Demodulation method of linear spread spectrum signal with sampling frequency deviation based on frequency domain difference

Abstract

The present invention discloses a method for demodulating linear spread spectrum signals with sampling frequency deviation based on frequency domain difference, which is mainly used to solve the problem of reduced demodulation performance of linear spread spectrum signals with sampling frequency deviation in scenarios where the source of the sampling clock signal and the sampling frequency are unknown. The implementation steps of the present invention are that the receiving end performs differential decoding and de-reduction set processing on the maximum position of the payload signal after incoherent demodulation; and determines the size relationship between the actual sampling frequency and the signal bandwidth of the linear spread spectrum signal according to the average value of the non-zero maximum position of the leading signal. The present invention improves the demodulation performance of the receiving end and has the advantage of many application scenarios in actual engineering.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and further relates to a method for demodulating a linear spread spectrum signal with sampling frequency deviation based on frequency domain difference in the field of wireless communication technology of the Internet of Things. The present invention can be used for a demodulation method of a linear spread spectrum signal at a receiving end in the Internet of Things when the sampling rate of the linear spread spectrum signal at the receiving end is inconsistent with that at the transmitting end. Background Art

[0002] One of the characteristics of the Internet of Things is the massive connections and numerous node devices, which requires low network deployment costs. In addition, the deployment scenarios of the Internet of Things are rich, and the impact of the environment on the Internet of Things devices is highly uncertain. For example, the frequency accuracy of cheap crystal oscillators is not high, and temperature changes can cause the crystal oscillator frequency to drift. The difference in crystal oscillator frequency between the transmitting and receiving devices causes the sampling rate of the linear spread spectrum signal at the receiving end to be inconsistent with that at the transmitting end, and the demodulation performance of the receiver drops sharply, that is, the bit error rate performance drops, resulting in reduced reliability of communication between Internet of Things devices.

[0003] Carolynn B et al. proposed a demodulation method for linear spread spectrum signals with sampling frequency deviation in their paper “Low Complexity LoRa Frame Synchronization for Ultra-Low Power Software-Defined Radios[J]” (IEEE Transaction on Communication. IEEE, 2020). This method calculates the sampling frequency deviation from the carrier frequency deviation based on the relationship between the carrier frequency deviation and the sampling frequency deviation when the RF carrier and the sampling clock come from the same clock, and then compensates the linear spread spectrum signal for demodulation. The disadvantage of this method is that in order to demodulate a linear spread spectrum signal with a sampling frequency deviation, the size of the carrier frequency deviation must be accurately known. However, the sampling clock and the RF carrier of this method come from different clocks, which will result in a decrease in the demodulation performance of the linear spread spectrum signal with a sampling frequency deviation, resulting in low system reliability.

[0004] Reza G et al. proposed a method for demodulating linear spread spectrum signals with sampling frequency deviation in their paper "LoRa Digital Receiver Analysis And Implementation [C]" (IEEE International Conference on Acoustics, Speech and Signal Processing, IEEE, 2019). This method uses an oversampling method to improve the resolution of the linear spread spectrum signal based on the fact that the error of the demodulation peak of the linear spread spectrum signal with sampling frequency deviation accumulates and increases with the increase of the number of signal symbols in a frame. By judging whether the sampling point of a symbol drifts into the adjacent symbol, the symbol boundary is continuously adjusted by discarding the sampling point, thereby preventing the error accumulation effect caused by the sampling frequency deviation from affecting the signal demodulation. However, the method still has the disadvantage that the premise of using this method is that the receiver needs to accurately know the actual sampling frequency of the linear spread spectrum signal, which is limited in practical engineering applications. Summary of the invention

[0005] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art and to propose a demodulation method for linear spread spectrum signals with sampling frequency deviation based on frequency domain differentiation, so as to solve the problem of decreased demodulation performance of linear spread spectrum signals with sampling frequency deviation in scenarios where the source of the sampling clock signal and the sampling frequency are unknown.

