Frequency division multiplexing RFID tag signal parallel receiving method based on SIC architecture

Abstract

The invention discloses a frequency division multiplexing RFID tag signal parallel receiving method based on an SIC architecture, which comprises the following steps: firstly, constructing a frequency division multiplexing UHF RFID system, performing subcarrier modulation on respective Miller codes by tags by using M values of 4, 8, 12 and 16, deducing the power spectrum density of a Miller subcarrier modulation sequence based on the Wiener-Singqin theorem, and finally, performing frequency division multiplexing on the Miller subcarrier modulation sequence according to the power spectrum density of the Miller subcarrier modulation sequence. The method comprises the following steps of: quantitatively analyzing the interference degree generated by tags of other channels in different frequency channels, performing power estimation through FFT (Fast Fourier Transform), estimating the amplitude and radio frequency phase of each tag by utilizing the known interference degree, calculating the SIR (Signal to Interference Ratio) of each tag, sequencing, and demodulating in sequence according to the SIR size until the last tag signal is demodulated. According to the method provided by the invention, the parallel receiving of the frequency division multiplexing RFID tags can be realized, so that the reader-writer can simultaneously inventory a plurality of tags in a single time slot, thereby effectively improving the identification efficiency and throughput rate of the UHF RFID system, and improving the ability of the UHF RFID system to adapt to application scenarios in which massive tags need to be quickly accessed in a large-scale Internet of Things.

Description

Technical Field

[0001] The present invention belongs to the technical field of radio frequency identification, and in particular relates to a method for parallel receiving frequency division multiplexing RFID tag signals based on a SIC architecture. Background Art

[0002] Radio Frequency Identification (RFID) technology is an automatic identification technology that performs contactless data communication between readers and passive tags. It is widely used in logistics management, inventory control, intelligent transportation systems and other scenarios, and is the basic support for building the Internet of Things. With the continuous increase in RFID devices and applications, higher requirements are placed on the terminal capacity and coverage of RFID systems.

[0003] When multiple tags are within the recognition range of the reader at the same time, tag collisions are likely to occur, resulting in reduced recognition efficiency and data loss. The current anti-collision algorithms mainly include tree structure algorithms and Aloha algorithms, both of which are based on time division multiplexing. The QT algorithm is a typical tree structure algorithm with a throughput of approximately 0.348. The maximum throughput of the dynamic Aloha algorithm is only 0.368. However, the communication distance and coverage of the reader are constantly increasing, and the number of tags connected is also increasing. However, the signal bandwidth returned by the tag is limited, and the total uplink throughput of the system is difficult to improve within the time slot returned by a single tag. The anti-collision algorithm based on the time division multiplexing mechanism has reached a bottleneck in improving the throughput.

[0004] At present, UHF RFID systems generally use the ISO / IEC 18000-6C protocol, which stipulates that a single reader can only communicate with a tag using one frequency channel through time division multiplexing. However, the ISO / IEC 18000-6C protocol allows tags to use Miller coded subcarrier modulation. Different tags select different M values ​​to achieve modulation on different subcarrier frequencies, making it possible to achieve uplink frequency division multiplexing on the tag side.

[0005] Due to the tag backscatter communication mechanism, it is impossible to filter and shape the reflected signal to suppress sidelobe / harmonic interference, resulting in interference between different channels. Therefore, solving the interference introduced by multiple tags using frequency division multiplexing technology to respond to reader commands at the same time is one of the key issues in improving the throughput of RFID systems. Summary of the invention

[0006] To solve the above technical problems, the present invention provides a method for parallel reception of frequency-division multiplexed RFID tag signals based on a SIC (Successive Interference Cancellation) architecture. The power spectral density of the Miller subcarrier modulation sequence is derived using the Wiener-Hinchin theorem, and the degree of interference caused by tags in other channels in different frequency channels is quantitatively analyzed. The SIC mechanism is introduced into the frequency-division multiplexed UHF RFID system to reduce the sidelobe and harmonic interference of the tag signal.

[0007] The technical solution adopted by the present invention is: a method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture, and the specific steps are as follows:

[0008] S1. Build a frequency division multiplexing UHF RFID system, including: 4 RFID tags and 1 reader / writer;

[0009] The reader sends a 920-925 MHz tag excitation signal through a transmitting antenna to activate four RFID tags. After receiving an inventory command from the reader, the four RFID tags simultaneously return a 920-925 MHz signal to the reader receiving antenna.

[0010] The reader / writer includes: a baseband signal processing module and a radio frequency front end. The baseband signal processing module includes: a downsampling module, a bandpass filter BLF, an FFT module, a power estimation module, an amplitude estimation module, a SIR calculation and sorting module, a radio frequency phase compensation module, a Costas loop, a decoding module, and a signal reconstruction module. The radio frequency front end includes: an orthogonal modulator and an ADC.

[0011] S2, the four RFID tags respectively use M values ​​of 4, 8, 12, and 16 to perform subcarrier modulation on their respective Miller codes to form Miller subcarrier modulation sequences;

[0012] After the reader sends the inventory command, the four RFID tags communicate and send the Miller subcarrier modulation sequence to the reader; the bandwidth of each frequency channel is the first zero point bandwidth of the RFID tag signal, that is, 4 times the data rate of the tag baseband signal.

[0013] S3, after the reader receives the signals returned by the four RFID tags at the same time, it first performs down-conversion and ADC sampling through the RF front end to convert the analog signal into a digital signal;

[0014] S4, performing power estimation on the total signal received by the reader through FFT;

[0015] First, estimate the signal power in each frequency band The power spectral density of the Miller subcarrier modulation sequence with M values ​​of 4, 8, 12, and 16 derived by the Wiener-Schinchin theorem is used. The power spectral density is numerically integrated to obtain the power of each Miller subcarrier modulation sequence in different frequency bands, and then the interference degree H caused by the tags in other channels in different frequency channels is obtained. Finally, the estimated value of the actual first zero point bandwidth power of each tag is obtained. Complete power estimation.

