplifier. The power amplifier working nonlinearly has a higher power efficiency, but at the same time it will cause signal distortion, thus affecting the information recovery at the receiving end. Existing power amplifier enhancement methods or distorted signal recovery methods such as digital pre-distortion technology and receiving end power amplifier nonlinear cancellation technology are based on the nonlinear model of the power amplifier. This type of method based on the nonlinear model of the power amplifier is difficult to implement in the millimeter wave frequency band, and it is difficult to effectively improve the power amplifier efficiency in the millimeter wave band, which seriously restricts the power efficiency of future millimeter wave-based wireless systems. Summary of the invention

[0003] The purpose of the present invention is to provide a nonlinear amplified three-dimensional signal modulation and demodulation method and device to solve the technical problem of being able to maintain a low error vector magnitude (Error Vector Magnitude, EVM) at a higher power efficiency, thereby improving the power efficiency of a wireless transmission system.

[0004] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0005] A nonlinear amplified three-dimensional signal modulation and demodulation method comprises the following steps:

[0006] Step 1: Selection of three-dimensional modulation signal space; the three-dimensional modulation method refers to introducing a Stokes constraint relationship between the orthogonally transmitted x-path and y-path symbols, and then converting the non-independent four-dimensional signal space formed by the x-path and y-path symbols into an independent three-dimensional Stokes state space, selecting L three-dimensional modulation state points, and maximizing their minimum distance, so as to carry log2L bits of information.

[0007] There is a Stokes constraint relationship between the x-path and y-path symbols of orthogonal transmission. By using the transformation relationship between the Stokes vector and the Jones vector, the selection problem of the non-independent four-dimensional modulation signal space is transformed into the selection problem of independent three-dimensional Stokes sphere points. Phase modulation is also performed on the x-path and y-path symbols at the same time to obtain the four-dimensional signal state space of three-dimensional modulation.

[0008] Step 2: Implementation of three-dimensional modulation: Group the binary bit stream input to the three-dimensional modulator, using log2(L×M) bits as a symbol, where M is the additional phase modulation order, and use the four-dimensional signal state space generated in step 1 to map the modulation symbol into two orthogonal I and Q signals;

[0009] Step 3: a transmitting device for a three-dimensional modulated signal; the three-dimensional modulation method is used for a single carrier frequency domain equalization SC-FDE transmission system; after the transmission bit stream is channel-coded, the transmission symbols of x-path and y-path are obtained according to the three-dimensional modulation method of step 1 and the three-dimensional modulation of step 2, but there are Stokes constraints between the information, and the transmission symbols of x-path and y-path are respectively modulated by baseband to obtain two orthogonal signals of x-path and y-path, and the signals are respectively transmitted through a radio frequency link including a nonlinear power amplifier and a transmitting antenna;

[0010] Step 4: a receiving device for three-dimensional modulated signals; using a receiving antenna and a radio frequency link to independently receive the x-channel and y-channel radio frequency signals, and independently performing carrier demodulation and analog-to-digital conversion to obtain the corresponding baseband signals, and then performing frequency domain equalization, and then estimating and compensating the amplitude distortion of the baseband signals, and estimating and compensating the phase distortion of the constellation diagram after baseband demodulation; and then performing maximum likelihood demapping and channel decoding according to the four-dimensional signal expression of the three-dimensional modulated signal space in step 1 to obtain a received bit stream;

[0011] Step 5: Obtain the two orthogonal baseband signals x and y obtained by the receiving antenna and the receiving RF link in step 4. Due to the nonlinear effect of the power amplifier, the baseband signals x and y obtained after equalization in step 4 and the modulation symbols obtained by baseband demodulation have amplitude and phase distortion. It is necessary to estimate and compensate the distortion amplitude of the baseband signal and estimate and compensate the distortion phase of the received modulation symbol. The specific steps are as follows: Initialize the modulation symbol power estimation value P s (0) =0, as the decision criterion for the first iteration in step 7;

[0012] Step 6: baseband demodulate the two orthogonal baseband signals x and y received in step 5 or the baseband estimation signal obtained in step 10 to obtain three-dimensional modulation reception symbols of x and y;

[0013] Step 7: Calculate the average power of the modulation symbol based on the three-dimensional modulation received symbol obtained in step 6, and use the convergence of the modulation symbol power as the decision condition for the amplitude compensation iteration. The decision method is as follows: determine whether the difference between the average power of the modulation symbol obtained in this iteration and the average power of the modulation symbol in the previous iteration is less than the threshold value. If it is less than the threshold value, the average power of the modulation symbol converges; if it is greater than or equal to the threshold value, it does not converge; if it converges, output the modulation symbols of the x-path and y-path, and directly go to step 11 to estimate and compensate the distortion position; if it does not converge, proceed to step 8;

[0014] Step 8: using the x-path modulation symbol to obtain an estimated value of the y-path modulation symbol, and using the y-path modulation symbol to obtain an estimated value of the x-path modulation symbol;

