Channel decoupling and equalization method and system for wifi backscatter

Abstract

The application relates to the field of backscattering technology and discloses a channel decoupling and equalization method and system for WiFi backscattering, which comprises the following steps: constructing an equation group based on a reference signal and a backscattering signal to solve a forward channel and a backscattering channel, wherein the equation group comprises 56 equations constructed by using a VHT-LTF field and 64 equations constructed by using a reference symbol; obtaining a matrix M through frequency domain to time domain conversion, time domain modulation and time domain to frequency domain conversion based on the reference symbol and a known tag sequence; solving a particular solution of the forward channel and a particular solution of the backscattering channel by simultaneously solving the equation group and combining the matrix M; and using a double-end channel equalization method to decode tag data through obtained first equalization signals and second equalization signals. The application designs a channel evaluation and equalization method, so that the modulation granularity of the tag can be further reduced, and the system transmission rate is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of backscatter technology, and in particular to a channel decoupling and equalization method and system for WiFi backscatter. BACKGROUND

[0002] With the vigorous development of Internet of Things technology, billions of Internet of Things devices are deployed in various scenarios such as homes, cities, and factories to collect data and perceive environmental conditions. However, the rapid increase in the number of devices has led to resource consumption problems, limiting large-scale deployment. To solve this problem, environmental backscatter technology has emerged. It uses a passive communication mechanism and has advantages such as low power consumption and low cost. The backscatter tag itself does not generate a signal, but instead carries data on the environmental signal for transmission, thus having lower resource consumption.

[0003] WiFi signals are one of the most widely used signals in daily life and can provide abundant carrier resources. Therefore, WiFi backscatter technology has become one of the research hotspots in recent years. Researchers have been working to reduce the modulation granularity to improve tag data transmission efficiency and speed up system operation. However, recent research has encountered a bottleneck: although the modulation granularity of existing systems has been reduced from packet symbol level to sub-symbol level, the latest results still require four sampling points to modulate a data bit, and none of them can achieve single-sampling-point-level modulation, still with modulation redundancy. Through systematic analysis, we found that the root cause of this limitation is that the tag modulation changes the transmission model of the system, causing the traditional channel estimation and equalization methods to fail.

[0004] Figure 1 shows the structure diagram of a traditional active WiFi system, Figure 2 shows the structure diagram of a WiFi backscatter system. By comparing the structure diagrams of the two systems, it can be seen that in the traditional active system, there is only one end-to-end wireless channel between the sending end and the receiving end While the backscatter system consists of three parts: the sending end generates WiFi signals as carriers, the tag modulates data onto the carrier and reflects it out, and the receiving end receives the backscattered signal. To avoid interference from the original signal, the backscattered signal is shifted to another frequency band; in this system, the tag divides the original channel into two independent channels: the forward channel from the sending end to the tag and the backscatter channel from the tag to the receiving end Generally speaking, within a short timeframe, a wireless channel can be approximated as a linear time-invariant (LTI) system in the frequency domain, and multiple channels can be linearly superimposed. However, when tag modulation is introduced between two channels, if this linearity is disrupted, the forward and backscatter channels can no longer be linearly superimposed. This leads to traditional channel estimation methods failing to accurately estimate the two channels, resulting in the tag data not being correctly decoded.

[0005] We found that when the tag's modulation unit is greater than or equal to one OFDM WiFi symbol, the linearity can be maintained. Therefore, symbol-level systems, such as FreeRider and RapidRider, can use traditional channel assessment and equalization methods to eliminate the additional phase difference introduced by the channel. However, these systems sacrifice transmission efficiency, exhibiting high modulation redundancy and low throughput. For example, the throughput of the single-symbol-level system RapidRider is only 250 kbps. When the modulation unit is smaller than one OFDM WiFi symbol, this linearity is disrupted, and traditional methods cannot properly eliminate the channel's influence. Existing work Hiscatter did not find this problem, but instead used phase difference between adjacent sampling points to approximate the elimination of the additional phase offset introduced by the channel. This method is only applicable to simple channel conditions and fails in complex real-world environments. Furthermore, this method relies on modulation redundancy, requiring 4 sampling points to encode 1 bit of tag data. Therefore, how to correctly assess and equalize the channel of the backscatter system is key to achieving efficient transmission. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a channel decoupling and equalization method and system for WiFi backscattering.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] In a first aspect, the present invention provides a channel decoupling and equalization method for WiFi backscattering, comprising:

[0009] The original signal from the transmitter is obtained through the original channel. The reference signal obtained later The original signal from the transmitting end passes through the forward channel. The tag then receives and modulates the tag data, which is then transmitted via the backscatter channel. Emit backscatter signal , obtain The backscattered signal includes a VHT-LTF field and a reference symbol; the reference symbol is the first OFDM WiFi symbol in the data field of the backscattered signal that has been tagged with a known tag sequence.

[0010] A set of equations is constructed based on the reference signal and the backscattered signal. The set of equations includes 56 equations constructed using the VHT-LTF field and 64 equations constructed using the reference symbol.