[0016] Where H is a 4×4 matrix, and the element h in the i-th row and j-th column is ij represents the ratio of the power of the RFID tag located in the jth frequency channel in the ith frequency channel to the first zero-point bandwidth power of the tag in the jth frequency channel, and

[0017] S5, based on step S4, according to Parseval's theorem, performing amplitude estimation and RF phase estimation;

[0018] S6. Based on step S4, after power estimation is completed, the SIR of each tag is calculated and the SIRs are sorted;

[0019] S7. Based on step S6, first perform SIC demodulation on the tag with the largest SIR;

[0020] First, it passes through a FIR bandpass filter to filter out out-of-band interference, then passes through the RF phase compensation module to offset the influence of the RF phase, and then inputs the signal into the Costas loop. The subcarrier signal and Miller code of the tag are obtained through the Costas loop, and then input into the decoding module to judge the Miller code to obtain the original data of the tag, and input into the signal reconstruction module to obtain the reconstructed tag signal, that is, to reconstruct the Miller subcarrier modulation sequence of the tag.

[0021] Among them, Costas generates a sinusoidal signal with the same frequency and phase as the original Miller subcarrier in an internal digitally controlled oscillator, and outputs a Miller coded waveform through an internal low-pass filter. The Miller coded waveform is polarized to obtain the Miller code.

[0022] S8, repeating the demodulation process based on step S7 until the last tag signal is demodulated, completing the RFID tag parallel reception;

[0023] Based on the reconstructed tag signal obtained by SIC demodulation of the tag with the largest SIR in step S7, the reconstructed signal is divided into I and Q paths using the estimated RF phase, the reconstructed tag signal is subtracted from the I and Q paths, and then the tag signal with the second largest SIR is demodulated. The demodulation process is repeated until the last tag signal is demodulated, completing the final reception process, that is, completing the parallel reception of RFID tags.

[0024] Furthermore, the M value used by the four RFID tags is not limited to 4, 8, 12, and 16, and can be expanded according to actual conditions.

[0025] Furthermore, the step S3 is specifically as follows:

[0026] The down-converted tag signal X received by the reader receiving antenna RF (t) is expressed as follows:

[0027]

[0028] Where t represents time, T i (t) = M i (t)Square i (2πf i t) represents the Miller subcarrier modulation sequence of tag i, where tag i=1, 2, 3, 4 represent tags with M values ​​of 4, 8, 12, and 16, respectively. i (t) represents the baseband Miller code, Square i (2πf i t) represents a square wave signal, a i , t i , and f i They represent the amplitude, RF phase, transmission delay, and subcarrier frequency of tag i respectively. N(t) represents additive white Gaussian noise.

[0029] Then X RF (t) After ADC sampling, the received complex signal X(n) is expressed as follows:

[0030]

[0031] Where n represents a discrete time series, f s represents the ADC sampling frequency; and the reader receiving end is orthogonal demodulation, then the I and Q components of X(n) are rewritten as follows:

[0032]

[0033] Among them, the two real factors and Includes the effect of RF phase.

[0034] Furthermore, the step S4 is specifically as follows:

[0035] The power of the I and Q digital signals after ADC sampling in step S3 is estimated by FFT. First, the power spectral density PSD of the Miller subcarrier modulation sequence with M values ​​of 4, 8, 12, and 16 derived by the Wiener-Hinchin theorem is used. The estimated calculation expression of PSD is as follows:

[0036]

[0037] Among them, D I (n), D Q (n) represents the estimated power spectrum density of I and Q paths respectively, represents Fourier transform, and N represents the length of Fourier transform.

[0038] In the FFT module, the reader calculates the power estimation value in the four frequency channels through FFT.

[0039]

[0040] in,[·] T Representing a transposition operation, the power estimation value includes the first zero-point bandwidth signal of the label corresponding to each channel and the side lobes or harmonic interference of the labels of the other three channels within the frequency channel.

[0041] For the I path, the power estimate entering the i-th channel is obtained by numerically integrating the power spectral density of the FFT estimate located in the i-th channel. set up represents the estimated value of the actual power of the i-th tag in the first zero-point bandwidth. and There are the following relationships:

[0042]

[0043] in, It represents the estimated value of the actual power of the tag in the four frequency channels of channel I in the first zero point bandwidth, H represents the linear expression of label interference in different channels, and its element h in the i-th row and j-th column is ij represents the ratio of the power of the tag located in the jth frequency channel in the ith frequency channel to the first zero-point bandwidth power of the tag in the jth frequency channel; and It represents the power estimation value in the four frequency channels of channel I.

[0044] Similarly, the power estimate of the i-th channel in the Q path is Estimate of the actual power of the i-th tag in the first zero-point bandwidth

[0045] In summary, the first zero point bandwidth power estimation of each tag distributed in the I and Q paths is calculated Get the power estimate of each frequency channel in I and Q paths

[0046] Furthermore, the step S5 is specifically as follows:

[0047] The power of the first zero point bandwidth of the label accounts for 0.77 of the power of the entire Miller subcarrier modulation signal. According to Parseval's theorem, the power in the time domain is equal to the power in the frequency domain. The estimated amplitude value of the I path is Similarly, the amplitude value of the Q-path estimate can be obtained

[0048] Then use the estimated amplitudes of the I and Q paths to estimate the RF phase , calculated as

[0049] Furthermore, the step S6 is specifically as follows:

[0050] After the power estimation is completed, the SIR of each subchannel label i is estimated by the SIR calculation and sorting module. The calculation expression of the SIR of label i is as follows:

[0051]

[0052] After the reader calculates the SIR of each tag, it determines whether the power value in each channel is lower than the preset threshold based on the power estimation value of each channel obtained in step S5. If it is lower, it is considered that there is no signal in the channel. In the subsequent SIC demodulation process, the channel is not considered. Then the SIRs of the remaining channels are sorted and demodulated in order from large to small SIR. During demodulation, the reader first selects the tag with the largest SIR for SIC demodulation.

