CN116405359A - LoRa concurrent communication demodulation method and system based on frequency domain interference iterative cancellation - Google Patents

Abstract

The invention discloses a LoRa concurrent communication demodulation method and system based on frequency domain interference iterative elimination, comprising signal transmission; taking the chirp signal as a processing unit, and transmitting the signal by a node end in a standard LoRa data packet frame format; receiving signals; the gateway receives the LoRa signal in a conflict state by taking the frequency domain characteristics of the pilot frequency part as a basis, wherein the conflict state is that a plurality of data packets are transmitted in the same time period; preprocessing signals; performing low-pass filtering and signal synchronization on the LoRa signal in the collision state, dividing the LoRa signal by taking the chirp length as a step length to obtain a plurality of chirp symbols, and performing carrier frequency offset correction on the chirp symbols; demodulating conflict signals; and D, carrying out frequency domain interference estimation and elimination on the chirp symbol preprocessed in the step three, and completing symbol demodulation. The demodulation method provided by the invention can complete demodulation of the conflict signal under the condition that the signal strength is far lower than the noise strength under the condition that the stronger robustness exists under the condition of low signal-to-noise ratio.

Description

A LoRa concurrent communication demodulation method and system based on iterative elimination of frequency domain interference

technical field

The invention belongs to the field of communication, and in particular relates to a LoRa concurrent communication demodulation method and system based on frequency domain interference iterative elimination.

Background technique

LoRa is based on linear spread spectrum modulation technology, and each signal it modulates, namely chirp, is a sin wave whose frequency linearly increases (upchirp) or decreases (downchirp). Within the frequency band (-BW/2, BW/2) the frequency of the chirp rises linearly from the initial frequency f ₀ to bw/2, and then returns to the lower boundary of the frequency band -bw/2, thereby sweeping the entire bandwidth. Different chirps correspond to different starting frequencies, and there are 2^SF types in total. As shown in Figure 1(a) and Figure 1(b), the time-frequency diagrams of two up-chirps with different starting frequencies are respectively shown. It can be seen from the figure that the frequency of the signal is gradually increasing within a certain interval. As shown in Figure 1(c), it is a time-frequency diagram of a standard down-chirp. It can be seen from the figure that the frequency of the signal gradually decreases with time. During demodulation, pulse compression is performed on the original chirp to obtain the starting frequency. Since there is a conjugate relationship between upchirp and downchirp, here you can directly use upchirp to multiply the standard downchirp, and perform Fourier transform to obtain its Frequency energy peak value of an interval. The frequency corresponding to the energy peak is its initial frequency for demodulation.

However, with the massive deployment of LoRa nodes, thousands of LoRa nodes are connected to one LoRa gateway. Severe packet collisions caused by the aggregated network structure lead to massive packet loss and reduced throughput. In the wireless transmission technology, the processing method of the conflict problem is divided into conflict avoidance and conflict resolution.

In terms of collision avoidance, LoRa uses multiple PHY technologies to transmit multiple LoRa nodes in parallel, and LoRa nodes can be configured with different radio parameters (eg, channel, spreading factor, etc.) to mitigate collisions, but this requires different operators and service providers The cooperation between suppliers, and the LoRa gateway can only support up to 8 LoRa nodes to transmit at the same time. However, since LoRa signals are mostly transmitted in an environment lower than the noise floor, it is difficult to complete signal-based channel monitoring, resulting in failure to avoid conflicts. At the same time, channel monitoring requires nodes to pay a lot of computing power and energy consumption, thereby shortening the working life of LoRa nodes. Limited by hardware capability and power supply, LoRa nodes usually adopt simple Aloha-based MAC to avoid collisions. According to the specification of the protocol, LoRa nodes can transmit data packets without channel detection. When the LoRa gateway is unable to decode one packet sent from the slave node due to collision, the packet will be retransmitted after a random backoff time, and the backoff retransmission further exacerbates the collision problem in the LoRa network. At the same time, it also brings unnecessary energy waste caused by nodes sending multiple times. Therefore, our more ideal state is that the node can send data packets whenever it wants to send, and the gateway will decode the conflicting signal packets.