[0006] The idea of ​​realizing the purpose of the present invention is that the present invention adopts the frequency domain difference method, and performs differential coding and reduced set coding on the data to be sent at the transmitting end of the Internet of Things. The mapped decimal data after coding are all even numbers, and then modulated using a linear spread spectrum signal. A frame of linear spread spectrum signal is directly demodulated at the receiving end, and the offset of the maximum position of the signal demodulation increases continuously with the increase of the symbol position. By using the differential encoding and decoding characteristics, the offset between the maximum position of the signal demodulation and the real position can be reduced. At this time, the average value of the non-zero maximum position of the leading signal and the parity of the maximum position of the demodulated load signal are fully utilized for judgment, which can further reduce the offset of the real position of the maximum position of the demodulated load signal. Finally, the set is reduced to solve the residual position offset, avoiding the problem that the receiving end needs to know the source of the sampling clock signal and the actual sampling frequency of the linear spread spectrum signal, and finally recovers the sent modulation information from the linear spread spectrum signal with sampling frequency deviation.

[0007] The specific steps of the present invention are as follows:

[0008] Step 1: Reduce the set and perform differential encoding on the random bits to be sent:

[0009] Step 1.1, the sender generates a random bit sequence consisting of 0s and 1s with a length of A, and divides the random bit sequence into equal parts with B bits as intervals, where the value of B is an integer randomly selected in the range of [4,10];

[0010] Step 1.2, add a bit sequence of length D to the end of each equally divided random bit sequence, so that the length of the added bit sequence is 12, and the added bit sequence has all bit values ​​0 except the first bit value 1, thus obtaining a bit sequence of length 12. The bit sequence of length 12 is mapped to decimal data, and the value of D is an integer selected in the range of [2,8];

[0011] Step 1.3, the first decimal data remains unchanged, and each of the remaining decimal data is added to the previous differentially encoded decimal data, and then modulo operation is performed with 4096, and the first decimal data and the results of all modulo operations are combined to form the differentially encoded decimal data;

[0012] Step 2: Generate and send a linear spread spectrum signal:

[0013] Step 2.1, using the linear spread spectrum signal sequence generation formula, generate 10 linear spread spectrum signal sequences of the transmitting end as the preamble signal sequence, and the length of each linear spread spectrum signal sequence is 4096;

[0014] Step 2.2, using the same method as step 2.1, generating a linear spread spectrum signal sequence carrying specific modulation information at the transmitting end as a payload signal sequence;

[0015] Step 2.3, adding the payload signal sequence to the end of the leading signal sequence to form a frame signal sequence, and the transmitting end sends the frame signal sequence;

[0016] Step 3: The receiving end demodulates the received signal sequence:

[0017] Step 3.1, the receiving end divides the received signal sequence into equal intervals with 4096 sampling points as the interval, and extracts the leading signal sequence and the payload signal sequence;

[0018] Step 3.2, the receiving end performs non-coherent demodulation on the pilot signal sequence and the payload signal sequence respectively;

[0019] Step 4: perform differential decoding on the maximum position of each non-coherently demodulated payload signal sequence:

[0020] Find the maximum value of the signal modulus value from the payload signal sequence after non-coherent demodulation, take the position of this maximum value in the sequence as the position of the maximum value of the payload signal sequence, subtract the maximum value position of the last leading signal sequence from the maximum value position of the first payload signal sequence, and then perform a modulo operation with 4096, subtract the maximum value position of the previous payload signal sequence from the maximum value position of the remaining payload signal sequences, and then perform a modulo operation with 4096, and use the results of all modulo operations as the maximum value position of the payload signal sequence after differential decoding;

[0021] Step 5, finding the position of the maximum value of each signal sequence from the pilot signal sequence after non-coherent demodulation, and averaging the non-zero maximum value positions of the pilot signal sequence;

[0022] Step 6: Process the maximum position of the payload signal sequence after differential decoding:

[0023] If the average value of the non-zero maximum position of the leading signal sequence is less than or equal to 2047, and the maximum position of the payload signal sequence after differential decoding is an odd number, the maximum position is reduced by 1; if the average value of the non-zero maximum position of the leading signal sequence is greater than 2047, and the maximum position of the payload signal sequence after differential decoding is an even number, the maximum position is increased by 1; in other cases, the maximum position is kept unchanged, and the processed maximum position and the unchanged maximum position are combined to form a new payload signal maximum position set;

[0024] Step 7: After performing a set reduction operation on each maximum position in the new payload signal maximum position set, modulation information is extracted therefrom.