[0011] Based on the reference symbol and the known tag sequence, matrix M is obtained through frequency-domain to time-domain transformation, time-domain tag modulation, and time-domain to frequency-domain transformation;

[0012] Solving the system of equations simultaneously, and combining it with matrix M, yields the solution for the forward channel. Special solution and backscatter channel Special solution ;

[0013] A dual-channel equalization method is used to apply equalization to the original signal. The first equalization signal is obtained, and an equalization signal is applied to the backscattered signal. The second equalization signal is obtained, and the phase difference between the first equalization signal and the second equalization signal is calculated to decode the tag data.

[0014] In one embodiment, the forward channel It is a 64-dimensional vector containing 56 unknown channel variables; The channel variable represents the forward channel corresponding to the j-th subcarrier;

[0015] Backscattering channel It is a 64-dimensional vector containing 64 unknown channel variables; This represents the channel variable of the backscatter channel corresponding to the i-th subcarrier.

[0016] In one embodiment, the 56 equations constructed using the VHT-LTF field specifically include:

[0017] VHT-LTF field of backscattered signal It is represented as:

[0018] ;

[0019] in, It is a VHT-LTF field predefined by the sender. When calculating the VHT-LTF field of the backscattered signal, only consider Data information of 56 non-empty subcarriers; It is a 56-dimensional identity matrix. It is a vector consisting of the channel variables of the forward channel corresponding to 56 non-empty subcarriers. for The vector consisting of the channel variables in the backscatter channel corresponding to the 56 channel variables in the data;

[0020] The 56 equations corresponding to the VHT-LTF field are:

[0021] (2)

[0022] This represents the data corresponding to the j-th subcarrier in the VHT-LFT field of the backscattered signal. This represents the data corresponding to the j-th subcarrier in the VHT-LTF field predefined by the transmitter.

[0023] In one embodiment, the 64 equations constructed using reference symbols specifically include:

[0024] Reference symbol of backscattered signal for:

[0025] ;

[0026] in, , It is the reference symbol in the original signal generated by the transmitting end. ; It is a known 64-bit tag sequence; Indicates the presence of a forward channel The vector of channel variables for the 64 subcarriers in the middle. , Indicates the presence of a backscatter channel A vector of channel variables for 64 subcarriers. ;

[0027] The 64 equations constructed using reference symbols are as follows:

[0028] (4)

[0029] The data representing the position of the i-th subcarrier in the reference symbol of the backscattered signal. The data representing the position of the i-th subcarrier in the reference symbol of the original signal. This represents the data in the i-th row and j-th column of the matrix M.

[0030] In one embodiment, obtaining matrix M based on reference symbols and known tag sequences through frequency-to-time domain transformation, time-domain tag modulation, and time-to-frequency domain transformation specifically includes:

[0031] The transmitting end constructs the reference symbol of the original signal. The zero subcarrier is located at the center position in the OFDM WiFi symbol, and before sending , the transmitting end performs ascending arrangement Shift() on the subcarriers of the OFDM WiFi symbol, and after the re-arrangement , the OFDM WiFi symbol is Shift( ). Then the transmitting end converts the reference signal from the frequency domain to the time domain through inverse Fourier transform: ; F1 is a coefficient matrix in the inverse Fourier transform, , each row in F1 is multiplied by Shift( ); is a time domain signal of Shift( );

[0032] The tag applies phase modulation on each sampling point of the time domain signal: ; wherein, , the i-th element in the known tag sequence in the time domain is multiplied by the i-th row of Shift( ), is the modulated signal;

[0033] The receiving end converts the received modulated signal from the time domain back to the frequency domain through Fourier transform, rearranges the order of the subcarriers, restores the zero subcarrier to the center position, and obtains the matrix M:

[0034] ;

[0035] , wherein, is a matrix multiplication operation, Shift2() is a subcarrier rearrangement operation, and F2 is a coefficient matrix in the Fourier transform.

[0036] In one embodiment, the calculation method of the coefficient matrix F1 in the inverse Fourier transform is:

[0037] ;

[0038] , the index of the k-th subcarrier corresponds to the index of the row in the coefficient matrix F1; , the index of the n-th time domain sampling point corresponds to the index of the column in the coefficient matrix F1, , , the number of rows and columns in the coefficient matrix F1, ; is an imaginary unit.

[0039] In one embodiment, the calculation method of the coefficient matrix F2 in the Fourier transform is:

[0040] .

[0041] denotes the index of the kth subcarrier, and corresponds to the index of the row in the coefficient matrix F1; denotes the index of the nth time domain sampling point, and corresponds to the index of the column in the coefficient matrix F1, , denotes the number of rows and columns in the coefficient matrix F1, ; is an imaginary unit.