[0053] Furthermore, the reader performs SIC demodulation on the current channel and reconstructs the tag signal of the channel. If the reader detects that a bit error occurs in the tag signal, when entering the SIC demodulation of the next level channel, it is divided into two cases, namely, subtracting the reconstructed signal from the remaining total signal and not subtracting the reconstructed signal, and comparing the two cases to see whether a bit error occurs in the SIC demodulation of the next channel; if no bit error occurs in either case, continue to reconstruct the current tag signal for subsequent SIC demodulation; if bit errors occur in both cases, stop SIC demodulation.

[0054] Furthermore, in step S7, the tag is demodulated as follows:

[0055] For each sub-channel label i, a bandpass filter BPF is used to filter out all out-of-band spectrum components, and then an RF phase compensation operation is performed. Then the digital signal X at the input of the Costas loop is i (n) The expression is as follows:

[0056]

[0057] The signal input to the Costas loop is the signal after RF phase compensation. i Indicates that the signal passes through a digital FIR bandpass filter.

[0058] The Costas loop includes: a phase detector, a loop filter and a digitally controlled oscillator; the Costas loop completes the demodulation, carrier phase locking and carrier frequency multiplication of the BPSK signal through two loops. The function of the Costas loop is to first demodulate the label subcarrier modulation sequence, first remove the subchannel carrier f i , use a low-pass filter LPF to filter the signal to remove 2f i The spectrum is then decoded using the zero-threshold decoding rule for bit decision making.

[0059] Among them, the phase detector adopts a multiplication phase detector in series with an FIR low-pass filter structure; the result of the I-way phase detector is multiplied by the sign of the result of the Q-way phase detector as the phase detection result of the entire loop and input into the loop filter; the frequency control word of the numerically controlled oscillator NCO is added to the loop filter output based on the initial control word; the numerically controlled oscillator generates a primary carrier sine and cosine carrier signal through the frequency control word, and generates a tertiary carrier cosine signal through the triple frequency control word; the primary carrier sine and cosine signals are used for the input of the phase detector multiplication end, and the tertiary carrier cosine signal is used as the module output for harmonic interference suppression.

[0060] Then the reconstructed reader receives input X i The tag signal at (n) is used in the loop to generate the sine wave sin[2πf i (n / f s -t i )], the sine wave has the same frequency and phase as the subcarrier of the Miller subcarrier modulation sequence, then the reconstructed Miller code M can be obtained by taking the polarity of the low-pass filter of the Costas loop i (n / f s -t i ), sin[2πf i (n / f s -t i )] After taking the polarity, the reconstructed subcarrier signal Square can be obtained i [2πf i (n / fs -t i )].

[0061] The reconstructed signal is in the form of a square wave, and then a method of using a series of odd harmonics to reconstruct the signal is proposed, that is, the square wave signal Square(2πf s The Fourier expansion expression of t) is as follows:

[0062]

[0063] Then, a finite number of odd-order sine harmonics are used to fit the square wave to obtain the reconstructed Miller subcarrier, which is expressed as follows:

[0064]

[0065] in, and It represents the subcarrier delay and subcarrier frequency obtained by Costas loop synchronization.

[0066] Finally, the sinusoidal signal output by the Costas loop is tripled, quintupled, and 7-folded, and the sinusoidal signal is superimposed with each odd harmonic component to fit the Miller subcarrier signal; the Miller subcarrier signal is multiplied by the Miller code to obtain the reconstructed label signal; the reconstructed signal expression of label i is as follows:

[0067]

[0068] Beneficial effects of the present invention: The method of the present invention first constructs a frequency division multiplexing UHF RFID system, and the tags use M values ​​of 4, 8, 12, and 16 to perform subcarrier modulation on their respective Miller codes to achieve different reverse link frequencies (BLFs), thereby achieving frequency division multiplexing of tag signals, and then derives the power spectral density of the Miller subcarrier modulation sequence with M values ​​of 4, 8, 12, and 16 based on the Wiener-Schinchin theorem, quantitatively analyzes the interference degree generated by tags of other channels in different frequency channels, estimates the power of the superimposed signal in each frequency band by FFT, and then estimates the amplitude and RF phase of each tag by using the known interference degree. When receiving, the reader uses FFT calculation to obtain the signal-to-interference ratio (SIR) of each signal and sorts them. The tag with the largest SIR is demodulated first. When demodulating, the Miller subcarrier modulation sequence of the tag reconstructed by the subcarrier and Miller code is obtained by using the Costas ring, and the reconstructed signal is subtracted from the total signal, and then the tag signal with the second largest SIR is demodulated; the demodulation process is repeated until the last tag signal is demodulated. The method of the present invention can realize the parallel reception of frequency division multiplexing RFID tags, so that the reader can simultaneously inventory multiple tags in a single time slot, thereby effectively improving the recognition efficiency and throughput rate of the UHF RFID system, and improving the ability of the UHF RFID system to adapt to application scenarios that require rapid access to massive tags in large-scale Internet of Things. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 The present invention is a flow chart of a method for parallel reception of frequency division multiplexing RFID tag signals based on a SIC architecture.

[0070] Figure 2 Schematic diagram of the structure of a frequency division multiplexing UHF RFID system in an embodiment of the present invention.

[0071] Figure 3 This is a parallel receiving block diagram of a frequency division multiplexing tag signal reader / writer based on the SIC architecture in an embodiment of the present invention.

