Multipath Suppression Method Based on LoRa Segment Demodulation

Abstract

The present invention discloses a multipath suppression method based on LoRa segmented demodulation. After the modulated LoRa signal passes through a multipath channel, while performing non-coherent demodulation on the received signal, this multipath suppression method segments the received signal, then performs non-coherent demodulation separately, and subtracts the demodulation output of the received signal from the segmented demodulation output, which can suppress the amplitude of the multipath signal part in the demodulation domain and reduce the decision error probability. The multipath suppression method disclosed by the present invention can effectively reduce the detection error rate caused by multipath interference and reduce the performance attenuation caused by multipath interference in LoRa wireless communication. This multipath suppression method has low complexity, does not require complex channel estimation, and can adapt to time-varying channels at the same time.

Description

Technical Field

[0001] The present invention relates to the technical fields of spread spectrum communication and Internet of Things, and particularly relates to a multipath suppression method based on LoRa segmented demodulation. Background Art

[0002] LoRa (Long Range) is a wireless modulation technology used to create long-distance communication links, originating from Chirp Spread Spectrum (CSS) communication, and capable of achieving long-distance, low-power, and secure data transmission. Due to its long communication distance and strong anti-interference ability, Chirp spread spectrum technology has been applied in the military and space communication fields for decades. LoRa adopts Frequency Shift Chirp Modulation (FSCM) technology, maintaining the same low-power characteristics as FSK modulation and significantly increasing the communication range, and is the first low-cost implementation solution for commercial use. LoRaWAN is a communication protocol and system architecture based on LoRa physical layer modulation technology, providing a long-distance, low-power Internet of Things communication platform, specifically designed for sensors and applications that need to work for a long time and can send data over medium to long distances in different environments. LoRa technology can be applied to application scenarios such as smart cities, homes and buildings, communities, agriculture, metering and utilities, healthcare, the environment, and supply chain and logistics. Traditional network technologies such as cellular and Bluetooth either require high bandwidth or high power, or have limited coverage or cannot penetrate deep into indoor environments, while the emergence of LoRa has filled the technical gaps of traditional network technologies in these application fields.

[0003] Due to the complex environment in indoor, urban and other communication scenarios, the transmitted signal will propagate through multiple paths such as reflection and refraction, resulting in multipath effects. The inter-symbol interference (ISI) caused by multipath effects will lead to signal fading and phase shift, affecting the signal transmission quality. In wireless communication, technologies such as spread spectrum, Rake receiver, channel coding, interleaving, equalization and diversity are used to eliminate the interference caused by multipath. LoRa technology achieves a higher receiver sensitivity by increasing the signal bandwidth in exchange for a lower signal-to-noise ratio, combats multipath interference, and improves the communication coverage. In addition, techniques such as forward error correction (FEC) and interleaving can be used in LoRa modulation to further improve the robustness. However, in some indoor environments, such as indoor industrial and body area network communications, there are many obstacles, and the signal propagation mostly exhibits the characteristics of dense multipath. The channel estimation methods for dense multipath are not perfect, and using techniques such as equalization and diversity will increase the computational complexity of the system and even increase the energy consumption of communication devices. Therefore, seeking effective low-cost techniques to eliminate multipath has always been a concern of researchers. Summary of the Invention

[0004] The purpose of the present invention is to reduce the performance attenuation caused by multipath interference in LoRa wireless communication, and to provide a multipath suppression method based on LoRa segmented demodulation, which can adapt to time-varying channels and has the characteristics of low complexity.

[0005] The purpose of the present invention can be achieved by adopting the following technical solutions:

[0006] A multipath suppression method based on LoRa segmented demodulation, the multipath suppression method includes the following steps:

[0007] Non-coherent demodulation: The receiving end synchronizes the received signal transmitted through the channel, multiplies the synchronized received signal r(k) by the conjugate of the base signal Then perform discrete Fourier transform and then take the absolute value to obtain the non-coherent demodulation output X of the received signal q , where the base signal is the LoRa modulation signal corresponding to when the LoRa symbol is 0, the LoRa symbol is a multi-level symbol mapped from a binary information source, and the LoRa modulation signal is the time-domain signal after the LoRa symbol is modulated by LoRa;

[0008] Segmented demodulation: While performing non-coherent demodulation on the received signal, the receiving end segments the received signal r(k), and then performs non-coherent demodulation on the signals of each segment;

[0009] Joint demodulation: At the receiving end, joint demodulation is performed. The demodulation outputs of each segment are subtracted to eliminate the direct path signal part in the segment demodulation output, and the demodulation output of the multipath signal is separated. Then, the demodulation output of the received signal is subtracted from the separated demodulation output of the multipath signal, and the absolute value is taken to obtain the joint demodulation output. Then, the decision of the LoRa symbol is made. Among them, the decision of the LoRa symbol is to detect the maximum amplitude of the joint demodulation output, and the index of the position where the maximum amplitude appears is determined as the received LoRa symbol.