There are also some related works in conflict decoding, such as Choir, which classifies conflict frames according to different decimal hardware frequency offsets of each LoRa node. However, in practical applications, due to the influence of interference and noise, it is difficult to accurately extract the carrier frequency offset caused by hardware. At the same time, as the number of nodes increases, the fractional hardware frequency offset will inevitably repeat, resulting in classification failure. Another work, mLoRa, deduces the time offset between two conflicting packets according to the designed preamble detection strategy, and then obtains chirp-level non-collision samples and corresponding frequencies. According to these non-conflict frequencies and samples, the preamble part is used The chirp amplitude and chirp linear spread spectrum modulation method are used to estimate and reconstruct the time-domain samples of the conflicting signal, and the time-domain samples of the estimated and reconstructed conflicting signal are subtracted from the time domain of the conflicting signal, through repeated estimation and Subtractive operation to decode colliding packets in one sample. However, since LoRa signals are mostly transmitted under low signal-to-noise ratio, and this method will greatly reduce the decoding accuracy when the SNR is reduced, which limits the communication performance of LoRa. CoLaRa utilizes packet time offsets to eliminate collisions. Cut the received signal into a series of receive windows, each of length equal to the chirp. Split symbols with misaligned windows. Then, for the signal in each window, collision decoding is performed according to the height of the peak in the frequency domain which is proportional to the length of the segment. However, in the process of reducing the SNR from 0dB to -15dB in this method, the system throughput is reduced to half of the original, and the performance is greatly reduced.

Contents of the invention

The present invention provides a LoRa concurrent communication demodulation method and system based on iterative elimination of frequency domain interference, which can demodulate conflicting LoRa signals in the case of SNR<0; the method combines sliding window detection and utilizes The frequency-domain correlation analysis of the received signal after pulse compression and the prior signal can ensure accurate data packet detection even if the strength of the LoRa signal is much lower than the noise strength.

In order to achieve the above tasks, the present invention adopts the following technical solutions, including:

A LoRa concurrent communication demodulation method based on iterative elimination of frequency domain interference is performed according to the following steps:

Step 1: Signal sending;

Taking the chirp signal as the processing unit, the node sends the signal in the standard LoRa packet frame format;

Step 2: Signal reception;

The gateway side uses the frequency domain characteristics of the pilot part as a basis to receive the LoRa signal in the conflict state. The conflict state means that multiple data packets are transmitted at the same time period;

Step 3: Signal preprocessing;

Low-pass filtering and signal synchronization are performed on the LoRa signal in the conflict state, and multiple chirp symbols are obtained by dividing with the chirp length as the step size, and carrier frequency offset correction is performed on the chirp symbols;

Step 4: demodulation of conflicting signals;

Perform frequency domain interference estimation and elimination on the chirp symbols preprocessed in step 3, and complete symbol demodulation.

Optionally, in step 4, the estimation of frequency domain interference includes obtaining frequency domain peak frequency points and frequency domain peak heights;

Among them, the acquisition of frequency domain peak frequency points includes:

Among them, f' _A1-1 , f' _A1-2 is the frequency domain peak frequency point of the interference generated by chirp symbol A1 to chirp symbol B1, Hz; f _A1-1 , f _A1-2 is the frequency domain of chirp symbol A1 Peak frequency point, Hz; BW is the frequency bandwidth, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s;

The acquisition of peak height in frequency domain includes:

时：when

hour:

时：when

hour:

Among them, _h'A1-1 , _h'A1-2 is the frequency domain peak height of the interference generated by chirp symbol A1 to chirp symbol B1; _hA1-1 , _hA1-2 is the frequency domain peak height of chirp symbol A1, Hz; h is the complete peak energy height of the chirp symbol in frequency domain, Hz.

Optionally, in step 4, after frequency domain interference estimation, the symbol demodulation process includes:

Eliminate the estimated frequency domain interference (f' _A1-1 , h' _A1-1 ), (f' _A1-2 , h' _A1-2 ) in the frequency domain;

In the frequency domain after eliminating the interference position, find the frequency point f where the frequency domain peak is the highest,

时：when

hour:

时：when

hour:

Among them: f is the frequency point of the highest peak in the frequency domain of the chirp symbol B1 after the interference is eliminated, Hz; BW is the frequency bandwidth, Hz; f _0-B is the starting frequency of the chirp symbol B1, Hz; find the frequency of the chirp symbol B1 The starting frequency is to complete the demodulation of chirp symbols.

Optionally, in the third step, the specific process of correcting the carrier frequency offset is shown in the following formula;

Among them: S′ is the chirp symbol after carrier frequency offset correction, S _chirp is the original chirp symbol, f ₀ is the starting frequency of the original chirp symbol, Hz, f _cfo is the carrier frequency offset frequency, Hz; k is the frequency change Rate, Hz/s; t is time, s; j is imaginary part signal.

Optionally, in step 3, the low-pass filtering is performed using an IIR low-pass filter.

Optionally, in step 3, perform pulse compression on the frequency domain energy of the LoRa signal after the low-pass filtering, use the sliding window to find the energy maximum position after pulse compression, which is the starting position of the payload of the chirp signal, and realize the signal Synchronization; use 10 downchirps for sliding windows;

Use the sliding window to perform signal synchronization on the low-pass filtered LoRa signal, and segment the Payload signal in the low-pass filtered LoRa signal according to the time unit.