[0025] Compared with the prior art, the present invention has the following advantages:

[0026] First, because the present invention directly demodulates the payload signal with sampling frequency deviation, and performs differential decoding and reduction set in turn on the maximum value position of the demodulated payload signal, it overcomes the shortcoming of the prior art that the demodulation performance of the linear spread spectrum signal with sampling frequency deviation decreases when the sampling clock and the RF carrier come from different clocks. This enables the present invention to reduce the bit error rate and improve the demodulation performance of the IoT receiving end.

[0027] Secondly, the present invention directly demodulates the leading signal with sampling frequency deviation and determines the relationship between the actual sampling frequency of the linear spread spectrum signal and the signal bandwidth according to the average value of the non-zero maximum position of the leading signal. This overcomes the deficiency in the prior art that the receiving end needs to accurately know the actual sampling frequency of the linear spread spectrum signal. The present invention can be applied not only to the scenario where the actual sampling frequency of the linear spread spectrum signal is known at the receiving end, but also to the scenario where the actual sampling frequency of the linear spread spectrum signal is unknown at the receiving end, to demodulate the linear spread spectrum signal with sampling frequency deviation, thereby enriching the application scenarios in actual engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a flow chart of the present invention;

[0029] Figure 2 It is a flow chart of non-coherent demodulation of linear spread spectrum signal in the present invention;

[0030] Figure 3 is a flow chart of obtaining a new load peak position according to the present invention;

[0031] Figure 4 It is a comparison diagram of the simulation experiment of the present invention. DETAILED DESCRIPTION

[0032] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0033] Reference Figure 1 , further describing the implementation steps of the embodiment of the present invention.

[0034] Step 1: Reduce the set and perform differential encoding on the random bits to be sent in turn to obtain payload modulated data.

[0035] Step 1.1, in the embodiment of the present invention, the sender generates a 24-bit random bit sequence {1,0,0,1,0,1,0,0,1,1,0,0,1,1,1,0,1,1,1,0,1,0,1}, divides the random bit sequence equally at intervals of 8 bits, and obtains three 8-bit equally divided random bit sequences, namely {1,0,0,1,0,1,0,0}, {1,1,0,0,1,1,1,0} and {1,1,1,1,0,1,0,1}.

[0036] Step 1.2, add {1,0,0,0} to the end of each equally divided random bit sequence so that the length of the added bit sequence is 12. Except for the first bit value of the added bit sequence which is 1, all the other bit values ​​are 0. The bit sequences after adding are {1,0,0,1,0,1,0,0,1,0,0,0}, {1,1,0,0,1,1,1,0,1,0,0,0} and {1,1,1,1,0,1,0,1,1,0,0,0} respectively. Since the added bit sequences {1,0,0,1,0,1,0,0,1,0,0,0}, {1,1,0,0,1,1,1,0,1,0,0,0} and {1,1,1,1,0,1,0,1,1,0,0,0} contain random bits {1,0,0,1,0,1,0,0}, {1,1,0,0,1,1,1,0} and {1,1,1,1,0,1,0,1} respectively, and the total number of random bits is 8, which is less than 12, the added bit sequence is the bit sequence after the reduced set. The bit sequences after the reduced set are mapped to decimal data, which are 2376, 3304 and 3928 respectively.

[0037] Step 1.3, add the first decimal data 2376 to the decimal data 0, and perform a modulo operation with 4096 to obtain a result of 2376, which is used as the first decimal data 2376 after differential coding. Add the second decimal data 3304 to the first decimal data 2376 after differential coding, and perform a modulo operation with 4096 to obtain a result of 1584, which is used as the second decimal data 1584 after differential coding. Add the third decimal data 3928 to the second decimal data 1584 after differential coding, and perform a modulo operation with 4096 to obtain a result of 1416. After the above operations, three differentially coded payload modulation data are obtained, which are 2376, 1584 and 1416 respectively.

[0038] Step 2: Generate and send a linear spread spectrum signal.