[0072] Figure 4 Schematic diagram of the power spectral density (PSD) of a Miller subcarrier modulation sequence with a reverse link frequency (BLF) of 80 kHz when the M value is 4 in an embodiment of the present invention.

[0073] Figure 5 This is a flow chart of a frequency domain channel collision solution in an embodiment of the present invention.

[0074] Figure 6 It is a schematic diagram of the structures of the radio frequency phase compensation module, the Costas loop, and the signal reconstruction module in the parallel receiving block diagram of the reader / writer in an embodiment of the present invention. DETAILED DESCRIPTION

[0075] The method of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0076] like Figure 1 As shown, a flow chart of a method for parallel reception of frequency division multiplexing RFID tag signals based on a SIC architecture of the present invention, the specific steps are as follows:

[0077] S1, such as Figure 2 As shown, a frequency division multiplexing UHF RFID system is constructed, including: 4 RFID tags and 1 reader / writer;

[0078] The reader sends a 920-925 MHz tag excitation signal through a transmitting antenna to activate four RFID tags. After receiving an inventory command from the reader, the four RFID tags simultaneously return a 920-925 MHz signal to the reader receiving antenna.

[0079] like Figure 3 As shown, the reader / writer includes: baseband signal processing module and radio frequency front end. The baseband signal processing module includes: downsampling module, bandpass filter BLF, FFT module, power estimation module, amplitude estimation module, SIR calculation and sorting module, radio frequency phase compensation module, Costas loop, decoding module, signal reconstruction module. The radio frequency front end includes: orthogonal modulator and ADC.

[0080] S2, the four RFID tags respectively use M values ​​of 4, 8, 12, and 16 to perform subcarrier modulation on their respective Miller codes to form Miller subcarrier modulation sequences;

[0081] After the reader sends the inventory command, the four RFID tags communicate and send the Miller subcarrier modulation sequence to the reader; the bandwidth of each frequency channel is the first zero point bandwidth of the RFID tag signal, that is, 4 times the data rate of the tag baseband signal.

[0082] S3, after the reader receives the signals returned by the four RFID tags at the same time, it first performs down-conversion and ADC sampling through the RF front end to convert the analog signal into a digital signal;

[0083] S4, performing power estimation on the total signal received by the reader through FFT;

[0084] First, estimate the signal power in each frequency band , using the power spectral density of the Miller subcarrier modulation sequence with M values ​​of 4, 8, 12, and 16 derived by the Wiener-Schinchin theorem, the power spectral density is numerically integrated to obtain the power size of each Miller subcarrier modulation sequence in different frequency bands, and then the interference degree H generated by the tags of other channels in different frequency channels is obtained, and finally the estimated value of the actual first zero point bandwidth power of each tag is obtained. , complete the power estimation.

[0085] Where H is a 4×4 matrix, and the element h in the i-th row and j-th column is ij represents the ratio of the power of the RFID tag located in the jth frequency channel in the ith frequency channel to the first zero-point bandwidth power of the tag in the jth frequency channel, and

[0086] S5, based on step S4, according to Parseval's theorem, performing amplitude estimation and RF phase estimation;

[0087] S6. Based on step S4, after power estimation is completed, the SIR of each tag is calculated and the SIRs are sorted;

[0088] S7. Based on step S6, first perform SIC demodulation on the tag with the largest SIR;

[0089] First, it passes through a FIR bandpass filter to filter out out-of-band interference, then passes through the RF phase compensation module to offset the influence of the RF phase, and then inputs the signal into the Costas loop. The subcarrier signal and Miller code of the tag are obtained through the Costas loop, and then input into the decoding module to judge the Miller code to obtain the original data of the tag, and input into the signal reconstruction module to obtain the reconstructed tag signal, that is, to reconstruct the Miller subcarrier modulation sequence of the tag.

[0090] Among them, Costas generates a sinusoidal signal with the same frequency and phase as the original Miller subcarrier in an internal digitally controlled oscillator, and outputs a Miller coded waveform through an internal low-pass filter. The Miller coded waveform is polarized to obtain the Miller code.

[0091] S8, repeating the demodulation process based on step S7 until the last tag signal is demodulated, completing the RFID tag parallel reception;

[0092] Based on the reconstructed tag signal obtained by SIC demodulation of the tag with the largest SIR in step S7, the reconstructed signal is divided into I and Q paths using the estimated RF phase, the reconstructed tag signal is subtracted from the I and Q paths, and then the tag signal with the second largest SIR is demodulated. The demodulation process is repeated until the last tag signal is demodulated, completing the final reception process, that is, completing the parallel reception of RFID tags.

[0093] In this embodiment, the M value used by the four RFID tags is not limited to 4, 8, 12, and 16, and can be expanded according to actual conditions, as shown in Table 1.

[0094] In order to facilitate the reader to receive the data returned by the frequency division multiplexing UHF RFID tag in parallel, this embodiment divides the channels based on different M values ​​and BLF at the same data rate. The ISO / IEC 18000-6C standard protocol stipulates that the BLF of the tag code should be in the range of 40 to 640kHz, so the BLF of the remaining channels except the data rate of 110kbps is within this range. And when dividing the channels, it is necessary to minimize the interference between adjacent channels. Based on this, the channel division results are shown in Table 1.

[0095] Table 1

[0096]

[0097] In Table 1, the unit of data rate is kbps, and the unit of BLF is kHz.

[0098] As shown in Table 1, the number of channels that a tag can select in a time slot can be 2, 3, 4, 5, and 8. When the data rate is 5 kbps, 10 kbps, and 20 kbps, there are 8 channels for the tag to select in a time slot. When there is only one channel in a time slot, the tag complies with the provisions of the ISO / IEC 18000-6C standard protocol. Although there is harmonic interference on some channels, the time required to suppress the harmonic interference is less than the time consumed to re-identify the tag, and when the harmonic interference signal is much smaller than the useful signal, the interfered tag can be correctly identified even without harmonic interference. Therefore, after comprehensive consideration, the channel with harmonic interference can be added when dividing the channel.