[0042] In one embodiment, the simultaneous equations are solved in combination with the matrix M to obtain a particular solution of the forward channel and a particular solution of the backscatter channel , and specifically comprises:

[0043] solving the forward channel , the equations corresponding to the non-empty subcarrier positions in formula (4) are simultaneously solved with formula (2), and the variable is eliminated, to obtain a matrix with a dimension of [56, 56], and theoretically ; since is a non-zero vector, when the rank of the matrix is less than 56, is the null space of ; the solution is obtained by calculating the eigenvectors and obtaining the smallest eigenvalue , and using the eigenvalue as the particular solution of the forward channel, that is, and have a multiple relationship, , K is a constant coefficient; after obtaining , the particular solution of the backscatter channel is obtained by substituting formula (4):

[0044] ;

[0045] is the reference symbol of the backscatter signal.

[0046] In one embodiment, the double-ended channel equalization method is used to apply to the original signal to obtain a first equalized signal, and to apply to the backscatter signal to obtain a second equalized signal, and the phase difference between the first equalized signal and the second equalized signal is calculated to decode the tag data, and specifically comprises:

[0047] The signal before tag modulation is denoted as , and the signal after tag modulation is denoted as , the original signal is multiplied by the specific solution of the forward channel obtained , the backscattering signal received by the receiving end is divided by the specific solution of the backscattering channel obtained ; by calculating and The phase difference of each sampling point pair in the signal is the label information, which realizes single sampling point level label data modulation and demodulation.

[0048] In a second aspect, the present application provides a computer system comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method of any one of the embodiments of the first aspect when executing the computer program.

[0049] Compared with the prior art, the present application has the beneficial technical effects that:

[0050] In order to solve the problem that the channel in the sub-symbol level modulation OFDM WiFi backscattering system cannot be accurately evaluated and equalized, the present application proposes a new composite channel decoupling method based on equation solving, independently evaluates the channel state of the forward channel and the backscattering channel in the backscattering link, and further proposes a double-end channel equalization method to correctly equalize the influence of the channel. The present application aims to design a customized channel evaluation and equalization method for the backscattering system, so as to further reduce the modulation granularity of the label, improve the system transmission rate, and promote the application of backscattering technology in real life. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 is a transmission architecture diagram of a traditional WiFi system.

[0052] Figure 2 is a transmission architecture diagram of a WiFi backscattering system.

[0053] Figure 3 is a channel decoupling and channel equalization algorithm flowchart.

[0054] Figure 4 is a WiFi backscattering system modulation and demodulation flowchart in the embodiment of the present application.

[0055] Figure 5 is a basic principle diagram of channel decoupling in the embodiment of the present application.

[0056] Figure 6 is a backscattering link transmission process modeling diagram in the embodiment of the present application.

[0057] Figure 7 is a generation process diagram of matrix M in the embodiment of the present application.

[0058] Figure 8 This is a schematic diagram of a backtracking-based channel estimation and equalization method in an embodiment of the present invention.

[0059] Figure 9 This is a diagram showing the forward channel evaluation results in an embodiment of the present invention.

[0060] Figure 10 This is a diagram showing the backscatter channel evaluation results in an embodiment of the present invention.

[0061] Figure 11 These are channel response diagrams for different channel models in embodiments of the present invention.

[0062] Figure 12 This is a comparison chart of tag decoding error rates under different channel models in embodiments of the present invention.

[0063] Figure 13 This is a performance comparison chart between the present invention and the traditional WiFi channel evaluation method.

[0064] Figure 14 This is a phase difference constellation diagram for a traditional WiFi channel evaluation method. In the diagram, I represents the in-phase component and Q represents the quadrature component.

[0065] Figure 15 This is a phase difference constellation diagram of the channel evaluation method in this embodiment of the invention. In the diagram, I represents the in-phase component and Q represents the quadrature component.

[0066] Figure 16 This is a flowchart of the method in an embodiment of the present invention. Detailed Implementation

[0067] A preferred embodiment of the present invention will now be described in detail with reference to the accompanying drawings.

[0068] like Figure 16 As shown, a channel decoupling and equalization method for WiFi backscattering in this invention includes the following steps:

[0069] S1, Obtain the original signal from the transmitting end via the original channel. The reference signal obtained later The original signal from the transmitting end passes through the forward channel. The tag then receives and modulates the tag data, which is then transmitted via the backscatter channel. Emit backscatter signal , obtain The backscattered signal includes a VHT-LTF field and a reference symbol; the reference symbol is the first OFDM WiFi symbol in the data field of the backscattered signal that has been labeled with a known tag sequence; the VHT-LTF field is the field in the preamble used for channel estimation, and the tag does not modulate this field;

[0070] S2, Construct a set of equations based on the reference signal and the backscattered signal. The set of equations includes 56 equations constructed using the VHT-LTF field and 64 equations constructed using the reference symbol.

[0071] S3, based on the reference symbol and the known tag sequence, matrix M is obtained through frequency-domain to time-domain transformation, time-domain tag modulation, and time-domain to frequency-domain transformation;

[0072] S4. Solve the forward channel equations by combining the given system of equations and matrix M. Special solution and backscatter channel Special solution ;

[0073] S5 uses a dual-end channel equalization method to apply equalization to the original signal. The first equalization signal is obtained, and an equalization signal is applied to the backscattered signal. The second equalization signal is obtained, and the phase difference between the first equalization signal and the second equalization signal is calculated to decode the tag data.