A LoRa concurrent communication demodulation system based on iterative elimination of frequency domain interference, including:

The signal sending module takes the chirp signal as the processing unit, and the node sends the signal in the standard LoRa packet frame format;

The signal receiving module, the gateway side uses the frequency domain characteristics of the pilot part as a basis to receive the LoRa signal in the conflict state. The conflict state means that multiple data packets are transmitted at the same time period;

The signal preprocessing module performs low-pass filtering and signal synchronization on the LoRa signal in the conflict state, divides the chirp length into multiple chirp symbols, and corrects the carrier frequency offset of the chirp symbols;

The conflicting signal demodulation module performs frequency domain interference estimation and elimination on the chirp symbols preprocessed in step 3, and completes symbol demodulation.

Optionally, in the conflicting signal demodulation module, the frequency domain interference estimation includes obtaining frequency domain peak frequency points and frequency domain peak heights;

Among them, the acquisition of frequency domain peak frequency points includes:

Among them, f' _A1-1 , f' _A1-2 is the frequency domain peak frequency point of the interference generated by chirp symbol A1 to chirp symbol B1, Hz; f _A1-1 , f _A1-2 is the frequency domain of chirp symbol A1 Peak frequency point, Hz; BW is the frequency bandwidth, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s;

The acquisition of peak height in frequency domain includes:

时：when

hour:

时：when

hour:

Among them, _h'A1-1 , _h'A1-2 is the frequency domain peak height of the interference generated by chirp symbol A1 to chirp symbol B1; _hA1-1 , _hA1-2 is the frequency domain peak height of chirp symbol A1, Hz; h is the complete peak energy height of the chirp symbol in frequency domain, Hz.

Optionally, after frequency domain interference estimation, the symbol demodulation process includes:

Eliminate the estimated frequency domain interference (f' _A1-1 , h' _A1-1 ), (f' _A1-2 , h' _A1-2 ) in the frequency domain;

In the frequency domain after eliminating the interference position, find the frequency point f where the frequency domain peak is the highest,

时：when

hour:

时：when

hour:

Among them: f is the frequency point of the highest peak in the frequency domain of the chirp symbol B1 after the interference is eliminated, Hz; BW is the frequency bandwidth, Hz; f _0-B is the starting frequency of the chirp symbol B1, Hz; find the frequency of the chirp symbol B1 The starting frequency is to complete the demodulation of chirp symbols.

Optionally, in the signal preprocessing module, the specific process of correcting the carrier frequency offset is shown in the following formula;

Among them: S′ is the chirp symbol after carrier frequency offset correction, S _chirp is the original chirp symbol, f ₀ is the starting frequency of the original chirp symbol, Hz, f _cfo is the carrier frequency offset frequency, Hz; k is the frequency change Rate, Hz/s; t is time, s; j is imaginary part signal.

Compared with the prior art, the present invention has the following characteristics:

The present invention will propose a method and system for LoRa signal conflict demodulation during concurrent communication based on the original LoRa communication mechanism. By iteratively estimating and eliminating the frequency domain interference when the LoRa signal collides, the LoRa signal demodulation in the conflict state is completed. The conflict demodulation method does not require additional hardware modules. At the same time, the LoRa signal conflict demodulation method can also be used together with channel parameter optimization and other solutions to further improve network performance. The idea of demodulating conflicting signals by iteratively estimating and canceling interference in the frequency domain has never been considered in previous work. Compared with other methods for demodulating conflicting signals based on signal time-domain characteristics or signal energy characteristics, this method has stronger robustness at low SNR, and can be completed when the signal strength is much lower than the noise strength. Demodulation of conflicting signals.

Description of drawings

The accompanying drawings are used to provide a further understanding of the present disclosure, and constitute a part of the description, together with the following specific embodiments, are used to explain the present disclosure, but do not constitute a limitation to the present disclosure. In the attached picture:

Figure 1 is the time-frequency diagram of three different chirps, (a) up-chirp with a starting frequency of -250kHz; (b) up-chirp with a starting frequency of -125kHz; (c) standard down-chirp;

Figure 2 is a time-frequency diagram of LoRa data packets under normal conditions;

Figure 3 is a time-frequency diagram of the LoRa data packet in the conflict state;

Fig. 4 shows the demodulation accuracy of conflicting signals under different signal-to-noise ratios when SF=8.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only a part of the embodiments of the present invention, not all embodiments, and do not limit the present invention in any form. All changes belong to the protection scope of the present invention.

The standard LoRa signal data packet includes the following parts: the pilot used for signal detection, that is, the Preamble, the start frame delimiter (SFD) used for signal synchronization, and the payload data part Payload that records the original transmission information.