[0015] Step 9: Using the estimated values ​​of the modulation symbols of the x-path and the y-path obtained in step 8, baseband shaping is performed respectively to obtain estimated values ​​of the two-path baseband signals;

[0016] Step 10: Analyze the original x-path and y-path baseband signals received in step 5 to determine whether the amplitude value of each point is greater than a specified threshold; if greater, use the amplitude of the corresponding baseband signal estimation value obtained in step 9 to replace the received signal amplitude, but retain the phase of the original received signal; if not greater, retain the original received signal amplitude and phase; after each iteration of amplitude compensation is completed, obtain the x-path and y-path baseband estimation signals and go to step 6;

[0017] Step 11: After the baseband signal amplitude is estimated and compensated in steps 5 to 10, the phase of the two-path symbol is estimated and compensated according to the correlation between the phases of the modulation symbols of the x-path and the y-path.

[0018] Furthermore, in step 1, the Stokes constraint relationship is introduced between the x-path and y-path symbols of orthogonal transmission as follows:

[0019]

[0020] Where E x is the amplitude of the modulation symbol of x channel, E y is the amplitude of the modulation symbol of the y-path, is the phase of the x-channel modulation symbol, is the phase of the modulation symbol of the y-path, S0 is the imposed constraint, expressed as the three-dimensional modulation symbol power;

[0021] The Fibonacci network is used to select the modulation state space on the Stokes sphere. The specific method for obtaining the Stokes sphere coordinates of the lth three-dimensional modulation symbol is as follows:

[0022]

[0023] Among them, S0, S1, S2 and S3 are Stokes parameters, S0 represents the sum of the power of the x-path modulation symbol and the y-path modulation symbol, that is, the power required to transmit a modulation symbol. M-order phase modulation is also performed simultaneously on the two orthogonal signals to obtain a four-dimensional signal state space of three-dimensional modulation.

[0024] Furthermore, in step 6, the x-path and y-path baseband signals with amplitude and phase distortion received in step 5 and the baseband estimation signal obtained in step 10 are baseband demodulated, and the amplitude of the baseband signal is estimated and compensated using the correlation between the three-dimensionally modulated x-path and y-path modulation symbols.

[0025] Furthermore, in step 7, the convergence of the modulation symbol power is used as the decision condition for the amplitude compensation iteration: the specific method is as follows:

[0026]

[0027] in is the average power of the modulation symbol obtained in the i-th iteration, ε is the decision threshold, and if the above formula is true, it is judged as convergence, otherwise it is not convergence.

[0028] Furthermore, in step 8, the distorted modulation symbols are used to perform cross estimation of the modulation symbol amplitudes of the x-path and y-path, and the specific method is as follows:

[0029]

[0030] in, is the symbol power estimate obtained in step 7, is the estimated value of the nth x-path modulation symbol, is the estimated value of the nth y-path modulation symbol, E x,n is the original value of the nth x-path modulation symbol, E y,n is the original value of the nth y-path modulation symbol.

[0031] Furthermore, in step 9, pulse shaping is used to obtain the estimated values ​​of the x-path and y-path baseband signals from the estimated values ​​of the modulation symbols.

[0032] Furthermore, in step 10, the original received baseband signal is reconstructed by threshold screening, the amplitude of the baseband signal point exceeding the threshold is replaced by the corresponding baseband signal estimation value, and the phase is maintained; the amplitude and phase of the baseband signal point not exceeding the threshold are maintained, and the amplitude threshold is specifically selected as follows:

[0033]

[0034] in, is the average power of the baseband signal on the x-path, is the average power of the baseband signal on the y-channel, σ is the received signal-to-noise ratio, and κ is the noise threshold coefficient.

[0035] Further, in step 11, the phase opposite relationship between the x-path modulation symbol component and the y-path modulation symbol component of the same modulation symbol of three-dimensional modulation is used to average the phase sum of the two-path modulation symbols, and then the symbol distortion phase is estimated and compensated in a manner of allocating the power ratio to the x-path and y-path modulation symbols. The specific method is as follows:

[0036]

[0037] in, and are the phase estimates of the modulation symbols for x and y, respectively, x,n and P x,n are the phase and power of the x-th modulation symbol obtained after amplitude compensation and baseband demodulation, θ y,n and P y,n is the phase and power of the n-th y-channel modulation symbol obtained after amplitude compensation and baseband demodulation, and N is the number of symbols in the data frame.