[0074] The following sections will first introduce the transmission flow of the backscatter system, then describe the design method for conformal channel decoupling, and finally introduce the design method for dual-end channel equalization. The system structure is as follows: Figure 3 As shown, Figure 3 This demonstrates the overall process of algorithm design. First, the two signals received by the two receivers R1 and R2 are analyzed. and Perform channel decoupling and obtain the forward channel. and backscatter channel Then, a dual-ended channel equalization module is used to eliminate the influence of the channel, resulting in a channel-equalized signal. and Then, the phase difference between the two signals is calculated, and finally the tag data is obtained.

[0075] 1. OFDM WiFi backscattering system.

[0076] like Figure 4 As shown, a typical OFDM WiFi backscatter system consists of three parts. The transmitter provides the ambient signal. The tag uses the ambient signal as a carrier for data modulation. The system includes two receivers: the first receiver R1 receives data transmitted through the original channel. Original signal after transmission As reference signal The second receiving end R2 receives the modulated backscatter signal, which is obtained after the original signal is transmitted through the forward channel , modulated by the tag, and finally transmitted through the backscatter channel . The tag embeds information by imposing a phase shift on the time-domain signal transmitted in the air. The receiving end decodes the tag data by comparing the phase difference between the backscatter signal and the reference signal and according to the code table. To ensure that the WiFi signal can be correctly identified, the tag only modulates the data field, while keeping the preamble field unchanged. In addition, the tag will shift the modulated signal to a different frequency band to avoid interference with the original signal. Therefore, the backscatter channel after frequency shift is independent of the forward channel , but the two channels exist in the form of a composite channel at R2. In addition, the receiving end R1 is located differently from the tag, so the original channels are also independent of each other. The traditional channel assessment method in WiFi can only assess the information of the composite channel, and cannot assess the two channels independently, so the present application proposes a new method of composite channel decoupling (Fascatter).

[0077] 2. Composite channel decoupling

[0078] First, the process of modeling the channel is introduced, and then the method of channel decoupling is introduced.

[0079] Channel modeling: WiFi is a wideband signal, and the channel has frequency selectivity, that is, it exhibits different channel response characteristics at different frequency bands. The multi-carrier modulation technology proposed by OFDM WiFi makes full use of this channel characteristic, realizes high-speed and high-capacity data transmission, and can effectively combat multipath interference. Specifically, in OFDM WiFi signals, the wideband signal is decomposed into multiple mutually orthogonal narrowband subcarriers, and the channel response of each subcarrier can be approximately modeled as flat fading, that is, all frequency components in this narrowband frequency band experience the same attenuation during transmission. At this time, the channel response exhibits constant gain and linear phase within the bandwidth range, which can be represented by a complex constant. Different subcarriers experience different channel fading, so the OFDM WiFi channel can be represented by a complex vector, and the length of the vector is the number of subcarriers.

[0080] For example, in 802.11n, one OFDM WiFi symbol contains 64 subcarriers, among which there are 8 null subcarriers, 4 pilot subcarriers and 52 data subcarriers. The null subcarriers do not carry signals and have an energy close to 0, and are almost not affected by the channel, and the channel response corresponding to the positions of these subcarriers is close to 0. The pilot subcarriers and the data subcarriers contain effective information, and the channel corresponding to each subcarrier is represented by a constant . Therefore, the forward channel can be modeled as , which has 56 unknowns. After tag modulation, the bandwidth of the backscatter signal is widened, and the original null subcarriers are no longer null, that is, there is information on all 64 subcarriers. At this time, when the backscatter signal passes through the backscatter channel, all 64 subcarriers will be affected by the channel. Therefore, the channel states of all subcarriers need to be evaluated, and the backscatter channel is modeled as , which contains 64 unknowns.

[0081] Channel decoupling: As can be seen from the above modeling, evaluating two channel states is converted into solving 120 unknowns (56+64=120). As shown in Figure 5 , the basic idea to solve this problem is to construct a sufficient number of equation systems to solve the unknowns. The traditional channel estimation utilizes the VHT-LTF field, which can provide information of the composite channel and can generate 56 independent equations. The VHT-LTF field received by the second receiving end R2 can be represented as:

[0082] ; (1)

[0083] wherein is the predefined VHT-LTF field of the sending end, only 56 data information of which is considered in the calculation . is a 56-dimensional unit matrix (because this field is not modulated by the tag), and represent vectors carrying 56 effective channel variables. Among them, 56 represents the number of channel variables corresponding to the 56 non-null subcarrier frequency bands (52 data subcarriers and 4 pilot subcarriers) of the 64 subcarriers of the OFDM WiFi signal; the remaining 8 null subcarriers do not carry signal energy and are almost not affected by the channel, so the VHT-LTF field does not evaluate the channel variables of the null subcarrier frequency bands.