[0010] Further, the received signal after synchronization in the non-coherent demodulation is the received signal passing through the multipath channel, which can be modeled as the following formula:

[0011]

[0012] where h0 represents the direct path attenuation, x a (k) is the transmitted LoRa modulation signal, L represents the number of multipaths, i represents the path sequence number, h i and k i represent the attenuation and delay of the i-th path. Denote I = {k i , i = 1, …, L - 1} as the delay set of the multipath interference, and δ(k - k i ) represents the impulse function with a delay of k i . The received signal can be divided into two parts, the direct path signal part and the multipath signal part. The delay of the direct path signal part is 0, and each path in the multipath signal part has a certain delay relative to the direct path signal part.

[0013] Further, in the non-coherent demodulation, the receiving end multiplies the synchronized received signal r(k) by the conjugate of the base signal and then performs a discrete Fourier transform, and then takes the absolute value to obtain the demodulation output X q of the received signal. The calculation formula is as follows:

[0014]

[0015] where, represents the conjugate of the base signal, and DFT[·] represents the discrete Fourier transform. Since the LoRa modulation signal belongs to the spread spectrum signal, multiplying the received signal by is for despreading. The multiplied signal is a single-frequency signal, and the frequency of the signal is related to the transmitted LoRa symbol. The purpose of performing the discrete Fourier transform is to convert the detection of the single-frequency signal to the frequency domain. After the discrete Fourier transform, the single-frequency signal forms an impulse signal, and the transmitted LoRa symbol can be judged by detecting the position of this impulse signal.

[0016] Further, in the segmented demodulation, the segmentation follows the principle of equal division, and the number of segments is a power of 2. The received signals after segmentation are respectively subjected to non-coherent demodulation to obtain the demodulation outputs of each segment. The impulse signals formed by the direct-path signal parts in the demodulation domains of the demodulation outputs of each segment are completely identical in amplitude and position. This is because LoRa modulation has the cyclic shift property, and the direct-path signal parts are evenly distributed in each segment. The impulse signals formed by the multipath signal parts in the demodulation domains of the demodulation outputs of each segment are not identical in amplitude and position. This is due to the time delay between the multipath signals and the direct-path signals. This difference enables the obtained segmented demodulation outputs to be used for joint demodulation to separate the demodulation output of the multipath signal.

[0017] Further, in the segmented demodulation, the receiving end segments the synchronized received signal and then performs non-coherent demodulation. Among them, the received signal is divided into two segments, and the signals after segmentation are processed as follows

[0018]

[0019] where r f (k) is the first half of the signal after segmentation of the received signal, and r s (k) is the second half of the signal after segmentation of the received signal. u(k) is the step function, and u(k - M + 1) is the step function obtained by delaying u(k) by the corresponding time. M = 2 SF , SF is the spreading factor in LoRa modulation. The two segments of signals are respectively subjected to non-coherent demodulation, and the demodulation outputs of the segmented demodulation are obtained as follows:

[0020]

[0021] where X f is the demodulation output of the first half, and X s is the demodulation output of the second half.

[0022] Further, in the joint demodulation, the receiving end performs joint demodulation on the demodulation output of the received signal and the demodulation outputs of each segment. First, the demodulation outputs of each segment are subtracted to separate the demodulation output X d of the multipath signal. The calculation formula is as follows:

[0023] X d = |X f - X s |

[0024] Since the direct-path signal parts are completely identical in the demodulation outputs of the two segments, X d does not contain the direct-path signal part, while for the multipath signal part, due to the difference in the demodulation outputs of the two segments, X dOnly a part of the multipath signals exist in it, that is, X d is the demodulation output of a part of the multipath signals separated from the received signal. Then, use the non-coherent demodulation output X of the received signal q and the demodulation output X of the separated multipath signals d to make a difference to obtain the joint demodulation output Y q as follows:

[0025] Y q =|X q -X d | Finally, the receiving end performs detection and decision according to the joint demodulation output Y q as follows:

[0026]

[0027] wherein, is the LoRa symbol after decision-making through segmented demodulation, and arg max(·) represents finding and outputting the corresponding independent variable value for the elements in the brackets.