[0038] The present invention also discloses a nonlinear amplified three-dimensional signal modulation and demodulation device, which realizes the transmission and receiving structure of a single-carrier frequency domain equalization wireless communication system based on three-dimensional modulation; in the transmission structure, after the bit stream to be transmitted is channel-coded, two independently transmitted x-path and y-path symbols are obtained through a three-dimensional modulation method, and after independent baseband modulation, x-path and y-path orthogonal transmission baseband signals are obtained, and the signals are transmitted through a radio frequency link and an antenna containing a nonlinear power amplifier; the structure of the radio frequency link is related to the carrying mode of the x-path and y-path signals, if different time slots are used to carry the two-path orthogonal modulation signals, the x-path and y-path signals are carrier-modulated and successively transmitted through the same radio frequency link and transmitting antenna; if different carriers are used to carry the two-path orthogonal modulation signals, the x-path and y-path signals are respectively carrier-modulated to different carrier frequencies and transmitted through a broadband radio frequency link and transmitting antenna; if different polarized antennas are used to carry the two-path orthogonal modulation signals, the two baseband signals are respectively transmitted through radio frequency links and power amplifiers of different channels through horizontal and Two vertical orthogonal polarized antennas are used for signal transmission; for the receiving structure, the corresponding antenna reception and coherent demodulation are performed on the receiving end according to the carrying mode of the orthogonal signal; for the two orthogonal modulated signals carried in different time slots, the receiving end uses a single-channel corresponding polarized antenna and a radio frequency link to receive the two orthogonal signals in time slots; for the two orthogonal modulated signals carried by different carriers, the receiving end uses coherent demodulation of different carrier frequencies to obtain the two orthogonal modulated signals; for the two orthogonal modulated signals carried by different polarizations, the receiving end uses two horizontal and vertical polarized antennas and a radio frequency link corresponding to two channels to receive the two orthogonal modulated signals; due to the nonlinear operation of the power amplifier at the transmitting end, the x-path and y-path signals obtained by the above method have amplitude and phase distortion after frequency domain equalization, and the x-path and y-path receiving symbols are obtained after the distortion amplitude and phase are estimated and compensated, and the four-dimensional signal state space of the obtained three-dimensional modulation and the maximum likelihood method are used for soft demodulation to obtain the logarithmic likelihood ratio of the transmission bit, and finally the receiving bit stream is obtained through channel decoding.

[0039] The nonlinear amplified three-dimensional signal modulation and demodulation method and device of the present invention have the following advantages:

[0040] 1. The present invention generates two orthogonal transmission signals, x-path and y-path, based on three-dimensional modulation. When the two orthogonal signals transmit the same symbol information, the total power remains unchanged and the power values ​​of the two signals and their changes show opposite characteristics. At the receiving end, the x-path and y-path symbols are mutually estimated, and then pulse shaping is performed to obtain mutually estimated baseband signals. The total power value of the two signals is estimated in an iterative manner, and the signal amplitude distortion can be estimated and compensated.

[0041] 2. The present invention performs baseband demodulation on the amplitude compensated signal, estimates the distortion of the symbol phase based on the phase correlation between the x-path and y-path symbols generated by the three-dimensional modulation, and thereby compensates the phase of the two-path modulation symbols. Compared with the three-dimensional modulation symbols without signal estimation and compensation, the EVM of the received three-dimensional modulation symbols after amplitude compensation and phase compensation has obvious advantages when working in the nonlinear region of the power amplifier. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a device for implementing the three-dimensional modulation method of the present invention;

[0043] Figure 2 The demodulation method of the three-dimensional modulation of the present invention is realized;

[0044] Figure 3 It is a flow chart of the three-dimensional modulation distortion signal estimation and compensation method of the present invention;

[0045] Figure 4 It is a power amplifier model used to verify the algorithm proposed in the present invention, wherein (a) is the power amplifier amplitude modulation-amplitude modulation (Amplitude-to-Amplitude modulation, AM-AM) effect, (b) is the power amplifier amplitude modulation-phase modulation (Amplitude-to-Phase modulation, AM-PM) effect, and (c) is a schematic diagram of power amplifier gain compression;

[0046] Figure 5 Schematic diagram of the amplitude distortion compensation effect of the present invention, wherein (a) is a comparison between a baseband signal without amplitude compensation and an original transmission signal, and (b) is a comparison between a baseband signal with amplitude compensation and an original transmission signal.

[0047] Figure 6 Schematic diagram of the phase compensation effect of the present invention, wherein (a) is a comparison between a received constellation diagram without phase compensation and an original transmitted symbol constellation diagram, and (b) is a comparison between a received constellation diagram with phase compensation and an original transmitted symbol constellation diagram.

[0048] Figure 7 It is a schematic diagram of the performance improvement of the error vector amplitude under different back-off powers by the signal reconstruction of the present invention. DETAILED DESCRIPTION

[0049] In order to better understand the purpose, structure and function of the present invention, the following further describes in detail a nonlinear amplified three-dimensional signal modulation and demodulation method and device of the present invention in conjunction with the accompanying drawings.

[0050] In the present invention, the nonlinear amplification effect of the power amplifier model is used to demonstrate the effectiveness of the proposed method. Figure 4As shown in the figure, the AM-AM effect, AM-PM effect and power amplifier gain curve of the power amplifier model are given. The specific model parameters are as follows:

[0051] The nonlinear distortion AM-AM effect of the power amplifier adopts the Rapp model:

[0052]

[0053] Among them, G(A) is the power amplifier gain, A is the input signal amplitude, g, A sat , s are the parameters of the power amplifier AM-AM effect Rapp model, and the specific configuration is shown in Table 1.