[0099] In this embodiment, step S3 is specifically as follows:

[0100] The down-converted tag signal X received by the reader receiving antenna RF (t) is expressed as follows:

[0101]

[0102] Where t represents time, Ti (t) = M i (t)Square i (2πf i t) represents the Miller subcarrier modulation sequence of tag i, where tag i=1, 2, 3, 4 represent tags with M values ​​of 4, 8, 12, and 16, respectively. i (t) represents the baseband Miller code, Square i (2πf i t) represents a square wave signal, a i , , t i , and f i They represent the amplitude, RF phase, transmission delay, and subcarrier frequency of tag i respectively. N(t) represents additive white Gaussian noise.

[0103] Then X RF (t) After ADC sampling, the received complex signal X(n) is expressed as follows:

[0104]

[0105] Where n represents a discrete time series, f s represents the ADC sampling frequency; and the reader receiving end is orthogonal demodulation, then the I and Q components of X(n) are rewritten as follows:

[0106]

[0107] Among them, the two real factors and Includes the effect of RF phase.

[0108] In this embodiment, step S4 is specifically as follows:

[0109] The power of the I and Q digital signals after ADC sampling in step S3 is estimated by FFT. First, the power spectral density PSD of the Miller subcarrier modulation sequence with M values ​​of 4, 8, 12, and 16 derived by the Wiener-Schinchin theorem is used. Under the assumption of a stationary signal, FFT is a good approximation of the PSD of the received signal (using a rectangular window). The estimated calculation expression of PSD is as follows:

[0110]

[0111] Among them, D I (n), D Q (n) represents the estimated power spectrum density of I and Q paths respectively, represents Fourier transform, and N represents the length of Fourier transform.

[0112] Due to the tag backscatter communication mechanism, it is impossible to filter and shape the reflected signal to suppress sidelobe / harmonic interference, so the bandwidth of its reflected scattered signal is very wide. Figure 4 is the power spectral density of the Miller subcarrier modulation sequence with an M value of 4 and a reverse link frequency of 80kHz, Figure 4 The side lobes or harmonic components of the tags distributed in other frequency channels will cause interference. The Miller subcarrier modulation sequence belongs to a first-order Markov random process. Based on the Wiener-Schinchin theorem, the power spectral density D4, D8, D 12 ,D 16 , quantitatively analyzed the interference degree caused by tags in other channels in different frequency channels, that is, the interference degree of D4, D8, D 12 ,D 16 Through the method of numerical integration, the power of each Miller subcarrier modulation sequence in different frequency bands is obtained, as shown in Table 2.

[0113] Table 2

[0114] Channel Label 1 Label 2 Label 3 Label 4 Sum 1 \-12.3dB-17.6dB-20.6dB -10.7dB 2 -24.6dB\-13.4dB-18.8dB -12.1dB 3 -9.8dB-20.0dB\-14.0dB -8.1dB 4 -31.0dB-37.8dB-18.4dB\ -18.1dB

[0115] In the FFT module, the reader calculates the power estimation value in the four frequency channels through FFT.

[0116]

[0117] in,[·] T Representing a transposition operation, the power estimation value includes the first zero-point bandwidth signal of the label corresponding to each channel and the side lobes or harmonic interference of the labels of the other three channels within the frequency channel.

[0118] For the I path, the power estimate entering the i-th channel is obtained by numerically integrating the power spectral density of the FFT estimate located in the i-th channel. set up represents the estimated value of the actual power of the i-th tag in the first zero-point bandwidth. and There are the following relationships:

[0119]

[0120] in, It represents the estimated value of the actual power of the tag in the four frequency channels of channel I in the first zero point bandwidth, H represents the linear expression of the label interference in different channels (obtained by converting the values ​​in Table 1 into a linear expression), and its element h in the i-th row and j-th column is ijrepresents the ratio of the power of the tag located in the jth frequency channel in the ith frequency channel to the first zero-point bandwidth power of the tag in the jth frequency channel; and It represents the power estimation value in the four frequency channels of channel I.

[0121] Similarly, the power estimate of the i-th channel in the Q path is , the estimate of the actual power of the ith tag in the first zero bandwidth

[0122] In summary, the first zero point bandwidth power estimation of each tag distributed in the I and Q paths is calculated , get the power estimate of each frequency channel in I and Q paths

[0123] In this embodiment, step S5 is specifically as follows:

[0124] The power of the first zero point bandwidth of the label accounts for 0.77 of the power of the entire Miller subcarrier modulation signal. According to Parseval's theorem, the power in the time domain is equal to the power in the frequency domain. The estimated amplitude value of the I path is Similarly, the amplitude value of the Q-path estimate can be obtained

[0125] Then use the estimated amplitudes of the I and Q paths to estimate the RF phase , calculated as

[0126] In this embodiment, step S6 is specifically as follows:

[0127] The interference of the tags in Table 2 in different channels makes the demodulation of the frequency division multiplexing tag signal very challenging. Therefore, this embodiment proposes that the reader receiver uses the SIC mechanism to perform parallel demodulation of multiple tag signals. Figure 3 After the power estimation is completed, the SIR of each sub-channel tag i is estimated by the SIR calculation and sorting module. The calculation expression of the SIR of tag i is as follows:

[0128]

[0129] In the frequency division multiplexing UHF RFID system, tags randomly select different time slots and M values ​​to communicate with the reader. If multiple tags select the same M value in the same time slot, it will cause frequency domain collision. This embodiment proposes a frequency domain channel collision solution. The solution flow chart is as follows: Figure 5After calculating the SIR of each tag, the reader determines whether the power value in each channel is lower than the preset threshold based on the power estimation value of each channel obtained in step S5. If it is lower, it is considered that there is no signal in the channel. In the subsequent SIC demodulation process, the channel is not considered. Then the SIRs of the remaining channels are sorted and demodulated in order from large to small SIRs. During demodulation, the reader first selects the tag with the largest SIR for SIC demodulation.