[0028] The present invention has the following advantages and effects compared with the prior art:

[0029] (1) The multipath suppression method disclosed by the present invention can suppress the amplitude formed by multipath signals in the demodulation domain, thereby reducing the interference of multipath signals, effectively reducing the detection error rate caused by multipath, and improving the communication performance of LoRa technology under multipath channels.

[0030] (2) The multipath suppression method disclosed by the present invention has low complexity. It only segments the received signal and performs non-coherent demodulation on the segmented received signal. The multipath suppression method is simple and easy to use.

[0031] (3) The multipath suppression method disclosed by the present invention can be applied to both time-varying and non-time-varying multipath channels without complex channel estimation, and has a wide range of application scenarios. Description of the Drawings

[0032] The drawings described herein are used to provide a further understanding of the present invention, form a part of this application, and the schematic embodiments of the present invention and their descriptions are used to explain the present invention, and do not constitute an improper limitation to the present invention. In the drawings:

[0033] Figure 1 is the flow block diagram of realizing 2-segment demodulation of the multipath suppression method based on LoRa segmented demodulation in Embodiment 1;

[0034] Figure 2 is the schematic diagram of energy aggregation of non-coherent demodulation in Embodiment 1;

[0035] Figure 3It is a schematic diagram of the demodulation output of the received signal affected by multipath interference in Embodiment 1;

[0036] Figure 4 It is a time-frequency schematic diagram of the received signal after multipath superposition in Embodiment 2;

[0037] Figure 5 It is a schematic diagram of the demodulation output of the received signal in Embodiment 2;

[0038] Figure 6 It is a time-frequency schematic diagram of the first half of the segmented received signal in Embodiment 2;

[0039] Figure 7 It is a schematic diagram of the demodulation output of the first half of the segmentation in Embodiment 2;

[0040] Figure 8 It is a time-frequency schematic diagram of the second half of the segmented received signal in Embodiment 2;

[0041] Figure 9 It is a schematic diagram of the demodulation output of the second half of the segmentation in Embodiment 2;

[0042] Figure 10 It is a schematic diagram of the demodulation output of the separated multipath in Embodiment 2;

[0043] Figure 11 It is a schematic diagram of the combined demodulation output after segmented demodulation in Embodiment 2. Detailed implementation manners

[0044] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0045] Embodiment 1

[0046] This embodiment discloses a multipath suppression method based on LoRa segmented demodulation. While the receiving end demodulates the received signal, the receiving end segments the received signal, then demodulates the segmented signal, and subtracts the demodulation output of the received signal from the demodulation output of the segmented signal, which can reduce the amplitude of the multipath signal in the demodulation domain and achieve the purpose of suppressing multipath. The flow block diagram of the multipath suppression method is as Figure 1 shown. The specific steps are as follows:

[0047] Non-coherent demodulation: As a frequency shift Chirp spread spectrum technology, LoRa modulation is an M-ary digital modulation method in the frequency band range It can transmit M symbols, where f0 and B are the center frequency and signal bandwidth respectively, and M = 2 SF , SF is the spreading factor, and the signal duration is Within each symbol period, if the entire bandwidth B is divided into M parts, then the starting frequency of the LoRa modulation signal corresponding to the LoRa symbol a (a ∈ {0, 1,..., M - 1}) to be transmitted is The instantaneous frequency of the LoRa modulation signal linearly increases from the starting frequency to Then it jumps to And then continues to linearly increase. The LoRa modulation signal x a (k) is calculated as follows:

[0048]

[0049] Where a is the LoRa symbol and k is the independent variable in the discrete time domain. One of the commonly used demodulation methods in LoRa modulation is non - coherent demodulation. The non - coherent demodulation method is that the receiving end first multiplies the received signal r(k) by Where represents the conjugate of the base signal, then performs a discrete Fourier transform (Discrete Fourier Transform, DFT), and then takes the absolute value. For a Gaussian flat channel, that is, when r(k) = x a (k), the demodulation output X q is obtained, and the calculation formula is as follows:

[0050]

[0051] Where q represents the independent variable in the demodulation domain, and δ(q - a) represents the impulse function with a time delay of a. Combining the above formula, the energy of a LoRa symbol in the demodulation domain is concentrated at one position, forming an impulse signal δ(q - a). As Figure 2 shown, the position of the impulse signal is related to the value of the LoRa symbol a, and the amplitude of the impulse signal is related to the energy of the LoRa modulation signal and the channel gain. When q ≠ a, the amplitude is very low. Therefore, the LoRa symbol can be judged by detecting the position of the maximum amplitude on the demodulation output X q , and the judgment method is as follows:

[0052]

[0053] Where is the judged LoRa symbol.