[0054] The AM-PM effect adopts the modified Rapp model:

[0055]

[0056] Among them, Ψ(A) is the output phase of the power amplifier, α, β, q1 and q2 are the parameters of the Rapp model of the AM-PM effect of the power amplifier, and the specific configuration is shown in Table 1.

[0057] The parameter configurations of the above two models are shown in Table 1.

[0058] Table 1

[0059] g <![CDATA[A sat ]]> s α β <![CDATA[q1]]> <![CDATA[q2]]> 4.65 0.58 0.81 2560 0.114 2.4 2.3

[0060] The three-dimensional modulated signal is distorted under the influence of the nonlinear region of the power amplifier. The distorted signal is transmitted to the receiving end through the transmitting and receiving antennas, and a distorted baseband signal is obtained after passing through the radio frequency link.

[0061] according to Figure 1 The three-dimensional modulation signal transmitting device shown, Figure 2 The three-dimensional modulation signal receiving device shown in Figure 3 The flowchart of the distortion signal estimation and compensation steps shown in the figure is as follows. The specific three-dimensional modulation transmitting and receiving device and the signal distortion estimation and compensation implementation method are described below. The present invention includes the following steps:

[0062] Step 1: Design of three-dimensional modulation signal space. In order to suppress the influence of power amplifier nonlinear distortion on the error rate of information transmission, a certain constraint relationship is introduced between the x-path and y-path modulation symbols carried in the orthogonal dimension. The constraint relationship is as follows:

[0063]

[0064] Among them, E x and E y Respectively represent the amplitude values ​​of the x and y path symbols corresponding to the modulation symbols, and They represent the phases of the x-path and y-path symbols corresponding to the modulation symbols. S0 is the imposed constraint.

[0065] The four-dimensional I / Q signal space formed by the x-path and y-path symbols with such constraints is not independent and can be converted into an independent three-dimensional Stokes state space. Three-dimensional modulation refers to a modulation method that selects L (three-dimensional modulation order) state points with the smallest distance and the largest value based on this three-dimensional Stokes state space to carry log2L bits of information.

[0066] The selection of the modulation state space of this modulation method must ensure that the minimum Euclidean distance of each symbol in the four-dimensional signal space is maximized. The existence of the constraint relationship makes the four-dimensional signals not independent of each other, so it cannot be performed directly in the signal domain.

[0067] The transformation relationship between the Stokes vector and the Jones vector can satisfy the above constraints, and its transformation expression is as follows:

[0068]

[0069] Among them, S0, S1, S2 and S3 are Stokes parameters, S0 represents the sum of the power of the x-modulation symbol and the y-path modulation symbol, that is, the power required to transmit a modulation symbol. When the amplitude constraint S0 is determined, the selection of the four-dimensional signal state space of the three-dimensional modulation is transformed into the selection of the spherical state space with a radius of S0. For L×M-order three-dimensional modulation, L is the number of Stokes sphere state points, and it is necessary to select L points with the largest minimum distance on the sphere as the modulation state space. M is the additional phase modulation order. The present invention introduces the Fibonacci network to select L points on the sphere so that the minimum Euclidean distance of the L spherical points is maximized. After S0 is normalized, the coordinates of the lth point on the sphere are:

[0070]

[0071] Since this three-dimensional modulation space selection method based on the Stokes sphere only occupies the first and fourth quadrants of the constellation diagram in the constellation diagram of the I / Q signal space of the x-path and y-path, the signal space utilization is insufficient and there is a DC component. Therefore, binary phase shift keying (BPSK) is also performed on the two orthogonal signals of the x-path and y-path at the same time, that is, M=2, to obtain a four-dimensional signal space of three-dimensional modulation.

[0072] Step 2: Implementation of 3D modulation. Group the binary bit stream input to the 3D modulator, take log2(L×M) (L×M is the 3D modulation order) bits as a symbol, and use the 4D modulation signal space generated in step 1 to map the modulation symbol to I and Q signals carried orthogonally by x and y channels. The modulation symbol carried orthogonally by the nth channel is expressed as:

[0073]

[0074] Among them, φ n is the symbol phase obtained by BPSK, which is 0 or π. Therefore, the sum of the phases of the x-path and y-path signals is 0 or 2π. BPSK does not affect the above constraints.

[0075] Step 3: Transmitter of 3D modulated signal. For a single carrier frequency domain equalization (SC-FDE) transmission system based on 3D modulation, its transmitter is as follows: Figure 1 As shown. After the transmission bit stream is channel coded, two transmission symbols with orthogonal transmission but Stokes constraints between information are obtained according to the three-dimensional modulation method in step 2. The two orthogonal transmission symbols are modulated by baseband to obtain two orthogonal signals, which are transmitted through the RF link containing a nonlinear power amplifier and the transmitting antenna respectively. The baseband signal expression at the transmitting end is:

[0076]

[0077] Where p(t) is the shaped pulse, T s is the symbol period.