The payload data part Payload is composed of various chirp signals, and the Payload signal is divided according to the time unit to obtain multiple chirp signals; in the signal sending stage, the sending signal is not changed, and it is carried out according to the standard frame format specified by LoRaWAN, so that it retains the original The sending settings and communication protocol do not conflict with the original communication protocol, ensuring that the system can be compatible with the original LoRa communication system;

With the massive deployment of LoRa nodes, thousands of LoRa nodes are connected to one LoRa gateway. The serious data packet collision caused by the aggregated network structure leads to a large amount of data packet loss and a decrease in throughput. Based on this problem, a demodulation system for conflicting data packets during concurrent communication is proposed. In order to further complete LoRa concurrent communication under low SNR For the demodulation of transmission conflicting signals, the present invention proposes a method for iteratively estimating and eliminating the frequency domain interference of conflicting signals, and the present invention can demodulate conflicting LoRa signals in the case of SNR<0; the method combines Sliding window detection is used, and the frequency domain correlation analysis between the received signal after pulse compression and the prior signal is used to ensure accurate packet detection even if the strength of the LoRa signal is much lower than the noise strength. include:

Step 1: The node (signal sending end) sends a data packet in the standard LoRaWAN frame format to the gateway (signal receiving end). For example, according to the communication environment, the sending end selects the appropriate spreading factor SF, frequency bandwidth BW, and the number of pilot chirps according to the LoRaWAN protocol, and does not change or limit the signal transmission of the node end;

Step 2: Signal reception.

Based on the frequency domain characteristics of the pilot part, the gateway side receives the LoRa signal in the conflict state as shown in Figure 3. The conflict state means that multiple data packets are transmitted at the same time period, as shown in Figure 2. The time-frequency diagram of the LoRa data packet in the transmission state, as shown in Figure 3, is the time-frequency diagram of the LoRa data packet in the conflict state; the frequency domain characteristics after the pulse compression of the pilot signal are used to correlate with the frequency domain characteristics of the prior signal Analysis, if the degree of correlation is higher than the degree of correlation analysis between the noise and the frequency domain characteristics of the prior signal, it is determined that it is a LoRa signal, and the data packet is received.

Step 3: Perform data preprocessing on the received signal.

Perform high-frequency noise filtering and signal segmentation on the LoRa signal, that is, divide and obtain each chirp signal and carrier frequency offset correction according to the length of time;

In the signal preprocessing stage, the signal is filtered using an IIR low-pass filter to remove high-frequency noise outside the signal bandwidth. Signal synchronization and segmentation using sliding windows. Carrier frequency offset is due to mismatch between local crystal oscillators between transceivers and Doppler shift during signal transmission. And this frequency offset will cause the frequency point value offset of the LoRa signal after pulse compression, thus affecting the decoding result. In the signal preprocessing stage, it is necessary to correct the carrier frequency offset to obtain the accurate original frequency domain characteristics of the signal.

In step three, data preprocessing is required. The first thing that needs to be done in this step is signal synchronization and accurate interception. Here, a sliding window is used to multiply the signal and Down chirp in the window, and the energy in the frequency domain is pulse-compressed, and the sliding window is used to find the pulse-compressed The energy maximum position is the starting position of the chirp signal payload, which realizes precise signal synchronization. Afterwards, each chirp signal is divided according to the time length of the chirp signal, and finally the carrier frequency offset frequency f _cfo is calculated according to the frequency offset value of the pilot to complete the frequency offset correction.

In step 3, the pilot part of the LoRa data packet is used to estimate the carrier frequency offset and perform frequency correction, so as to obtain an accurate frequency point value. .

Among them: S′ is the chirp symbol after carrier frequency offset correction, S _chirp is the original chirp symbol, f ₀ is the starting frequency of the original chirp symbol, Hz, f _cfo is the carrier frequency offset frequency, Hz; k is the frequency change Rate, Hz/s; t is time, s; j is imaginary part signal.

Step 4: Perform frequency domain interference estimation and elimination, and complete demodulation.

For the chirp signal preprocessed in step 3, calculate the time window offset t _{1_2} generated by the signal arrival time difference at the chirp symbol level according to the starting position between the signal synchronizations.

t _{1_2} = p ₂ -p ₁ -N×T;

Among them, the value range of the time window offset t _{1_2} is between 0 and T, s; p ₂ is the arrival time of the second packet, s; p ₁ is the arrival time of the first packet, s; N is positive Integer, T is the duration of the chirp symbol, s.

Then demodulate the clean segment of the first chirp symbol A1 where the previous data packet A collides with the subsequent data packet B. The process is as follows: the clean segment signal is first multiplied by the down chirp of the corresponding length. The specific process is as follows shown;

Among them: Cu is upchirp, Cd is downchirp, f _o is the initial frequency of Cu, Hz; k is the frequency change rate, Hz/s; BW is the frequency bandwidth, Hz; t is time, and the value range is from 0 to t _{1_2} , s; fBw/2 is the starting frequency of Cd, Hz; j is the imaginary part signal.