[0054] When the LoRa modulated signal passes through a multipath channel, the direct-path signal received at the receiving end will be interfered by the LoRa modulated signal copies with different arrival times. This kind of multipath signal will also form a signal with a large amplitude at q≠a in the demodulation domain. Under the influence of additive noise, it will greatly increase the symbol misjudgment probability and lead to the degradation of communication performance. The received signal passing through the multipath channel can be modeled by the following formula:

[0055]

[0056] where h0 represents the direct-path attenuation, L represents the number of multipaths, i represents the path sequence number, h i and k i represent the attenuation and delay of the i-th path. Denote I={k i ,i = 1,…,L - 1} as the delay set of multipath interference, and δ(k - k i ) represents the impulse function with a delay of k i .

[0057] After the receiving end synchronizes the received signal, it demodulates the signal to obtain the demodulation output X q , and the calculation formula is as follows:

[0058]

[0059]

[0060] Since the second term in the above formula represents the demodulation output of the multipath signal, the main lobe width is where M = 2 SF and the SF value range is 7 - 12. Therefore, the main lobe width formed by the multipath signal is relatively narrow and the main energy is concentrated in the main lobe, which is different from the position formed by the direct-path signal in the demodulation domain. Specifically, as shown in Figure 3 , the demodulated signal can be approximated as the superposition of the direct-path signal and the demodulated signals of each multipath signal respectively.

[0061] Segmented demodulation: While the receiving end performs non-coherent demodulation on the received signal, it divides the received signal into two segments as follows:

[0062]

[0063] where r f (k) is the first half segment signal after the received signal is segmented, r s (k) is the second half segment signal after the received signal is segmented, u(k) is the step function, and u(k - M + 1) is the step function of u(k) with the corresponding delay. Then the receiving end performs non-coherent demodulation on these two segments of signals to obtain the demodulation output of the segmented demodulation as follows:

[0064]

[0065] Where X f is the demodulation output of the first half, and X s is the demodulation output of the second half. After demodulating the first half signal, the demodulation output X f is obtained, and the calculation formula is as follows:

[0066]

[0067] Where I left ={k i : 0 < k i < M / 2}, representing the multipath signals that will appear in the demodulation domain of the first half, denoted as the first half multipath signals. After demodulating the second half signal, the demodulation output X s is obtained, and the calculation formula is as follows:

[0068]

[0069] Where I right ={k i : M / 2 ≤ k i < M - 1}, representing the multipath signals that will appear in the demodulation domain of the second half, denoted as the second half multipath signals. Since the received signal is divided into two segments of signals, whether it is the direct path signal part or the multipath signal part in the demodulation output X q will be distorted in the segmented demodulation outputs X f and X s . When distorted, the amplitude of the signal changes but the position remains unchanged. The difference is that the direct path signal is evenly divided into two parts, while the multipath signal is unevenly divided into two parts due to the existence of time delay. The direct path signal evenly divided into two parts forms exactly the same amplitude on the segmented demodulation outputs X f and X s , but the amplitudes formed by the multipath signal on the segmented demodulation outputs X f and X s are inconsistent. This characteristic can be used to separate the direct path signal and the multipath signal.

[0070] Joint demodulation: Since the direct path signal is distorted in the two segmented demodulation outputs, but the positions and amplitudes of the distorted signals are exactly the same, that is, the first terms of the formulas of X f and X s are exactly the same. However, due to the existence of time delay, the multipath parts on the two segmented demodulation outputs are inconsistent, not only different in amplitude. When the multipath time delay k i ≥ M / 2, the multipath will only appear in the demodulation output X s of the second half. Thus, the receiving end combines Xf and X s Taking the difference with X, the demodulation output X for separating the multipath signal can be obtained d , and the calculation formula is as follows:

[0071]

[0072] X d is equivalent to separating a part of the multipath signal from the overlap of the direct-path signal and the multipath signal. Taking the difference between X d and X q can obtain the joint demodulation output Y q , and the calculation formula is as follows:

[0073]

[0074] Y q Relative to X q , the direct-path signal remains unchanged, but the number of multipaths is reduced. The latter half of the multipath signal is completely eliminated, leaving only the first half of the multipath signal, and the amplitude in the demodulation domain is suppressed. Finally, the receiving end performs detection and decision based on the joint demodulation output:

[0075]

[0076] where is the LoRa symbol determined after segmented demodulation.