[0084] , which represents the channel variables of the forward channel on the 56 non-null subcarriers, , representing the channel variables of the backscatter channel on 56 non-empty subcarriers. Since the VHT-LTF field is not modulated by the tag, the position corresponding to the empty subcarrier in the received VHT-LTF field is still approximately empty. Only this is considered in the calculation. Data information on 56 non-empty subcarriers of the China-Africa 0 Since the data at the position corresponding to the empty subcarrier in the above variables are all zero, 56 equations as shown in formula (2) can be constructed using the VHT-LTF field.

[0085] Figure 5 This demonstrates the basic principle of channel decoupling, which can be solved using two fields in the WiFi data packet. One is the VHT-LTF field in the pilot, which is unmodulated by the tag; this field can be used to construct a system of 56 equations. The other is the first symbol in the data field; this symbol is used as a reference symbol, and a known tag sequence is modulated onto it; this field can be used to construct a system of 64 equations. Solving these 120 equations simultaneously allows us to determine the two channel states. and .

[0086] The 56 equations corresponding to the VHT-LTF field are:

[0087] (2)

[0088] This represents the data corresponding to the j-th subcarrier in the VHT-LFT field of the backscattered signal. This represents the channel variable of the backscatter channel corresponding to the j-th subcarrier. This represents the channel variable of the forward channel corresponding to the j-th subcarrier. This represents the data corresponding to the j-th subcarrier in the predefined VHT-LTF field. The value of j ranges from 5 to 32 and from 34 to 61, for a total of 56 positions.

[0089] However, using only this one field is insufficient to meet the requirements for the number of equations. Therefore, it is necessary to find other fields to assist in constructing more equations.

[0090] Considering that the 64 channel parameters of the backscatter channel need to be observed through the modulated field, this invention selects the first OFDM WiFi symbol in the data field of the backscatter signal data packet as the reference symbol. The tag applies known tag modulation data to this reference symbol. The reference symbol received by the second receiver R2 is:

[0091] (3)

[0092] in, is a 64-dimensional non-zero vector , due to the influence of tag modulation, the null subcarriers are no longer null, and each of the 64 subcarriers carries valid data, is a reference symbol in the original signal generated by the sending end, The information can be obtained from the first receiving end R1. is known 64-bit tag modulation data, . and denotes a vector containing 64 subcarrier channel variables, The position of the null subcarrier is represented by 0. Since the channel acts on the frequency domain and the tag modulation acts on the time domain, the influence of the tag modulation cannot be simply represented by the product shown in formula (2). Therefore, the present application models the transmission process of the reference field in the backscatter link in detail, as shown in formula (3). Figure 6 .

[0093] The transmission process is divided into three steps. In the first step, the reference symbol in the original signal is transmitted through the forward channel , the original signal and the channel influence corresponding to the subcarrier position are in a product relationship, and the output of this step is a 64-dimensional vector. In the second step, the tag is independently modulated at each sampling point in the time domain, and this process is represented by the symbol. This step includes a frequency domain to time domain conversion and a time domain to frequency domain conversion, and the detailed process will be described later. The output of this step is still a 64-dimensional vector, and the output result can be represented as Mx , M is a [64, 64] dimensional matrix obtained by and operated by, the column corresponding to the null subcarrier in M is 0, i.e. the elements in columns [1-4, 33, 62-64] are 0, is a self-defined operator symbol, representing the process of frequency domain to time domain conversion, time domain tag modulation and time domain to frequency domain conversion of the reference symbol and the known tag sequence. In the third step, the modulated signal passes through the backscatter channel , the channel in this step has the same effect as the first step and is equivalent to a product. The 64 equations corresponding to this field are:

[0094] ; (4)

[0095] In summary, formulas (2) and (4) construct a total of 120 independent equations, which can be solved together to solve the 120 unknown variables. Among them, can be obtained from the first receiving end R1, and can be obtained from the second receiving end R2, is a known tag sequence, and are unknowns to be solved, and matrix M is unknown. Therefore, the key to solve this problem is to obtain matrix M. As shown above, matrix M is obtained by a specific operation of frequency domain reference symbols and time domain modulation marker data . In order to obtain M, the present application models the signal processing process in detail using matrix operations, as shown in Figure 7 .

[0096] Figure 7 The generation process of matrix M is shown, which is solved by modeling the changes of the signal in the transmission process. The frequency domain signal sent by the sending end is first transformed into the time domain by inverse Fourier transform (IFFT), and then the tag is phase-modulated in the time domain. The modulated signal is transformed into the frequency domain by Fourier transform (FFT).

[0097] First, the sending end arranges the subcarriers in ascending order Shift() when constructing the OFDM WiFi symbol, and adjusts the zero subcarrier to the first position. The symbol after rearrangement is Shift( ) . Then the sending end converts the signal from the frequency domain to the time domain by inverse Fourier transform (IFFT). This process can be modeled as , where represents the dot product of each row in F1 and the Shift( ) vector, and F1 is the coefficient matrix in inverse Fourier transform (IFFT):

[0098] ; (5)

[0099] The dimension of this matrix is [64, 64].