[0130] In this embodiment, since the reader cannot calculate the bit error rate of each tag when receiving the tag signal, the mechanism adopted by the reader when receiving is as follows:

[0131] The reader performs SIC demodulation on the current channel and reconstructs the tag signal of the channel. If the reader detects that a bit error occurs in the tag signal (CRC check fails), when entering the SIC demodulation of the next level channel, there are two cases, namely, subtracting the reconstructed signal from the remaining total signal and not subtracting the reconstructed signal. The two cases are compared to see whether a bit error occurs in the SIC demodulation of the next channel; if a bit error does not occur in one of the cases, the current tag signal continues to be reconstructed for subsequent SIC demodulation; if a bit error occurs in both cases, the SIC demodulation is stopped.

[0132] In this embodiment, in step S7, the tag is demodulated as follows:

[0133] The structure of the signal demodulation and reconstruction of the tag signal is as follows: Figure 6 For each sub-channel label i, a bandpass filter BPF is used to filter out all out-of-band spectrum components, and then RF phase compensation is performed (RF phase compensation is to offset the impact of uneven amplitudes of the IQ paths caused by the RF phase). The digital signal X at the input of the Costas loop is i (n) The expression is as follows:

[0134]

[0135] The signal input to the Costas loop is the signal after RF phase compensation. i Indicates that the signal passes through a digital FIR bandpass filter. Considering that the output of the low-pass filter is a baseband Miller coded waveform, a strict linear phase characteristic is required within the passband, so a FIR low-pass filter is used, and its passband bandwidth is the first zero point bandwidth of the tag.

[0136] Costas loop is widely used to track carrier suppression signals with low signal-to-noise ratio. The Costas loop includes: a phase detector, a loop filter and a digitally controlled oscillator; the Costas loop completes the demodulation of BPSK signals, carrier phase locking and carrier frequency multiplication through two loops. Its principle is as follows Figure 6 As shown. The function of the Costas loop is to first demodulate the label subcarrier modulation sequence and remove the subchannel carrier f i , use a low-pass filter LPF to filter the signal to remove 2f i The spectrum is then decoded using the zero-threshold decoding rule for bit decision making.

[0137] Among them, in order to further reduce the overhead, the phase detector does not use the traditional multiplier, but adopts the phase detection algorithm with sign judgment, that is, the multiplication phase detector is connected in series with the FIR low-pass filter structure, and the polarity of the baseband signal output by the in-phase branch is used to control whether the low-pass filter output of the orthogonal branch is flipped; the result of the I-way phase detector is multiplied by the sign of the result of the Q-way phase detector as the phase detection result of the entire loop input to the loop filter, and the loop filter uses a second-order active proportional loop filter; the frequency control word of the numerical control oscillator NCO is added to the loop filter output on the basis of the initial control word; the numerical control oscillator uses the method of direct digital synthesis (Direct Digital Synthesis, DDS) signal generator when implementing it, and generates a primary carrier sine and cosine carrier signal through the frequency control word, and generates a tertiary carrier cosine signal through the triple frequency control word; the primary carrier sine and cosine signals are used for the input of the phase detector multiplication terminal, and the tertiary carrier cosine signal is used as the module output for harmonic interference suppression.

[0138] Then the reconstructed reader receives input X i The tag signal at (n) is used in the loop to generate the sine wave sin[2πf i (n / f s -t i )], the sine wave has the same frequency and phase as the subcarrier of the Miller subcarrier modulation sequence, then the reconstructed Miller code M can be obtained by taking the polarity of the low-pass filter of the Costas loop i (n / f s -t i ), sin[2πf i (n / f s -t i )] After taking the polarity, the reconstructed subcarrier signal Square can be obtained i [2πf i (n / f s -t i )].

[0139] Since the reconstructed signal is in the form of a square wave, there will be a large deviation from the original signal at the edge of the signal frequency hopping. After the interference suppression is implemented, the remaining signal has a spike. Therefore, in order to suppress interference, it is not appropriate to reconstruct the original signal into a square wave form. This embodiment proposes a new method for signal reconstruction using a series of odd harmonics, namely, a square wave signal Square(2πf s The Fourier expansion expression of t) is as follows:

[0140]

[0141] From formula (8), we can see that the square wave signal can be expressed as the sum of an infinite number of odd-order sinusoidal signals, the amplitude of each sinusoidal signal is 1, 1 / 3, 1 / 5, 1 / 7, etc., and the frequency of each sinusoidal signal is an odd multiple of the baseband frequency. Since the sinusoidal modulated signal can be expressed as a spectrum shift of the baseband signal, the Miller subcarrier modulation sequence is equivalent to the sum of the baseband Miller code modulated to an infinite number of odd-order sinusoidal signals. Therefore, the square wave can be fitted by a finite number of odd-order sinusoidal harmonics to obtain the reconstructed Miller subcarrier, which is expressed as follows:

[0142]

[0143] in, and Represents the delay and subcarrier frequency of the subcarrier obtained by Costas loop synchronization. The sinusoidal signal output by the Costas loop is tripled, quintupled, and 7-folded. The frequency multiplication operation is to multiply the frequency control word of the input numerically controlled oscillator by 3, 5, and 7 to achieve triple, quintupled, and 7-fold of the sinusoidal signal output by the numerically controlled oscillator; the sinusoidal signal is superimposed with each odd harmonic component to fit the Miller subcarrier signal; the Miller subcarrier signal is multiplied by the Miller code to obtain the reconstructed label signal; the reconstructed signal expression of label i is as follows:

[0144]

[0145] In summary, the method of the present invention can realize the parallel reception of frequency-division multiplexed RFID tags, so that the reader can simultaneously inventory multiple tags in a single time slot, thereby effectively improving the recognition efficiency and throughput of the UHF RFID system, and improving the ability of the UHF RFID system to adapt to application scenarios that require rapid access to massive tags in large-scale Internet of Things.