[0077] Embodiment 2

[0078] This embodiment discloses a multipath suppression method based on LoRa segmented demodulation, which specifically includes the following steps:

[0079] S1. Non-coherent demodulation: The transmitting end sends a LoRa symbol with a symbol of a. The LoRa modulated signal x a (k) has the following calculation formula:

[0080]

[0081] where a is the LoRa symbol, k is the independent variable in the discrete time domain, and M is the number of possible transmitted symbols. The spreading factor SF is set to 9, then M = 2 SF = 512, and the transmitted symbol a = 256. For a LoRa signal passing through a channel with two-path multipath interference, the received signal r(k) obtained by the receiving end is as follows:

[0082] r(k) = h0x a (k) + h1x a (k - k1) + h2x a (k - k2)

[0083] Among them, h0 represents the attenuation of the direct path, h1 and h2 represent the attenuation of the multipath, and k1 and k2 represent the time delay of the multipath. The parameter settings are: h0 = h1 = 1, k1 = 128, k2 = 384.

[0084] At the receiving end, the received signal passing through the multipath channel is received, and after synchronizing the direct path signal, demodulation is performed to obtain the demodulation output X q , and the calculation formula is as follows:

[0085]

[0086] Among them, q represents the independent variable in the demodulation domain, and δ(q - a) represents the impulse function with a time delay of a. The time-frequency schematic diagram of the received signal after multipath superposition is as Figure 4 shown. The solid line represents the direct path signal, and the dashed line represents the multipath signal. The obtained demodulation output is as Figure 5 shown. The direct path signal will form an impulse signal (amplitude of 512) at the position of q = 256. The multipath signal will form a certain amplitude at the positions of q = a - k1 = 128 and q = a - k2 = -128 (due to the cyclic shift characteristic of the LoRa modulation signal, corresponding to {(a - k2) mod M} = 384) (the amplitudes are 384 and 128.8 respectively). The higher the amplitude formed at the position where q ≠ a, the more likely it is to produce misjudgment due to the influence of noise during decision-making, thus causing attenuation of the communication performance.

[0087] S2. Piecewise demodulation: After receiving the signal passing through the multipath channel and synchronizing it, while performing non-coherent demodulation, the receiving end divides the received signal into two segments as follows:

[0088]

[0089]

[0090] Among them, r f (k) is the first half of the signal after the received signal is segmented, r s (k) is the second half of the signal after the received signal is segmented, and u(k) is the step function, and u(k - M + 1) are the step functions of u(k) with corresponding time delays. Then the receiving end demodulates these two segments of signals respectively to obtain the demodulation output X f , and the calculation formula is as follows:

[0091]

[0092] The demodulation output X s of the second half is as follows:

[0093]

[0094]

[0095] The time-frequency schematic diagram of the first half of the signal is as shown in Figure 6 where the solid line represents the direct path signal part in the first half, and the dashed line represents the multipath signal part in the first half. The demodulation output of the first half is as shown in Figure 7 It can be seen that there is only a direct path signal (amplitude 256) at q = 256 and a multipath signal (amplitude 128) at q = 128, and there is no multipath signal at q = 384. The time-frequency schematic diagram of the second half of the signal is as shown in Figure 8 where the solid line represents the direct path signal part in the second half, and the two-hop dashed line represents the multipath signal part in the second half. The demodulation output of the second half is as shown in Figure 9 It can be seen that there is a direct path signal (amplitude 256) at q = 256, a multipath signal (amplitude 128) at q = 128, and a multipath signal (amplitude 128.8) at q = 384.

[0096] S3. Joint demodulation: The direct path signal is distorted in the two segmented demodulation domains. The distorted signals are both at q = a and have the same amplitude. However, the multipath with a time delay k2 = 384 only appears in the second half demodulation domain. The receiver subtracts X f and X s to obtain the demodulation output X d for separating the multipath signal. The calculation formula is as follows:

[0097]

[0098] The demodulation output of the multipath separated in the demodulation domain is as shown in Figure 10 It can be seen that there are only multipath signals (amplitudes 128 and 128.8 respectively) at q = 128 and q = 384, and there is no direct path signal at q = 256. Then subtracting X d from Y q can obtain the joint demodulation output Y q The calculation formula is as follows:

[0099]

[0100] Y q Compared with Y q , the multipath signal with a time delay k2 = 384 is completely eliminated, and the amplitude of the multipath signal with a time delay k1 = 128 is suppressed in the demodulation domain. The demodulation output after segmented demodulation is as shown in Figure 11 At q = 384, the multipath signal is completely eliminated, and the amplitude of the multipath signal at q = 128 changes from 384 (Figure 5 as shown) is reduced to 256, while the direct path signal remains unchanged. Suppressing the amplitude of the multipath signal part in the demodulation domain can reduce the possibility of misjudgment during decision-making and improve communication performance. Finally, the receiving end performs detection and decision-making based on the output of joint demodulation:

[0101]

[0102] where is the LoRa symbol determined after segmented demodulation.

[0103] In summary, this embodiment can suppress the amplitude of multipath in the LoRa demodulation domain and reduce the error code caused by multipath interference.

[0104] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope of the present invention.

Claims

1. A multipath suppression method based on LoRa segmented demodulation, characterized in that, The multipath suppression method includes the following steps: Non - coherent demodulation: The receiving end synchronizes the received signal transmitted through the channel, multiplies the synchronized received signal r(k) by the conjugate of the base signal Then, perform a discrete Fourier transform and then take the absolute value to obtain the non - coherent demodulation output X of the received signal q , where the base signal is the LoRa modulation signal corresponding to when the LoRa symbol is 0, the LoRa symbol is a multi - level symbol mapped from a binary information source, and the LoRa modulation signal is the time - domain signal after the LoRa symbol is modulated by LoRa; Segmented demodulation: While performing non-coherent demodulation on the received signal, the receiving end segments the received signal r(k), and then performs non-coherent demodulation on the signal of each segment; Joint demodulation: The receiving end performs joint demodulation, subtracts the demodulation outputs of each segment to eliminate the direct path signal part in the segmented demodulation output, separates the demodulation output of the multipath signal, then subtracts the separated demodulation output of the multipath signal from the demodulation output of the received signal, takes the absolute value to obtain the joint demodulation output, and then makes a decision on the LoRa symbol. Among them, the decision on the LoRa symbol is to detect the maximum amplitude of the joint demodulation output, and the index of the position where the maximum amplitude appears is determined as the received LoRa symbol.

2. The multipath suppression method based on LoRa segmented demodulation according to claim 1, characterized in that, In the segmented demodulation, the segmentation follows the principle of equal division, the number of segments is a power of 2, and the received signals after segmentation are respectively subjected to non-coherent demodulation to obtain the demodulation outputs of each segment.

3. The multipath suppression method based on LoRa segmented demodulation according to claim 1, wherein The non-coherent demodulation is that the receiving end multiplies the received signal r(k) after synchronization by the conjugate of the base signal and then performs a discrete Fourier transform to obtain the demodulation output X of the received signal q , and the calculation formula is as follows: where represents the conjugate of the base signal, and DET[·] represents the discrete Fourier transform 4. The multi-path suppression method based on LoRa segmented demodulation according to claim 2, wherein The segmented demodulation is that the receiving end segments the synchronized received signal and then performs non-coherent demodulation. Among them, the received signal is divided into two segments, and the processed signals after segmentation are as follows where r f (k) is the first half of the segmented received signal, and r s (k) is the second half of the segmented received signal, u(k) is the step function, and u(k - M + 1) is the step function with a corresponding time delay of u(k), where M = 2 SF , SF is the spreading factor in LoRa modulation. The two signals are demodulated non - coherently respectively, and the demodulation outputs of the segmented demodulation are as follows: Among them, X f is the demodulation output of the first half, and X s is the demodulation output of the second half.

5. The multipath suppression method based on LoRa segmented demodulation according to claim 4, wherein The joint demodulation is to perform joint demodulation on the demodulation output of the received signal and the demodulation outputs of each segment. First, the demodulation outputs of each segment are subtracted to separate the demodulation output X of the multipath signal d , and the calculation formula is as follows: X d = |X f -X s | Then, use the non-coherent demodulation output X of the received signal q to perform a difference operation with the demodulation output X of the separated multipath signals d to obtain the joint demodulation output Y q as follows: Y q = |X q - X d | Finally, the receiving end performs detection and decision based on the joint demodulation output Y q as follows: Among them, is the LoRa symbol after segmented demodulation and decision, and arg max(·) represents finding the value of the independent variable corresponding to the maximum value in the elements in the parentheses and outputting it.

Images