[0100] IFFT formula cannot be used directly here, because the output of this operation is a 64-dimensional vector, while the present application needs a [64, 64] coefficient matrix, and the subsequent Fourier transform (FFT) process is similar.

[0101] Then, the tag applies phase modulation to each sampling point of the time domain signal. This process can be modeled as , where represents the multiplication of the i-th row of and the i-th element in T. Finally, the receiving end converts the received modulated signal from the time domain back to the frequency domain using FFT, and rearranges the order of the subcarriers to restore the zero subcarrier to the center position. This process can be modeled as , where Shift2() performs matrix multiplication, and Shift2() performs subcarrier rearrangement. Here, because... It is a matrix of size [64, 64], and both rows and columns of the matrix need to be rearranged.

[0102] F2 is the coefficient matrix of the FFT process, with dimensions [64, 64].

[0103] (6)

[0104] In summary, matrix M is generated as follows:

[0105] (7)

[0106] Since M is a [64, 64] matrix, Shift2() means rearranging both row and column vectors.

[0107] Figure 7 The diagram illustrates the process of generating matrix M using a 4-dimensional vector as an example. , The coefficient matrix F1 is:

[0108] ;

[0109] The coefficient matrix F2 is:

[0110] .

[0111] The calculation results for each step are as follows: Figure 7 As shown in the image. In The label data '1' indicates that the phase will be flipped. 1 represents the label data '0', indicating that the phase remains unchanged.

[0112] After obtaining M, the solution can be obtained by simultaneously solving formulas (2) and (4). and First, solve... Extract the equation corresponding to the non-empty subcarrier position in formula (4) and combine it with formula (2) to eliminate variables. Taking the 5th subcarrier as an example, we can obtain the result from formula (2). Substituting this into the fifth equation in formula (4) yields:

[0113] (8)

[0114] After combining like terms, we get:

[0115] (9)

[0116] After the above calculations, a matrix with dimensions [56, 56] can be obtained. And theoretically .

[0117] because It is a non-zero vector, when the matrix When the rank is less than 56, that is, Rank( )<56, for The null space. However, in actual calculations, due to factors such as computational precision, the result of matrix multiplication can only approach 0 and cannot be completely equal to 0, that is... Therefore, this invention calculates the eigenvectors and obtains the minimum eigenvalue. To solve this problem, we will use this value as... The approximate particular solution, i.e. and There was a multiple relationship before. K is a constant coefficient. In obtaining... Then, by substituting it into formula (4), the particular solution of the backscattering channel can be obtained. :

[0118] (10)

[0119] For reference.

[0120] 3. Dual-end channel equalization.

[0121] Currently, the forward channel and the backscatter channel have been successfully decoupled, and the corresponding particular solutions are as follows: and Considering that the addition of tags disrupts the linear superposition characteristic of the channel, this invention designs as follows: Figure 8 The dual-end channel equalization method shown starts with the original signal and the backscattered signal respectively, and performs channel equalization by combining the estimated channel state characteristics. The original signal incorporates the influence of the forward channel, while the backscattered signal removes the influence of the backscattered channel. This method allows us to obtain the signals before and after tag modulation. The only difference between these two signals is the tag phase shift; therefore, the phase difference can be calculated to decode the tag data. The reference signal... Without label modulation, the original channel can be equalized using traditional channel estimation and equalization methods. The influence of this, therefore, can be utilized using a reference signal. And evaluate the original channel Obtain the original signal .

[0122] likeFigure 8 The signals before and after the label modulation are respectively shown as and The channel equalization is respectively performed from the sending end and the second receiving end R2. The original signal of the sending end (multiplied by the first receiving end R1) is multiplied by the particular solution of the forward channel to obtain The backscattering signal received by the second receiving end R2 is divided by the particular solution of the backscattering channel to obtain Since K is a constant coefficient, only the amplitude of the signal is changed, and the phase is not changed, the label information can be solved by calculating the phase difference of each sample point pair in the and signal, and the label data modulation and demodulation at the single sample point level are realized.

[0123] 4. The present application shows the effectiveness of the method from two aspects. First, the present application compares the comparison results of the evaluated channel and the real channel. Second, the decoding performance of the method of the present application is compared with that of the HiScatter method in different channel environments. Finally, the decoding performance of the method of the present application is compared with that of the traditional channel evaluation and equalization method.

[0124] (1) Channel evaluation results:

[0125] The channel response obtained by the new channel decoupling method of the present application is compared with the real channel response. The experimental results show that the evaluated channel response and the real result have the same trend, and the gain size is different. This is because the obtained result is a particular solution, and there is a certain multiple relationship with the real solution. Figure 9 The result comparison of the forward channel is shown. The measured channel gain is 7 dB higher than the real result. Figure 10 The result comparison of the backscattering channel is shown. The measured channel gain is 7 dB lower than the real result. This result is consistent with the theoretical analysis in the foregoing. When the particular solution of the forward channel is K times of the real result, the particular solution of the backscattering channel is 1 / K of the real result, that is, and .