[0078] The signal transmission method is related to the dimension of the two orthogonal signals. If different time slots are used to carry the two orthogonal modulated signals, the two signals are sent successively through the same RF link and transmitting antenna; if different carriers are used to carry the two orthogonal modulated signals, the two baseband signals are respectively modulated to different carrier frequencies and transmitted through broadband RF links and transmitting antennas; if different polarized antennas are used to carry the two orthogonal modulated signals, the two baseband signals are respectively transmitted through RF links and power amplifiers of different channels through horizontal and vertical orthogonal polarization antennas.

[0079] Step 4: Receive the three-dimensional modulated transmission signal of step 3. Perform corresponding processing on the receiving end according to the carrying method of the orthogonal signal. For two orthogonal modulated signals carried in different time slots, the receiving end uses a single-channel antenna with corresponding polarization and a radio frequency link to receive the two orthogonal signals in time slots; for two orthogonal modulated signals carried by different carriers, the receiving end uses coherent demodulation of different carrier frequencies to obtain two orthogonal modulated signals; for two orthogonal modulated signals carried by different polarizations, the receiving end uses two antennas with horizontal and vertical polarizations and a radio frequency link corresponding to two channels to receive the two orthogonal modulated signals. After receiving the baseband signal, frequency domain equalization is performed to obtain two orthogonal modulated receiving baseband signals, and the received signal is expressed as:

[0080]

[0081] in, The baseband signals received by x and y channels, S x (t), S y (t) is the baseband signal transmitted by x and y paths, h xx (t), h xy (t), h yx (t), h yy (t) is the channel impulse response including the RF link, Z x (t) and Z y (t) is Gaussian white noise, and ⊙ represents the convolution operation.

[0082] Step 5: The x and y baseband signals obtained in step 4 are distorted due to the nonlinear operation of the power amplifier. Directly using them for three-dimensional signal demodulation will result in a large error vector amplitude and symbol error rate. Therefore, the following steps estimate and compensate for the signal distortion. Based on the two orthogonal baseband signals x and y obtained in step 4, the symbol power (the total power of the two orthogonal symbols of x and y) estimate is initialized at the same time. As the decision criterion for the first iteration in step 7.

[0083] Step 6: baseband demodulate the two orthogonal baseband signals x and y received in step 5 or the baseband estimation signals obtained through steps 6 to 10 to obtain three-dimensional modulation reception symbols for x and y;

[0084] Step 7: Calculate the average power of the modulation symbol of the i-th iteration based on the modulation symbols of the x and y paths obtained in step 6. The calculation method is:

[0085]

[0086] in, represents the average power of the modulation symbol obtained in the i-th iteration, and Respectively represent the amplitude values ​​of the x-path and y-path symbols corresponding to the n-th modulation symbol in the ith iteration. Determine whether the average power of the modulation symbol converges. If so, output the x-path and y-path symbols and go to step 11; if not, go to step 8.

[0087] The convergence decision is expressed as follows:

[0088]

[0089] in, is the average power of the modulation symbol obtained in the i-th iteration, ε is the decision threshold, and if the above formula is true, it is judged as convergence, otherwise it is not convergence.

[0090] Step 8: For the three-dimensional system, there is a constraint relationship between the amplitudes of the x-path symbol component and the y-path symbol component of the same modulation symbol, namely:

[0091]

[0092] It can be seen that the peak value of the x-channel modulation signal and the peak value of the y-channel modulation signal will not be reached at the same time, and the amplitudes of the two signals are offset and complementary. When the x-channel modulation signal is in the nonlinear region of the power amplifier, the y-channel modulation signal is often in the linear region, and vice versa. In this way, the amplitudes of the two signals can be cross-estimated to compensate for the amplitude of the distorted signal. The symbol estimation method is as follows:

[0093]

[0094] in, is the symbol power estimate obtained in step 7, E x,n and E y,n Respectively represent the amplitude values ​​of the x and y path symbols corresponding to the nth modulation symbol, and They respectively represent the amplitude estimation values ​​of the x-path and y-path symbols corresponding to the n-th modulation symbol.

[0095] Step 9: Using the estimated values ​​of the modulation symbols of paths x and y obtained in step 8, perform raised cosine pulse shaping filtering respectively to obtain estimated values ​​of the two baseband signals. According to the analysis in step 8, when the modulation symbol power value converges, the estimated value can restore the peak part of the original signal.