[0146] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture, the specific steps are as follows: S1. Build a frequency division multiplexing UHF RFID system, including: 4 RFID tags, 1 reader / writer; The reader sends a 920-925MHz tag excitation signal through the transmitting antenna to activate the four RFID tags. After receiving the inventory command from the reader, the four RFID tags simultaneously return a 920-925MHz signal to the reader receiving antenna. The reader / writer comprises: a baseband signal processing module and a radio frequency front end; the baseband signal processing module comprises: a downsampling module, a bandpass filter BLF, an FFT module, a power estimation module, an amplitude estimation module, a SIR calculation and sorting module, a radio frequency phase compensation module, a Costas loop, a decoding module, and a signal reconstruction module; the radio frequency front end comprises: an orthogonal modulation modulator and an ADC; S2, the four RFID tags respectively use M values ​​of 4, 8, 12, and 16 to perform subcarrier modulation on their respective Miller codes to form Miller subcarrier modulation sequences; After the reader sends the inventory command, the four RFID tags communicate and send the Miller subcarrier modulation sequence to the reader; the bandwidth of each frequency channel is the first zero point bandwidth of the RFID tag signal, that is, 4 times the data rate of the tag baseband signal; S3, after the reader receives the signals returned by the four RFID tags at the same time, it first performs down-conversion and ADC sampling through the RF front end to convert the analog signal into a digital signal; S4, performing power estimation on the total signal received by the reader through FFT; First, estimate the signal power in each frequency band The power spectral density of the Miller subcarrier modulation sequence with M values ​​of 4, 8, 12, and 16 derived by the Wiener-Schinchin theorem is used. The power spectral density is numerically integrated to obtain the power of each Miller subcarrier modulation sequence in different frequency bands, and then the interference degree H caused by the tags in other channels in different frequency channels is obtained. Finally, the estimated value of the actual first zero point bandwidth power of each tag is obtained. Complete power estimation; Where H is a 4×4 matrix, and the element h in the i-th row and j-th column is ij represents the ratio of the power of the RFID tag located in the jth frequency channel in the ith frequency channel to the first zero-point bandwidth power of the tag in the jth frequency channel, and S5, based on step S4, according to Parseval's theorem, performing amplitude estimation and RF phase estimation; S6. Based on step S4, after power estimation is completed, the SIR of each tag is calculated and the SIRs are sorted; S7. Based on step S6, first perform SIC demodulation on the tag with the largest SIR; First, the signal passes through the FIR bandpass filter to filter out the out-of-band interference, and then passes through the RF phase compensation module to offset the influence of the RF phase. The signal is then input into the Costas loop to obtain the subcarrier signal and Miller code of the tag. The signal is then input into the decoding module to judge the Miller code and obtain the original data of the tag. The signal is then input into the signal reconstruction module to obtain the reconstructed tag signal, that is, the Miller subcarrier modulation sequence of the tag is reconstructed. Among them, Costas generates a sinusoidal signal with the same frequency and phase as the original Miller subcarrier in an internal digitally controlled oscillator, and outputs the Miller coded waveform through an internal low-pass filter. The Miller coded waveform is polarized to obtain the Miller code; S8, repeating the demodulation process based on step S7 until the last tag signal is demodulated, completing the RFID tag parallel reception; Based on the reconstructed tag signal obtained by SIC demodulation of the tag with the largest SIR in step S7, the reconstructed signal is divided into I and Q paths using the estimated RF phase, the reconstructed tag signal is subtracted from the I and Q paths, and then the tag signal with the second largest SIR is demodulated. The demodulation process is repeated until the last tag signal is demodulated, completing the final reception process, that is, completing the parallel reception of RFID tags.

2. The method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture according to claim 1, characterized in that: The M value used by the four RFID tags is not limited to 4, 8, 12, and 16, and can be expanded according to actual conditions.

3. The method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture according to claim 1, characterized in that: The step S3 is specifically as follows: The down-converted tag signal X received by the reader receiving antenna RF (t) is expressed as follows: Where t represents time, T i (t) = M i (t)Square i (2πf i t) represents the Miller subcarrier modulation sequence of tag i, where tag i=1, 2, 3, 4 represent tags with M values ​​of 4, 8, 12, and 16, respectively. i (t) represents the baseband Miller code, Square i (2πf i t) represents a square wave signal, a i , t i , and f i They represent the amplitude, RF phase, transmission delay, and subcarrier frequency of tag i respectively; N(t) represents additive white Gaussian noise; Then X RF (t) After ADC sampling, the received complex signal X(n) is expressed as follows: Where n represents a discrete time series, f s represents the ADC sampling frequency; and the reader receiving end is orthogonal demodulation, then the I and Q components of X(n) are rewritten as follows: Among them, the two real factors and Includes the effect of RF phase.