Then perform Fourier transform on the multiplied signal to obtain the initial frequency f ₀ .

Then, according to the obtained starting frequency f ₀ of the first chirp symbol A1 where the previous data packet A collides with the subsequent data packet B, the frequency domain peak frequency point and the frequency domain peak height of the symbol A1 are estimated.

Among them, f _A1-1 and f _A1-2 are the peak frequency points in the frequency domain of symbol A1, Hz; h _A1-1 and h _A1-2 are their corresponding peak heights; f ₀ is the starting frequency of symbol A1, Hz ; BW is the frequency bandwidth, Hz; h is the complete peak energy height of the chirp frequency domain.

According to the frequency domain peak frequency point and frequency domain peak height estimated by symbol A1, calculate the frequency domain peak frequency point and frequency domain peak height of the interference generated by the chirp symbol B1 that will collide with A1 in the subsequent data packet B.

Specifically, the calculation process of the peak frequency point in the frequency domain where the interference is located is as follows:

Among them, f′ _A1-1 and f′ _A1-2 are the peak frequency points in the frequency domain of the interference generated by symbol A1 on B1, Hz; f _A1-1 and f _A1-2 are the peak frequency points in the frequency domain of symbol A1, Hz; BW is the frequency bandwidth, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s.

Specifically, the calculation process of the peak height in the frequency domain where the interference is located is as follows:

时：when

hour:

时：when

hour:

Among them, h′ _A1-1 and h′ _A1-2 are the frequency-domain peak heights of the interference caused by symbol A1 to B1; h _A1-1 and h _A1-2 are the frequency-domain peak heights of symbol A1, in Hz; h is The peak energy height of the complete chirp frequency domain, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s.

Next, first demodulate the chirp symbol B1 that collides with A1 in the latter data packet B, then perform Fourier transform on the multiplied signal to obtain its frequency domain characteristics, and estimate the frequency domain interference obtained in the frequency domain (f' _A1-1 , h' _A1-1 ), (f' _A1-2 , h' _A1-2 ) are eliminated.

In the frequency domain after eliminating the interference position, find the frequency point f where the frequency domain peak is the highest,

时：when

hour:

时：when

hour:

Among them: f is the frequency point of the highest peak in the frequency domain of the chirp symbol B1 after the interference is eliminated, Hz; BW is the frequency bandwidth, Hz; f _0-B is the starting frequency of the chirp symbol B1, Hz; find the frequency of the chirp symbol B1 The starting frequency is to complete the demodulation of chirp symbols.

Find the start frequency of the B1 symbol, that is, complete the demodulation of the conflicting symbols. Similarly, use the starting frequency of the B1 symbol to repeat steps 4.4 and 4.5 to complete the demodulation of the A2 symbol.

By analogy, iterative estimation and elimination of interference in the frequency domain is performed on conflicting signals.

Embodiment one:

This embodiment proposes a LoRa splicing communication method based on segmentation neural network decoding, including the following steps:

Step 1: The standard LoRa signal packet contains the following parts: the pilot used for signal detection, that is, the Preamble, the start frame delimiter (SFD) used for signal synchronization, and the payload data part that records the original transmission information Payload. In the signal sending stage, this method does not change the sending signal, and proceeds according to the standard frame format specified by LoRaWAN, so that it retains the original sending settings and communication protocol, does not conflict with the original communication protocol, and ensures that the system can communicate with the original LoRa communication systems are compatible.

Step 2: Use the gateway to receive the conflicting signal, and the gateway receives the conflicting signal based on the frequency domain characteristics of the pilot part; use the frequency domain characteristics of the pilot signal after pulse compression to correlate with the frequency domain characteristics of the prior signal If the degree of correlation is higher than the degree of correlation analysis between the noise and the frequency domain characteristics of the prior signal, it is determined that it is a LoRa signal and the data packet is received.

Step 3: Perform data preprocessing on the received signal. Perform high-frequency noise filtering on the LoRa signal to reduce the impact of out-of-band noise, and signal segmentation means dividing each chirp signal according to the length of time and correcting the carrier frequency offset of the received signal;

Step 3.1: First, filter the signal using an IIR low-pass filter to remove high-frequency noise.

Step 3.2: Signal synchronization and payload segmentation. Since the splicing mechanism preserves the preamble of the original LoRa signal, LoRa standard packet detection can be used to ensure signal synchronization. That is, the sliding window is realized by using a window size of 10 downchirps to monitor the channel. Then, the sliding window of channels is multiplied by 2.25 upchirps and a Fourier transform is performed to synchronize the signals. When the fast Fourier transform peak strength of the signal exceeds the preset threshold, the signal synchronization is completed, and this moment is the exact initial position p _i of the payload part of the signal, and the peak frequency point position is the carrier frequency offset frequency f _cfo . According to the length of the signal, starting from the exact starting position p _i of the payload part, the signal in the payload part is divided to realize the payload acquisition, and at the same time, the frequency-domain peak height h of the chirp symbol is obtained according to the pilot part, which is used for Calculation in subsequent steps.