[0039] Step 2.1, use Formula, the transmitter generates 10 linear spread spectrum signal sequences as the leading signal sequences, the modulation information carried by the 10 linear spread spectrum signal sequences is 0, and the length of each linear spread spectrum signal sequence is 4096, where c(·) represents the generated linear spread spectrum signal sequence, k represents the index value of the linear spread spectrum signal sequence data, which is an integer in the range of [0,4095], and T represents the time interval between adjacent data in the linear spread spectrum signal sequence, which is p represents the modulation information carried by the linear spread spectrum signal, e (·) It represents the exponential operation with the natural constant e as the base, j represents the imaginary unit symbol, and π represents pi.

[0040] In step 2.2, the same method as in step 2.1 is used to generate three linear spread spectrum signal sequences at the transmitter as payload signal sequences. The length of each linear spread spectrum sequence is 4096, and the modulation information carried by the three linear spread spectrum signal sequences is {2376, 1584, 1416} respectively.

[0041] Step 2.3, the payload signal sequence is added to the end of the leading signal sequence to form a frame signal sequence, and the transmitting end sends the frame signal sequence.

[0042] Step 3: The receiving end extracts the preamble signal sequence and the payload signal sequence, and demodulates the preamble signal sequence and the payload signal sequence respectively.

[0043] Step 3.1, the receiving end resamples the received signal sequence of one frame at 124.995kHz to obtain a resampled signal sequence of one frame, and divides the resampled signal sequence into equal intervals of 4096 sampling points to obtain the divided signal sequence {C1, C2, C3, C4, C5, C6, C7, C8, C9, C10, D1, D2, D3}, where C1, C2, C3, C4, C5, C6, C7, C8, C9, C10, D1, D2, D3 all contain 4096 sampling points. C1, C2, C3, C4, C5, C6, C7, C8, C9, C10 are preamble signal sequences, and D1, D2, D3 are payload signal sequences.

[0044] Step 3.2, the receiving end performs non-coherent demodulation on the C1, C2, C3, C4, C5, C6, C7, C8, C9, C10 signal sequences respectively to obtain demodulated preamble signal sequences E1, E2, E3, E4, E5, E6, E7, E8, E9, E10, and performs non-coherent demodulation on the D1, D2, D3 signals respectively to obtain demodulated payload signal sequences F1, F2, F3. The specific process of the non-coherent demodulation is as shown in the attached figure. Figure 2 shown.

[0045] Step 4: Find the position of the maximum value of each signal sequence from the payload signal sequence after incoherent demodulation, and perform differential decoding on the position of the maximum value of the payload signal sequence after demodulation.

[0046] The maximum value 2378 of the demodulated payload signal sequence F1 is subtracted from the maximum value 1 of the demodulated preamble signal sequence E10, and the result is modulo 4096, which is 2377. The maximum value 1586 of the demodulated payload signal sequence F2 is subtracted from the maximum value 2378 of the demodulated payload signal sequence F1, and the result is modulo 4096, which is 3304. The maximum value 1586 of the demodulated payload signal sequence F2 is subtracted from the maximum value 1418 of the demodulated payload signal sequence F3, and the result is modulo 4096, which is 3928. The payload modulation data 2377, 3304 and 3928 after differential decoding are obtained.

[0047] Step 5, finding the position of the maximum value of each signal sequence from the pilot signal sequence after non-coherent demodulation, and performing average processing on the non-zero maximum value positions of the demodulated pilot signal sequence.

[0048] Step 5.1, find the position of the maximum value from the leading signal sequences E1, E2, E3, E4, E5, E6, E7, E8, E9, and E10 after incoherent demodulation. The positions of the maximum value are 0, 0, 0, 0, 1, 1, 1, 1, 1 respectively.

[0049] Step 5.2, the non-zero maximum position of the pilot signal sequence after non-coherent demodulation is 1,1,1,1, and the average value is 1.

[0050] Step 6: Process the maximum position of the payload signal sequence after differential decoding according to the size of the average position of the maximum value after demodulation of the pilot signal.

[0051] The average value of the non-zero maximum position of the pilot signal sequence after demodulation is 1, which is less than 2047. Therefore, the first payload modulation data 2377 after differential decoding is an odd number, which is subtracted by 1 to obtain 2376. The second payload modulation data 3304 after differential decoding is an even number, which is kept unchanged. The third payload modulation data 3928 after differential decoding is an even number, which is kept unchanged. Three new payload signal maximum positions are obtained, which are 2376, 3304 and 3928. The process of obtaining three new payload peak positions is as follows: Figure 3 shown.