[0126] (2) Decoding performance comparison of the present application and the HiScatter method under different channel models:

[0127] MATLAB's wlanTGnChannel tool provides different channel models. This invention selected four representative channel environments for performance testing, as shown in Table 1. From Model A to Model D, four propagation scenarios were simulated: flat fading, indoor residential, small office, and office. The number of taps and delay time increased sequentially with increasing scenario complexity. These parameters were used to model the OFDM WiFi channel response:

[0128] (11)

[0129] Where L is the number of taps (number of propagation paths). It is the time-varying complex gain of the k-th path. This represents the delay of the k-th path, where t is a time variable. It is a time delay function.

[0130] Table 1, Channel Model Parameters:

[0131]

[0132] Specifically, Model A describes an ideal flat fading propagation scenario with only a single tap, where both the root mean square delay spread (RMS) and the maximum delay are zero. Figure 11 As shown in the WiFi bandwidth diagram, Model A has the same effect on all frequency bands. However, this characteristic does not exist in real-world environments. From Model B to Model D, the responses of each subcarrier are different. The channel response model becomes more complex as the number of taps and latency increase.

[0133] The Hiscatter system equalizes the additional phase difference introduced by the wireless channel by phase differentially adjusting the phase of adjacent sampling points. This method approximates the channel model as flat fading, in which case the phase shift introduced by all sampling points is the same and can be eliminated using differential methods. However, in complex channel environments where the channel response on different subcarriers differs, the impact on each sampling point varies, rendering this method ineffective. The method proposed in this invention accurately evaluates the channel response of each subcarrier on the OFDM WiFi wireless channel and can adapt to complex channel models. Therefore, this invention compares the bit error rate of the proposed channel equalization method with that of the method proposed in Hiscatter under different channel models. Figure 12The results shown show that the Hiscatter method can only correctly decode under Model A with a bit error rate (BER) of 1.17%, and the method fails in the multi-tap channel (models B / C / D) with a bit error rate exceeding 42%. In contrast, the method of the application not only outperforms HiScatter (0.22% bit error rate) in Model A, but also adapts to diversified complex channels, with a bit error rate still remaining at a low level of 2.88% in the scenario of channel model D. This shows that for broadband frequency-selective fading signals such as OFDM WiFi, it is necessary to model each subcarrier in detail.

[0134] (3) Performance comparison with traditional channel estimation methods:

[0135] The decoding performance of the method of the application and the traditional channel estimation and equalization is compared through experiments, and the experimental results show that the decoding accuracy of the method of the application is much higher than that of the traditional method, because the method of the application correctly equalizes the influence of the channel. As shown in Figure 13 , the bit error rate (BER) of the method of the application is 0.97%, while the bit error rate (BER) of the traditional method is as high as 21.54%. The reason is as shown in Figure 14 and Figure 15 , the phase difference obtained by the traditional method is randomly distributed in the interval [0, 2 ], and does not have distinguishability, so the decoding error rate of the method is high. In comparison, the phase difference after correct channel equalization of the application is concentrated at 0 and , which is similar to the theoretical result, so the method can correctly decode.

[0136] Embodiment:

[0137] First, an OFDM WiFi backscatter system is built, and the overall structure of the system is as shown in Figure 4 , which consists of three parts: a sending end, a tag, and a receiving end. The sending end and the first receiving end use commercial WiFi equipment, use a Dell notebook with a Qualcomm AR938x network card, use CommView as a transceiver software, and the second receiving end uses a ZedBoard SDR and carries an AD9361 subboard for obtaining physical layer intermediate data, and uses MATLAB software for data analysis. The tag consists of three main parts: a detector equipped with a multi-stage demodulation logarithmic amplifier AD8313 and a threshold voltage tuning circuit TLV3501; an XILINX Artix-7 XC7A35T FPGA for tag data modulation and logic control; and an RF switch ADG902 controlled by a mixed-mode clock manager (MMCM).

[0138] The flow chart of the overall system is as shown in Figure 3As shown, first, two signals received by two receiving ends R1 and R2 are used to calculate the phase difference between the two signals and channel decoupling is performed to obtain the channel state of the forward channel and the backscatter channel and then, a double-end channel equalization module is used to eliminate the influence of the channel to obtain the channel equalized signals and then, the phase difference between the two signals is calculated, and finally, the tag data is solved.

[0139] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used herein, the term "includes" and tautological expressions thereof, such as "including", "includes", "include", "including", "include", "including", and the like, mean the existence of

[0140] It should be understood that, although each step in the flowchart of the drawings of the specification is shown in sequence according to the direction of the arrow, these steps are not necessarily executed in sequence according to the direction of the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other order. Moreover, at least part of the steps in the flowchart of the drawings of the specification can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or steps or stages in other steps.

[0141] In one embodiment, the present application provides a computer system, which can be a server. The computer system comprises a processor, a memory and a network interface connected by a system bus. Among them, the processor of the computer system is used to provide computing and control capability. The memory of the computer system comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer system is used to store the data used in the above method. The network interface of the computer system is used to communicate with the external terminal through the network connection. The computer program is executed by the processor to realize the above method.