Step 3.3: Use the pilot part of the LoRa data packet to estimate the carrier frequency offset and perform frequency correction to remove the noise caused by the mismatch between the local crystal oscillators between the transceivers and the Doppler frequency shift during signal transmission. frequency offset.

Among them: S′ is the chirp symbol after carrier frequency offset correction, S _chirp is the original chirp symbol, f ₀ is the starting frequency of the original chirp symbol, Hz, f _cfo is the carrier frequency offset frequency, Hz; k is the frequency change Rate, Hz/s; t is time, s; j is imaginary part signal.

Step 4: Iteratively estimate the interference in the frequency domain and eliminate the interference, so as to realize accurate demodulation of conflicting signals.

Step 4.1: For the chirp signal preprocessed in step 3, calculate the time window offset t _{1_2} generated by the signal arrival time difference at the chirp symbol level according to the starting position between the signal synchronizations.

t _{1_2} ＝p ₂ -p ₁ -N×T

Among them, the value range of the time window offset t _{1_2} is between 0 and T, s; p ₂ is the arrival time of the second packet, s; p ₁ is the arrival time of the first packet, s; N is positive Integer, T is the duration of the chirp symbol, s.

Step 4.3: Demodulate the clean segment of the first chirp symbol A1 where the previous data packet A collides with the subsequent data packet B. The process is as follows: the clean segment signal is first multiplied by the down chirp of the corresponding length. The specific process As shown in the following formula;

Among them: Cu is upchirp, Cd is downchirp, f _o is the initial frequency of Cu, Hz; k is the frequency change rate, Hz/s; BW is the frequency bandwidth, Hz; t is time, and the value range is from 0 to t _{i_2} , s; fBw/2 is the starting frequency of Cd, Hz; j is the imaginary part signal.

Then perform Fourier transform on the multiplied signal to obtain the initial frequency f ₀ .

Step 4.4: According to the starting frequency f ₀ of the first chirp symbol A1 where the previous data packet A collides with the subsequent data packet B obtained in step 4.3, estimate the frequency domain peak frequency point and frequency domain peak height of the symbol A1.

Among them, f _A1-1 and f _A1-2 are the peak frequency points in the frequency domain of chirp symbol A1, Hz; h _A1-1 and h _A1-2 are their corresponding peak heights; f ₀ is the starting frequency of chirp symbol A1 , Hz; BW is the frequency bandwidth, Hz; h is the complete peak energy height of the chirp frequency domain.

Step 4.5 Calculate the frequency domain peak frequency point and frequency domain peak value of the interference generated by the chirp symbol B1 that will collide with A1 in the next data packet B according to the estimated frequency domain peak frequency point and frequency domain peak height of symbol A1 high.

Specifically, the calculation process of the peak frequency point in the frequency domain where the interference is located is as follows:

Among them, f′ _A1-1 and f′ _A1-2 are the peak frequency points in the frequency domain of the interference generated by symbol A1 on B1, Hz; f _A1-1 and f _A1-2 are the peak frequency points in the frequency domain of symbol A1, Hz; BW is the frequency bandwidth, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s.

Specifically, the calculation process of the peak height in the frequency domain where the interference is located is as follows:

时：when

hour:

时：when

hour:

Among them, h′ _A1-1 and h′ _A1-2 are the frequency-domain peak heights of the interference caused by symbol A1 to B1; h _A1-1 and h _A1-2 are the frequency-domain peak heights of symbol A1, in Hz; h is The peak energy height of the complete chirp frequency domain, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s.

Step 4.5: Iterative cancellation of frequency domain interference

First, demodulate the chirp symbol B1 that conflicts between the latter data packet B and A1, and the process is as follows: the B1 signal is first multiplied by the down chirp of the corresponding length, and the specific process is shown in the following formula;

Among them: Cu is upchirp, Cd is downchirp, f _o is the initial frequency of Cu, Hz; k is the frequency change rate, Hz/s; BW is the frequency bandwidth, Hz; t is time, the value range is from 0 to T, s; f _BW /2 is the starting frequency of Cd, Hz; j is the imaginary part signal.

Then perform Fourier transform on the multiplied signal to obtain its frequency domain characteristics, and in the frequency domain, the frequency domain interference (f′ _A1-1 , h′ _A1-1 ), (f′ _{A1 -2} , h′ _A1-2 ) for elimination.