[0052] Step 7: Perform a reduction set operation on the new maximum value position of the load signal and extract modulation information therefrom.

[0053] The above-mentioned set reduction operation refers to mapping each maximum value position in the new payload signal maximum value position set to binary data with a bit width of 12 bits, and discarding D bits of the 12-bit data. In this embodiment, the three payload signal maximum value positions 2376, 3304 and 3928 are mapped to binary data, namely {1,0,0,1,0,1,0,0,1,0,0,0}, {1,1,0,0,1,1,1,0,1,0,0,0} and {1,1,1,1,0,1,0,1,1,0,0,0}, respectively, and {1,0,0,1,0,1,0,1,0,0,0} is discarded. The lower four bits of the data are {1,0,0,1,0,1,0,0}, {1,1,0,0,1,1,1,0,1,0,0,0} and {1,1,1,1,0,1,0,1,1,0,0,0}, and we get {1,0,0,1,0,1,0,0}, {1,1,0,0,1,1,1,0} and {1,1,1,1,0,1,0,1}. We concatenate them together and get {1,0,0,1,0,1,0,0,1,1,0,0,1,1,0,1,1,1,0,1,0,1,0,1}, which are the random bits sent by the sender.

[0054] The effect of the present invention can be further demonstrated through the following simulation experiments.

[0055] 1. Simulation experimental conditions.

[0056] The platforms for the simulation experiment of the present invention are: Windows 10 operating system and Matlab R2020b.

[0057] The linear spread spectrum signal with sampling frequency deviation used in the simulation experiment of the present invention is composed of a pilot signal and a payload signal. The pilot signal includes 10 linear spread spectrum signals with modulation information of 0, the payload signal includes 40 linear spread spectrum signals carrying specific modulation information, the transmission channel environment is an additive white Gaussian noise channel, the spreading factor is 12, the bandwidth is 125KHz, the sampling frequency deviation value is -40ppm, and the signal-to-noise ratio is from -30 to -10.

[0058] 2. Analysis of simulation content and results.

[0059] The simulation experiment of the present invention adopts the frequency domain difference method of the present invention and an existing technology (demodulation method of linear spread spectrum signal), demodulates the linear spread spectrum signal with sampling frequency deviation under 21 signal-to-noise ratio conditions, obtains the received binary bits, counts the number of received binary bits that are different from the transmitted binary bits, and takes the ratio of each counted number to the number of transmitted binary bits as the bit error rate under 21 signal-to-noise ratios. The relationship between the bit error rate and the signal-to-noise ratio after demodulation of the linear spread spectrum signal with sampling frequency deviation is plotted as shown in the figure. Figure 4 The two curves are shown.

[0060] In the simulation experiment, the existing technology used refers to: a demodulation method for linear spread spectrum signals proposed by Lorenzo Vangeliata in his published paper “Frequency Shift Chirp Modulation: The LoRa Modulation[J]” (IEEE SignalProcessing Letters IEEE, 2017).

[0061] Combine the following Figure 4 The simulation diagram of the present invention is further described.

[0062] Figure 4 The horizontal axis in the figure represents the signal-to-noise ratio in dB, and the vertical axis represents the bit error rate. The curve marked with a five-pointed star represents the relationship curve between the bit error rate and the signal-to-noise ratio obtained by simulating the method of the present invention, and the curve marked with a circle represents the relationship curve between the bit error rate and the signal-to-noise ratio obtained by simulating the prior art.

[0063] from Figure 4 It can be seen that the 21 bit error rates obtained by the method of the present invention are constantly decreasing with the increase of the signal-to-noise ratio. In the range of signal-to-noise ratio [-24, -10], the bit error rate obtained by the method of the present invention is always lower than the bit error rate obtained by the simulation of the prior art, indicating that the demodulation performance of the method of the present invention is better than that of the prior art.