4. The method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture according to claim 1, characterized in that: The step S4 is specifically as follows: The power of the I and Q digital signals after ADC sampling in step S3 is estimated by FFT. First, the power spectral density PSD of the Miller subcarrier modulation sequence with M values ​​of 4, 8, 12, and 16 derived by the Wiener-Hinchin theorem is used. The estimated calculation expression of PSD is as follows: Among them, D I (n), D Q (n) represents the estimated power spectrum density of I and Q paths respectively, represents Fourier transform, N represents the length of Fourier transform; In the FFT module, the reader calculates the power estimation value in the four frequency channels through FFT. in,[·] T represents a transposition operation, the power estimation value includes the first zero-point bandwidth signal of the label corresponding to each channel and the side lobes or harmonic interference of the labels of other three channels in the frequency channel; For channel I, the power estimate entering the i-th channel is obtained by numerically integrating the power spectral density of the FFT estimate in the i-th channel. set up represents the estimated value of the actual power of the i-th tag in the first zero-point bandwidth; then and There are the following relationships: in, It represents the estimated value of the actual power of the tag in the four frequency channels of channel I in the first zero point bandwidth, H represents the linear expression of label interference in different channels, and its element h in the i-th row and j-th column is ij represents the ratio of the power of the tag located in the jth frequency channel in the ith frequency channel to the first zero-point bandwidth power of the tag in the jth frequency channel; and It represents the power estimation value in the four frequency channels of channel I; Similarly, the power estimate of the i-th channel in the Q path is Estimate of the actual power of the i-th tag in the first zero-point bandwidth In summary, the first zero point bandwidth power estimation of each tag distributed in the I and Q paths is calculated Get the power estimate of each frequency channel in I and Q paths 5. The method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture according to claim 1, characterized in that: The step S5 is specifically as follows: The power of the first zero point bandwidth of the label accounts for 0.77 of the power of the entire Miller subcarrier modulation signal. According to Parseval's theorem, the power in the time domain is equal to the power in the frequency domain. The estimated amplitude value of the I path is Similarly, the amplitude value of the Q-path estimate can be obtained Then use the estimated amplitudes of the I and Q paths to estimate the RF phase The calculation method is 6. The method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture according to claim 1, characterized in that: The step S6 is specifically as follows: After the power estimation is completed, the SIR of each subchannel label i is estimated by the SIR calculation and sorting module. The calculation expression of the SIR of label i is as follows: After the reader calculates the SIR of each tag, it determines whether the power value in each channel is lower than the preset threshold based on the power estimation value of each channel obtained in step S5. If it is lower, it is considered that there is no signal in the channel. In the subsequent SIC demodulation process, the channel is not considered. Then the SIRs of the remaining channels are sorted and demodulated in order from large to small SIR. During demodulation, the reader first selects the tag with the largest SIR for SIC demodulation.

7. The method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture according to claim 1, characterized in that: The reader performs SIC demodulation on the current channel and reconstructs the tag signal of the channel. If the reader detects that a bit error occurs in the tag signal, when entering the SIC demodulation of the next level channel, it is divided into two cases, namely, subtracting the reconstructed signal from the remaining total signal and not subtracting the reconstructed signal, and comparing the two cases to see whether a bit error occurs in the SIC demodulation of the next channel; if a bit error does not occur in one of the cases, the current tag signal continues to be reconstructed for subsequent SIC demodulation; if a bit error occurs in both cases, the SIC demodulation is stopped.

8. The method for parallel reception of frequency division multiplexing RFID tag signals based on SIC architecture according to claim 1, characterized in that: In step S7, the tag is demodulated as follows: For each sub-channel label i, a bandpass filter BPF is used to filter out all out-of-band spectrum components, and then an RF phase compensation operation is performed. Then the digital signal X at the input of the Costas loop is i (n) The expression is as follows: The signal input to the Costas loop is the signal after RF phase compensation. i Indicates that the signal passes through a digital FIR bandpass filter; The Costas loop includes: a phase detector, a loop filter and a digitally controlled oscillator; the Costas loop completes the demodulation, carrier phase locking and carrier frequency multiplication of the BPSK signal through two loops; and the function of the Costas loop is to first demodulate the label subcarrier modulation sequence, first remove the subchannel carrier f i , use a low-pass filter LPF to filter the signal to remove 2f i Spectrum, then bit decision is made according to zero threshold decoding rule; Among them, the phase detector adopts a multiplication phase detector in series with an FIR low-pass filter structure; the result of the I-way phase detector is multiplied by the sign of the result of the Q-way phase detector as the phase detection result of the entire loop and input into the loop filter; the frequency control word of the numerical control oscillator NCO is added to the loop filter output on the basis of the initial control word; the numerical control oscillator generates a primary carrier sine and cosine carrier signal through the frequency control word, and generates a tertiary carrier cosine signal through the triple frequency control word; the primary carrier sine and cosine signals are used for phase detector multiplication input, and the tertiary carrier cosine signal is used as a module output for harmonic interference suppression; Then the reconstructed reader receives input X i The tag signal at (n) is used in the loop to generate the sine wave sin[2πf i (n / f s -t i )], the sine wave has the same frequency and phase as the subcarrier of the Miller subcarrier modulation sequence, then the reconstructed Miller code M can be obtained by taking the polarity of the low-pass filter of the Costas loop i (n / f s -t i ), sin[2πf i (n / f s -t i )] After taking the polarity, the reconstructed subcarrier signal Square can be obtained i [2πf i (n / f s -t i )]; The reconstructed signal is in the form of a square wave, and then a method of using a series of odd harmonics to reconstruct the signal is proposed, that is, the square wave signal Square(2πf s The Fourier expansion expression of t) is as follows: Then, a finite number of odd-order sine harmonics are used to fit the square wave to obtain the reconstructed Miller subcarrier, which is expressed as follows: in, and Represents the time delay and subcarrier frequency of the subcarrier obtained by Costas loop synchronization; Finally, the sinusoidal signal output by the Costas loop is tripled, quintupled, and 7-folded, and the sinusoidal signal is superimposed with each odd harmonic component to fit the Miller subcarrier signal; the Miller subcarrier signal is multiplied by the Miller code to obtain the reconstructed label signal; the reconstructed signal expression of label i is as follows:

Images