[0096] Step 10: Analyze the original baseband signals of x and y received in step 4 to determine whether the amplitude value of each point is greater than the specified threshold. If it is greater, the amplitude of the corresponding baseband signal estimation value obtained in step 9 is used to replace the received signal amplitude, but the phase of the original received signal is retained; if it is not greater, the original received signal amplitude and phase are retained. The setting of the threshold is related to the average power and received signal-to-noise ratio of the original received three-dimensional modulated baseband signal, and its threshold expression is:

[0097]

[0098] in, and are the average power of the baseband received signals on the x-path and y-path, σ is the received signal-to-noise ratio, and κ is the noise threshold coefficient. After the average power of the modulation symbol converges, the amplitude compensation effect is as follows: Figure 5 After the amplitude compensation of each iteration is completed, go to step 7.

[0099] Step 11: After the baseband signal amplitude is estimated and compensated in steps 5 to 10, the received baseband signal is demodulated by baseband demodulation to obtain the amplitude-compensated phase-distorted x and y modulation symbols. After passing through the power amplifier, the phase distortion of the large signal is worse than that of the small signal, and the degree of phase distortion is positively correlated with the signal amplitude. For the three-dimensional modulation method, the phases of the x-path modulation symbol component and the y-path modulation symbol component of the same modulation symbol are opposite, so the sum of the phases of the x-path modulation symbol and the y-path modulation symbol is 0°, and the power amplifier has a stable mean value for the distortion of the phase sum of each modulation symbol. Therefore, the average value of the phase sum of the two modulation symbols can be calculated and then distributed to the x and y modulation symbols according to the power ratio to obtain an estimated value of the distorted phase. The compensated phase can be expressed as

[0100]

[0101] in, and are the phase estimates of the modulation symbols for x and y, respectively, x,n and P x,n are the phase and power of the x-th modulation symbol obtained after amplitude compensation and baseband demodulation, θ y,n and P y,n is the phase and power of the nth y-channel modulation symbol obtained after amplitude compensation and baseband demodulation, N is the number of symbols in the data frame, and the phase compensation effect is as follows: Figure 6 shown.

[0102] Signal estimation and compensation results using the proposed method:

[0103] Figure 7 The error vector magnitude comparison results between the uncompensated received signal, the signal compensated by the above method and the original transmitted signal are given when the power amplifier works at different back-off powers. Figure 7 It can be seen that when the working area of ​​the power amplifier changes from the linear region to the nonlinear region (i.e. Figure 7 When the horizontal axis back-off power increases, the compensated 3D modulated received signal has a significant EVM reduction compared to the uncompensated 3D modulated received signal, thereby ensuring that the power amplifier operates at a higher efficiency without damaging the demodulation performance of the 3D modulation.

[0104] The present invention discloses a nonlinear amplified three-dimensional signal modulation and demodulation method and device. The method uses a wireless transmission system to realize three-dimensional modulation with Stokes constraints. The two signals generated by the three-dimensional modulation at the transmitting end are independently orthogonalized through a power amplifier and an antenna working in nonlinear amplification for signal transmission. The receiving end obtains a received signal with nonlinear distortion through a polarization orthogonal antenna (the transmitting end uses an orthogonal signal with a constraint relationship in the polarization dimension) or a single antenna (the transmitting end uses an orthogonal signal with a constraint relationship in the frequency dimension). The nonlinear distortion of the amplitude and phase of the received signal is estimated and compensated by using the Stokes constraint relationship of the two-way modulation symbols. The present invention can be applied to a single-carrier frequency domain equalization (polarization) wireless transmission system, and information is transmitted by using three-dimensional modulation with Stokes constraints. Even if the power amplifier works in a nonlinear region or even a saturated amplification region, the proposed method enables the wireless receiver to still achieve high-performance reception, thereby significantly improving the power efficiency of the wireless communication system.