Claims

1. A method for demodulating a linear spread spectrum signal with sampling frequency deviation based on frequency domain difference, characterized in that: The receiving end performs differential decoding and de-reduction set processing on the maximum position of the payload signal after non-coherent demodulation; the size relationship between the actual sampling frequency of the linear spread spectrum signal and the signal bandwidth is determined according to the average value of the non-zero maximum position of the pilot signal; the steps of the demodulation method include the following: Step 1: Reduce the set and perform differential encoding on the random bits to be sent: Step 1.1, the sender generates a random bit sequence consisting of 0s and 1s with a length of A, and divides the random bit sequence into equal parts with B bits as intervals, where the value of B is an integer randomly selected in the range of [4,10]; Step 1.2, add a bit sequence of length D to the end of each equally divided random bit sequence, so that the length of the added bit sequence is 12, and the added bit sequence has all bit values ​​0 except the first bit value 1, thus obtaining a bit sequence of length 12. The bit sequence of length 12 is mapped to decimal data, and the value of D is an integer selected in the range of [2,8]; Step 1.3, the first decimal data remains unchanged, and each of the remaining decimal data is added to the previous differentially encoded decimal data, and then modulo operation is performed with 4096, and the first decimal data and the results of all modulo operations are combined to form the differentially encoded decimal data; Step 2: Generate and send a linear spread spectrum signal: Step 2.1, using the linear spread spectrum signal sequence generation formula, generate 10 linear spread spectrum signal sequences of the transmitting end as the preamble signal sequence, and the length of each linear spread spectrum signal sequence is 4096; Step 2.2, using the same method as step 2.1, generating a linear spread spectrum signal sequence carrying specific modulation information at the transmitting end as a payload signal sequence; Step 2.3, adding the payload signal sequence to the end of the leading signal sequence to form a frame signal sequence, and the transmitting end sends the frame signal sequence; Step 3: The receiving end demodulates the received signal sequence: Step 3.1, the receiving end divides the received signal sequence into equal intervals with 4096 sampling points as the interval, and extracts the leading signal sequence and the payload signal sequence; Step 3.2, the receiving end performs non-coherent demodulation on the pilot signal sequence and the payload signal sequence respectively; Step 4: perform differential decoding on the maximum position of each non-coherently demodulated payload signal sequence: Find the maximum value of the signal modulus value from the payload signal sequence after non-coherent demodulation, take the position of this maximum value in the sequence as the position of the maximum value of the payload signal sequence, subtract the maximum value position of the last leading signal sequence from the maximum value position of the first payload signal sequence, and then perform a modulo operation with 4096, subtract the maximum value position of the previous payload signal sequence from the maximum value position of the remaining payload signal sequences, and then perform a modulo operation with 4096, and use the results of all modulo operations as the maximum value position of the payload signal sequence after differential decoding; Step 5, finding the position of the maximum value of each signal sequence from the pilot signal sequence after non-coherent demodulation, and averaging the non-zero maximum value positions of the pilot signal sequence; Step 6: Process the maximum position of the payload signal sequence after differential decoding: If the average value of the non-zero maximum position of the leading signal sequence is less than or equal to 2047, and the maximum position of the payload signal sequence after differential decoding is an odd number, the maximum position is reduced by 1; if the average value of the non-zero maximum position of the leading signal sequence is greater than 2047, and the maximum position of the payload signal sequence after differential decoding is an even number, the maximum position is increased by 1; in other cases, the maximum position is kept unchanged, and the processed maximum position and the unchanged maximum position are combined to form a new payload signal maximum position set; Step 7, performing a set reduction operation on each maximum position in the new payload signal maximum position set, and then extracting modulation information therefrom; The set reduction operation refers to mapping each maximum value position in the new payload signal maximum value position set to binary data with a bit width of 12 bits, and discarding D bits of the 12-bit data.

2. The method for demodulating a linear spread spectrum signal with sampling frequency deviation based on frequency domain difference according to claim 1, characterized in that: The linear spread spectrum signal sequence generation formula described in step 2.1 is as follows: Where c(·) represents the generated linear spread spectrum signal sequence, k represents the index value of the linear spread spectrum signal sequence data, which is an integer in the range of [0,4095], and T represents the time interval between adjacent data in the linear spread spectrum signal sequence, which is B represents the bandwidth of the signal, p represents the modulation information carried by the linear spread spectrum signal, and e (·) It represents the exponential operation with the natural constant e as the base, j represents the imaginary unit symbol, and π represents pi.

Images