[0142] The technical features of the above embodiments can be combined in any way. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0143] It will be obvious to a person skilled in the art that the application is not limited to the details of the foregoing exemplary embodiments and can be implemented in other concrete forms without departing from the spirit or essential characteristics of the application. The embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein and no

[0144] Furthermore, it should be understood that although the description is made on the basis of the embodiments, not every embodiment contains only one independent technical solution, and the description is made in this way only for the sake of clarity, and a person skilled in the art should consider the description as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by a person skilled in the art.

Claims

1. A channel decoupling and equalization method for WiFi backscattering, characterized in that, include: The original signal from the transmitter is obtained through the original channel. The reference signal obtained later The original signal from the transmitting end passes through the forward channel. The tag then receives and modulates the tag data, which is then transmitted via the backscatter channel. Emit backscatter signal , obtain The backscattered signal includes the VHT-LTF field and the reference symbol. The reference symbol is the first OFDM WiFi symbol in the data field of the backscattered signal that has been labeled with a known tag sequence; A set of equations is constructed based on the reference signal and the backscattered signal. The set of equations includes 56 equations constructed using the VHT-LTF field and 64 equations constructed using the reference symbol. Based on the reference symbol and the known tag sequence, matrix M is obtained through frequency-domain to time-domain transformation, time-domain tag modulation, and time-domain to frequency-domain transformation; Solving the system of equations simultaneously, and combining it with matrix M, yields the solution for the forward channel. Special solution and backscatter channel Special solution ; A dual-channel equalization method is used to apply equalization to the original signal. The first equalization signal is obtained, and an equalization signal is applied to the backscattered signal. The second equalization signal is obtained, and the phase difference between the first equalization signal and the second equalization signal is calculated to decode the tag data. in, It is a 64-dimensional vector containing 56 unknown channel variables; For the forward channel corresponding to the j-th subcarrier; It is a 64-dimensional vector containing 64 unknown channel variables; Let be the channel variable of the backscatter channel corresponding to the i-th subcarrier; VHT-LTF field of backscattered signal ;VHT-LTF field predefined by the sending end In calculation At that time, only consider Data information of 56 non-empty subcarriers; It is a 56-dimensional identity matrix. It is a vector consisting of the channel variables of the forward channel corresponding to 56 non-empty subcarriers. for The vector consisting of the channel variables in the backscatter channel corresponding to the 56 channel variables in the data; The 56 equations corresponding to the VHT-LTF field are: ；（2） for The data corresponding to the j-th subcarrier, for The data corresponding to the j-th subcarrier; Reference symbol of backscattered signal ; The reference symbol of the original signal generated by the transmitting end. ; It is a known 64-bit tag sequence; Indicates inclusion The vector of channel variables for the 64 subcarriers. , Indicates inclusion A vector of channel variables for 64 subcarriers. ; The 64 equations constructed using reference symbols are as follows: ；（4） express Data at the i-th subcarrier position, express Data at the i-th subcarrier position, This represents the data in the i-th row and j-th column of M; The sending end is constructing When transmitting OFDM WiFi symbols, the zero subcarrier is located in the center position. Beforehand, the transmitter sorts the subcarriers of the OFDM WiFi symbol in ascending order using Shift(), and then rearranges them... The OFDM WiFi symbol is Shift ( Then, the transmitting end uses an inverse Fourier transform to convert the reference signal from the frequency domain to the time domain: ; F1 is the coefficient matrix in the inverse Fourier transform. This indicates that each line in F1 is related to Shift( Perform dot product; for The time-domain signal; The tag applies phase modulation at each sampling point of the time-domain signal: ; express The i-th row and the known label sequence in the time domain Multiply by the i-th element in the matrix. The signal is modulated; The receiving end uses Fourier transform to convert the received signal into a signal. Transform back to the frequency domain from the time domain and rearrange the order of the subcarriers. By restoring the zero subcarrier to its center position, we obtain matrix M: ； For matrix multiplication, F2 is the coefficient matrix in the Fourier transform; F1 is: ; This represents the index of the k-th subcarrier, corresponding to the index of the row in F1; This represents the index of the nth time-domain sampling point, corresponding to the column index in F1. , This indicates the row and column numbers in F1. ; The imaginary unit; F2 is: ; Solve Extract the equation corresponding to the non-empty subcarrier position in formula (4) and combine it with formula (2) to eliminate... This yields a matrix with dimensions [56, 56]. And theoretically ;because It is a non-zero vector, when the matrix When the rank is less than 56, for The null space; by calculating the eigenvectors and finding the smallest eigenvalue. To solve, using As a particular solution of the forward channel, i.e. and There was a multiple relationship before. K is a constant coefficient; after obtaining Then, substituting into formula (4), we can obtain Special solution : ； The signal before tag modulation is denoted as The signal modulated by the tag is used as The original signal is multiplied by get The backscattered signal received at the receiving end is divided by get ; through calculation and The phase difference between each pair of sampling points in the signal is used to solve for the tag data.

2. A computer system comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method of claim 1.

Images