[0105] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A nonlinear amplified three-dimensional signal modulation and demodulation method, characterized in that: The following steps are involved: Step 1: Selection of three-dimensional modulation signal space; The three-dimensional modulation method refers to introducing a Stokes constraint relationship between the orthogonally transmitted x-path and y-path symbols, and then converting the non-independent four-dimensional signal space formed by the x-path and y-path symbols into an independent three-dimensional Stokes state space, selecting L three-dimensional modulation state points, and maximizing their minimum distance, so as to carry log2L bits of information. There is a Stokes constraint relationship between the x-path and y-path symbols of orthogonal transmission. By using the conversion relationship between the Stokes vector and the Jones vector, the selection problem of the non-independent four-dimensional modulation signal space is converted into the selection problem of independent three-dimensional Stokes sphere points. Since this Stokes sphere modulation signal space selection method only occupies the first and fourth quadrants of the constellation diagram in the constellation diagram of the x-path and y-path signals, the signal space utilization rate is insufficient and there is a DC component. Therefore, phase modulation is also performed on the x-path and y-path symbols at the same time to obtain a four-dimensional signal state space of three-dimensional modulation. Step 2: Implementation of three-dimensional modulation: Group the binary bit stream input to the three-dimensional modulator, using log2(L×M) bits as a symbol, where M is the additional phase modulation order, and use the four-dimensional signal state space generated in step 1 to map the modulation symbol into two orthogonal I and Q signals; Step 3: a transmitting device for a three-dimensional modulated signal; the three-dimensional modulation method is used for a single carrier frequency domain equalization SC-FDE transmission system; after the transmission bit stream is channel-coded, the transmission symbols of x-path and y-path are obtained according to the three-dimensional modulation method of step 1 and the three-dimensional modulation of step 2, but there are Stokes constraints between the information, and the transmission symbols of x-path and y-path are respectively modulated by baseband to obtain two orthogonal signals of x-path and y-path, and the signals are respectively transmitted through a radio frequency link including a nonlinear power amplifier and a transmitting antenna; Step 4: a receiving device for three-dimensional modulated signals; using a receiving antenna and a radio frequency link to independently receive the x-channel and y-channel radio frequency signals, and independently performing carrier demodulation and analog-to-digital conversion to obtain the corresponding baseband signals, and then performing frequency domain equalization, and then estimating and compensating the amplitude distortion of the baseband signals, and estimating and compensating the phase distortion of the constellation diagram after baseband demodulation; Then, according to the four-dimensional signal expression of the three-dimensional modulation signal space in step 1, maximum likelihood demapping and channel decoding are performed to obtain a received bit stream; Step 5: Obtain the two orthogonal baseband signals x and y obtained by the receiving antenna and the receiving RF link in step 4. Due to the nonlinear effect of the power amplifier, the baseband signals x and y obtained after equalization in step 4 and the modulation symbols obtained by baseband demodulation have amplitude and phase distortion. It is necessary to estimate and compensate the distortion amplitude of the baseband signal and estimate and compensate the distortion phase of the received modulation symbol. The specific steps are as follows: Initialize the modulation symbol power estimate As the decision criterion for the first iteration in step 7; Step 6: baseband demodulate the two orthogonal baseband signals x and y received in step 5 or the baseband estimation signal obtained in step 10 to obtain three-dimensional modulation reception symbols of x and y; Step 7: Calculate the average power of the modulation symbol according to the three-dimensional modulation reception symbol obtained in step 6, and use the convergence of the modulation symbol power as a decision condition for amplitude compensation iteration; The judgment method is as follows: judge whether the difference between the average power of the modulation symbol obtained in this iteration and the average power of the modulation symbol in the previous iteration is less than the threshold value, if it is less than the threshold value, the average power of the modulation symbol converges, if it is greater than or equal to the threshold value, it does not converge; if it converges, output the modulation symbols of the x-path and y-path, and directly go to step 11 to estimate and compensate the distortion position; if it does not converge, go to step 8; Step 8: using the x-path modulation symbol to obtain an estimated value of the y-path modulation symbol, and using the y-path modulation symbol to obtain an estimated value of the x-path modulation symbol; Step 9: Using the estimated values ​​of the modulation symbols of the x-path and the y-path obtained in step 8, baseband shaping is performed respectively to obtain estimated values ​​of the two-path baseband signals; Step 10: Analyze the original x-path and y-path baseband signals received in step 5 to determine whether the amplitude value of each point is greater than a specified threshold; if greater, use the amplitude of the corresponding baseband signal estimation value obtained in step 9 to replace the received signal amplitude, but retain the phase of the original received signal; If it is not greater than, the original received signal amplitude and phase are retained; after each iteration of amplitude compensation is completed, the x-path and y-path baseband estimation signals are obtained and the process goes to step 6; Step 11: After the baseband signal amplitude is estimated and compensated in steps 5 to 10, the received baseband signal performs two-path symbol phase estimation and compensation according to the correlation between the modulation symbol phases of the x-path and the y-path; In step 1, the Stokes constraint relationship is introduced between the x-path and y-path symbols of orthogonal transmission as follows: Where E x is the amplitude of the modulation symbol of x-path, E y is the amplitude of the modulation symbol of the y-path, is the phase of the x-channel modulation symbol, is the phase of the modulation symbol of the y-path, S0 is the imposed constraint, expressed as the three-dimensional modulation symbol power; The Fibonacci network is used to select the modulation state space on the Stokes sphere. The specific method for obtaining the Stokes sphere coordinates of the lth three-dimensional modulation symbol is as follows: Among them, S0, S1, S2 and S3 are Stokes parameters, S0 represents the sum of the power of the modulation symbol of the x-path and the modulation symbol of the y-path, that is, the power required to transmit a modulation symbol. M-order phase modulation is also performed on the two orthogonal signals at the same time, and the four-dimensional signal state space of the three-dimensional modulation is obtained; In the step 11, the phase inverse relationship between the x-path modulation symbol component and the y-path modulation symbol component of the same modulation symbol of three-dimensional modulation is used to average the phase sum of the two-path modulation symbols, and then the symbol distortion phase is estimated and compensated in a manner of allocating the power ratio to the x-path and y-path modulation symbols. The specific method is as follows: in, and are the phase estimates of the modulation symbols for x and y, respectively, x,n and P x,n are the phase and power of the x-th modulation symbol obtained after amplitude compensation and baseband demodulation, θ y,n and P y,n is the phase and power of the n-th y-channel modulation symbol obtained after amplitude compensation and baseband demodulation, and N is the number of data frame symbols.