In the frequency domain after eliminating the interference position, find the frequency point f where the frequency domain peak is the highest,

时：when

hour:

时：when

hour:

Among them: f is the frequency point of the highest peak in the frequency domain of the chirp symbol B1 after the interference is eliminated, Hz; BW is the frequency bandwidth, Hz; f _0-B is the starting frequency of the chirp symbol B1, Hz; find the frequency of the chirp symbol B1 The starting frequency is to complete the demodulation of chirp symbols.

Find the starting frequency of the chirp symbol B1, that is, complete the demodulation of the conflicting symbols.

Similarly, use the starting frequency of the chirp symbol B1 to repeat steps 4.4 and 4.5 to complete the demodulation of the A2 symbol.

By analogy, iterative estimation and elimination of interference in the frequency domain is performed on conflicting signals.

Using this method, the LoRa conflict signal demodulation is carried out under different parameter environments. The verification experiments carried out in this section all use the same transceiver equipment. The sending end uses the Arduino module equipped with the sx1276 chip to transmit the LoRa signal, and the receiving end uses the USRP-2954 manufactured by the American NI company to complete the receiving of the LoRa signal. The frequency band selected for signal transmission is 915.9MHz, and the signal bandwidth is 500KHz. The demodulation part of the conflicting data packets is implemented on MATLAB. The experiment adjusts the signal-to-noise ratio of the signal by adjusting the transmitting power of the sending node, changing the transmission distance between the sending end and the receiving end, adding an attenuator at the receiving end and changing the attenuation power.

When SF=8 and SNR is between -5dB～-10dB, the demodulation accuracy rate of the conflicting signal can reach 89.06%. When SF=10 and SNR is between -15dB～-20dB, the demodulation accuracy rate of the conflicting signal can reach 92.94%. As shown in FIG. 4 , when SF=8, the demodulation accuracy of conflicting signals under different signal-to-noise ratios.

The preferred embodiments of the present disclosure have been described in detail above, but the present disclosure is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present disclosure, various simple modifications can be made to the technical solutions of the present disclosure. These simple modifications All belong to the protection scope of the present disclosure.

In addition, it should be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner if there is no contradiction. The combination method will not be described separately.

In addition, various implementations of the present disclosure can be combined arbitrarily, as long as they do not violate the idea of the present disclosure, they should also be regarded as the content disclosed in the present disclosure.

Claims (10)

1. A LoRa concurrent communication demodulation method based on frequency domain interference iterative elimination, is characterized in that, executes according to the following steps: Step 1: Signal sending; Taking the chirp signal as the processing unit, the node sends the signal in the standard LoRa packet frame format; Step 2: Signal reception; The gateway side uses the frequency domain characteristics of the pilot part as a basis to receive the LoRa signal in the conflict state. The conflict state means that multiple data packets are transmitted at the same time period; Step 3: Signal preprocessing; Low-pass filtering and signal synchronization are performed on the LoRa signal in the conflict state, and multiple chirp symbols are obtained by dividing with the chirp length as the step size, and carrier frequency offset correction is performed on the chirp symbols; Step 4: demodulation of conflicting signals; Perform frequency domain interference estimation and elimination on the chirp symbols preprocessed in step 3, and complete symbol demodulation. 2. The LoRa concurrent communication demodulation method based on frequency domain interference iterative elimination according to claim 1, characterized in that, in described step 4, the estimation of frequency domain interference includes obtaining frequency domain peak frequency point and frequency domain peak height; Among them, the acquisition of frequency domain peak frequency points includes:

Among them, f' _A1-1 , f' _A1-2 is the frequency domain peak frequency point of the interference generated by chirp symbol A1 to chirp symbol B1, Hz; f _A1-1 , f _A1-2 is the frequency domain of chirp symbol A1 Peak frequency point, Hz; BW is the frequency bandwidth, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s; The acquisition of peak height in frequency domain includes: when

hour:

when

hour:

Among them, _h'A1-1 , _h'A1-2 is the frequency domain peak height of the interference generated by chirp symbol A1 to chirp symbol B1; _hA1-1 , _hA1-2 is the frequency domain peak height of chirp symbol A1, Hz; h is the complete peak energy height of the chirp symbol in frequency domain, Hz. 3. The LoRa concurrent communication demodulation method based on frequency domain interference iterative elimination according to claim 1 or 2, is characterized in that, in described step 4, after the frequency domain interference is estimated, the symbol demodulation process comprises: Eliminate the estimated frequency domain interference (f' _A1-1 , h' _A1-1 ), (f' _A1-2 , h' _A1-2 ) in the frequency domain; In the frequency domain after eliminating the interference position, find the frequency point f where the frequency domain peak is the highest, when

hour:

when

hour:

Among them: f is the frequency point of the highest peak in the frequency domain of the chirp symbol B1 after the interference is eliminated, Hz; BW is the frequency bandwidth, Hz; f _0-B is the starting frequency of the chirp symbol B1, Hz; find the frequency of the chirp symbol B1 The starting frequency is to complete the demodulation of chirp symbols. 4. the LoRa concurrent communication demodulation method based on frequency domain interference iterative elimination according to claim 1 or 2, is characterized in that, in described step 3, the specific process that carrier frequency offset is corrected is shown in the following formula;

Among them: S′ is the chirp symbol after carrier frequency offset correction, S _chirp is the original chirp symbol, f ₀ is the starting frequency of the original chirp symbol, Hz, f _cfo is the carrier frequency offset frequency, Hz; k is the frequency change Rate, Hz/s; t is time, s; j is imaginary part signal. 5. the LoRa concurrent communication demodulation method based on frequency domain interference iterative elimination according to claim 1 or 2, is characterized in that, in step 3, described low-pass filtering adopts IIR low-pass filter to carry out. 6. the LoRa concurrent communication demodulation method based on frequency domain interference iterative elimination according to claim 1 or 2 is characterized in that, in step 3, the frequency domain energy of the LoRa signal after the low-pass filter is carried out pulse compression, utilizes The sliding window finds the energy maximum position after pulse compression, which is the starting position of the payload of the chirp signal, and realizes signal synchronization; use 10 downchirps to perform the sliding window; Use the sliding window to perform signal synchronization on the low-pass filtered LoRa signal, and segment the Payload signal in the low-pass filtered LoRa signal according to the time unit. 7. A LoRa concurrent communication demodulation system based on frequency domain interference iterative elimination, is characterized in that, comprising: The signal sending module takes the chirp signal as the processing unit, and the node sends the signal in the standard LoRa packet frame format; The signal receiving module, the gateway side uses the frequency domain characteristics of the pilot part as a basis to receive the LoRa signal in the conflict state. The conflict state means that multiple data packets are transmitted at the same time period; The signal preprocessing module performs low-pass filtering and signal synchronization on the LoRa signal in the conflict state, divides the chirp length into multiple chirp symbols, and corrects the carrier frequency offset of the chirp symbols; The conflicting signal demodulation module performs frequency domain interference estimation and elimination on the chirp symbols preprocessed in step 3, and completes symbol demodulation. 8. The LoRa concurrent communication demodulation system based on iterative elimination of frequency domain interference according to claim 7, characterized in that, in the demodulation module of the conflicting signal, the estimation of frequency domain interference includes obtaining the frequency domain peak frequency point and frequency domain peak height; Among them, the acquisition of frequency domain peak frequency points includes:

Among them, f' _A1-1 , f' _A1-2 is the frequency domain peak frequency point of the interference generated by chirp symbol A1 to chirp symbol B1, Hz; f _A1-1 , f _A1-2 is the frequency domain of chirp symbol A1 Peak frequency point, Hz; BW is the frequency bandwidth, Hz; t _{1_2} is the time window offset, s; T is the complete time length of the chirp symbol, s; The acquisition of peak height in frequency domain includes: when

hour:

when

hour:

Among them, _h'A1-1 , _h'A1-2 is the frequency domain peak height of the interference generated by chirp symbol A1 to chirp symbol B1; _hA1-1 , _hA1-2 is the frequency domain peak height of chirp symbol A1, Hz; h is the complete peak energy height of the chirp symbol in frequency domain, Hz. 9. The LoRa concurrent communication demodulation system based on frequency domain interference iterative elimination according to claim 8, wherein, after the frequency domain interference is estimated, the symbol demodulation process includes: Eliminate the estimated frequency domain interference (f' _A1-1 , h' _A1-1 ), (f' _A1-2 , h' _A1-2 ) in the frequency domain; In the frequency domain after eliminating the interference position, find the frequency point f where the frequency domain peak is the highest, when

hour:

when

hour:

Among them: f is the frequency point of the highest peak in the frequency domain of the chirp symbol B1 after the interference is eliminated, Hz; BW is the frequency bandwidth, Hz; f _0-B is the starting frequency of the chirp symbol B1, Hz; find the frequency of the chirp symbol B1 The starting frequency is to complete the demodulation of chirp symbols. 10. The LoRa concurrent communication demodulation system based on iterative elimination of frequency domain interference according to claim 7, 8 or 9, characterized in that, in the described signal preprocessing module, the specific process of correcting the carrier frequency offset is as follows shown in the formula;

Among them: S′ is the chirp symbol after carrier frequency offset correction, S _chirp is the original chirp symbol, f ₀ is the starting frequency of the original chirp symbol, Hz, f _cfo is the carrier frequency offset frequency, Hz; k is the frequency change Rate, Hz/s; t is time, s; j is imaginary part signal.

Images