2. The nonlinear amplified three-dimensional signal modulation and demodulation method according to claim 1, characterized in that: In step 6, the x-path and y-path baseband signals with amplitude and phase distortion received in step 5 and the baseband estimation signal obtained in step 10 are baseband demodulated, and the amplitude of the baseband signal is estimated and compensated using the correlation between the three-dimensionally modulated x-path and y-path modulation symbols.

3. The nonlinear amplified three-dimensional signal modulation and demodulation method according to claim 2, characterized in that: In step 7, the convergence of the modulation symbol power is used as the decision condition for the amplitude compensation iteration: the specific method is as follows: in is the average power of the modulation symbol obtained in the i-th iteration, ε is the decision threshold, and if the above formula is true, it is judged as convergence, otherwise it is not convergence.

4. The nonlinear amplified three-dimensional signal modulation and demodulation method according to claim 3, characterized in that: In step 8, the distorted modulation symbols are used to perform cross estimation of the modulation symbol amplitudes of the x-path and y-path, and the specific method is as follows: in, is the symbol power estimate obtained in step 7, is the estimated value of the nth x-path modulation symbol, is the estimated value of the nth y-path modulation symbol, E x,n is the original value of the nth x-path modulation symbol, E y,n is the original value of the nth y-path modulation symbol.

5. The nonlinear amplified three-dimensional signal modulation and demodulation method according to claim 4, characterized in that: In step 9, pulse shaping is used to obtain the estimated values ​​of the x-path and y-path baseband signals from the estimated values ​​of the modulation symbols.

6. The nonlinear amplified three-dimensional signal modulation and demodulation method according to claim 5, characterized in that: In step 10, the original received baseband signal is reconstructed by threshold screening, the amplitude of the baseband signal point exceeding the threshold is replaced by the corresponding baseband signal estimation value, and the phase is maintained; the amplitude and phase of the baseband signal point not exceeding the threshold are maintained, and the amplitude threshold is specifically selected as follows: in, is the average power of the baseband signal on the x-path, is the average power of the baseband signal on the y-channel, σ is the received signal-to-noise ratio, and κ is the noise threshold coefficient.

7. A nonlinear amplified three-dimensional signal modulation and demodulation device, characterized in that: The device realizes the transmitting and receiving structure of a single-carrier frequency-domain equalization wireless communication system based on three-dimensional modulation; in the transmitting structure, after the bit stream to be transmitted is channel-coded, two independently transmitted x-path and y-path symbols are obtained by the three-dimensional modulation method described in claim 2, and after independent baseband modulation, x-path and y-path orthogonal transmission baseband signals are obtained, and the signals are transmitted through a radio frequency link and antenna containing a nonlinear power amplifier; the structure of the radio frequency link is related to the carrying method of the x-path and y-path signals. If different time slots are used to carry the two orthogonal modulated signals, the x-path and y-path signals are carrier-modulated and successively sent through the same radio frequency link and transmitting antenna; if different carriers are used to carry the two orthogonal modulated signals, the x-path and y-path signals are respectively carrier-modulated to different carrier frequencies and transmitted through a broadband radio frequency link and transmitting antenna; if different polarized antennas are used to carry the two orthogonal modulated signals, the two baseband signals are respectively transmitted through radio frequency links and power amplifiers of different channels, and are transmitted through two orthogonally polarized antennas of horizontal and vertical polarizations. signal transmission; for the receiving structure, according to the carrying mode of the orthogonal signal, the corresponding antenna reception and coherent demodulation are performed on the receiving end; for the two orthogonal modulated signals carried in different time slots, the receiving end uses a single-channel antenna with corresponding polarization and a radio frequency link to receive the two orthogonal signals in time slots; for the two orthogonal modulated signals carried by different carriers, the receiving end uses coherent demodulation of different carrier frequencies to obtain the two orthogonal modulated signals; for the two orthogonal modulated signals carried by different polarizations, the receiving end uses two antennas with horizontal and vertical polarizations and a radio frequency link corresponding to two channels to receive the two orthogonal modulated signals; due to the nonlinear operation of the power amplifier at the transmitting end, the obtained x-path and y-path signals have amplitude and phase distortion after frequency domain equalization, and the x-path and y-path receiving symbols are obtained after the estimation and compensation of the distortion amplitude and phase as described in any one of claims 2-6, and the four-dimensional signal state space of the three-dimensional modulation obtained in claim 2 and the maximum likelihood method are used for soft demodulation to obtain the logarithmic likelihood ratio of the transmission bit, and finally the received bit stream is obtained through channel decoding.

Images