CN109842430B

CN109842430B - LoRa modulation signal demodulation method under multipath channel

Info

Publication number: CN109842430B
Application number: CN201910286472.0A
Authority: CN
Inventors: 刘祖军; 郭玉蓉
Original assignee: Xidian University
Current assignee: Xidian University
Priority date: 2019-04-10
Filing date: 2019-04-10
Publication date: 2020-09-08
Anticipated expiration: 2039-04-10
Also published as: CN109842430A

Abstract

The invention discloses a demodulation method of LoRa modulation signals under a multipath channel, which mainly comprises the following stepsThe problem that in the prior art, the performance of directly demodulating LoRa modulation signals is poor is solved. The implementation scheme is as follows: directly demodulating the down-sampled preamble signal and the load signal to obtain a directly demodulated preamble signal A and a directly demodulated load signal D; searching the maximum value in A, reserving M points around the maximum value to obtain a preamble signal C with the length of (2M +1) after interception, and making C smaller than the maximum value

Resetting the point data to 0 to obtain a new leading signal B; carrying out matched filtering on the new preamble signal B and the load signal D after direct demodulation to obtain a new load signal E; and E, searching a position P corresponding to the maximum value of each symbol, and sequentially carrying out decimal to binary conversion and gray inverse mapping on the position P to obtain modulation information bits. The invention improves the demodulation performance of the system, and can be used for processing the LoRa modulation signal at the receiving end under the condition of a multipath channel.

Description

LoRa modulation signal demodulation method under multipath channel

Technical Field

The invention belongs to the technical field of communication, and particularly relates to a demodulation method of an LoRa modulation signal, which can be used for processing the LoRa modulation signal at a receiving end under the condition of a multipath channel.

Background

With the development of internet users from people to things, the internet of things IoT has attracted great interest. The internet of things enables objects around us to be connected with each other, and meanwhile, messages can be sent and received to the internet. To meet the various demands of internet of things applications, low power wide area LPWANs provide a novel communication paradigm. The LPWAN technology comprises an ultra-narrow band technology SigFox, a remote technology LoRa and the like, the technologies are proposed to meet the requirement of wide area connection of the Internet of things from several kilometers to dozens of kilometers, and the application of low data rate, low power consumption and low throughput is realized. In the recently proposed and dominant LPWAN technology, the semiconductor manufacturer Semtech has introduced its remote LoRa^TMThe advanced spread spectrum technology of the product series is widely applied. Unlike direct sequence spread spectrum DSSS, which modulates a message signal over a pseudo-random code, LoRa encodes the message signal using a swept frequency signal with a frequency that increases up-chirp or decreases down-chirp over time, and it can be seen that LoRa technology is essentially a derivative of chirp spread spectrum modulation.

The Chirp spread spectrum modulation technique CSS does not use a pseudo random code as a spreading code nor a frequency hopping pattern, but uses matched filtering and pulse compression characteristics of a Chirp signal to implement spread spectrum communication. In the field of wireless communication, chirp spread spectrum modulation technology is still a novel technology, and although the technology is listed as one of the physical standards of IEEE 802.15.4, the technology still has many defects. For example, if the capacity of a link is generally low if the chirp spread spectrum technique is used alone without combining other modulation techniques, the data transmission rate is improved by shortening the duration of the chirp signal or by using an overlapping technique, but both shortening the duration of the chirp signal and overlapping the signal increase the complexity and power consumption of the system.

In the LoRa modulation system, the conventional direct demodulation method is generally adopted for signal demodulation, and as shown in fig. 6, point multiplication is performed on a received signal and an original down-chirp, then fast fourier transform FFT is performed, and finally, determination of a maximum value is performed. The direct demodulation mode can obtain better demodulation performance under a single-path channel, but the demodulation performance of the system is sharply reduced under the condition that the channel is multipath, namely the bit error performance is reduced, so that the reliability of the system is reduced.

Disclosure of Invention

The present invention aims to provide a method for demodulating a LoRa modulated signal under a multipath channel to improve the bit error performance of a receiving end, thereby improving the reliability of a system.

In order to achieve the purpose, the technical scheme of the invention is as follows:

(1) performing corresponding point multiplication on the down-sampled preamble signal and the original down-chirp signal, and then sequentially performing FFT and absolute value calculation to obtain the length of 2 after direct demodulation^SFThe preamble signal a of (1), wherein SF is a spreading factor;

(2) intercepting and resetting data of the preamble signal after direct demodulation to obtain a new preamble signal B:

(2a) searching the maximum value of the preamble signal A after direct demodulation, reserving M points around the maximum value, intercepting the preamble signal C with the length of (2M +1) points, wherein the size of M is taken according to the condition of a channel and the range of M is 1 to 2^SF-1-1；

(2b) The value of each point of the preamble signal C after interception is compared with the maximum value

And (3) comparison: when the value at a certain point is greater than the maximum value

When the value of a certain point is less than the maximum value, the value of the point is kept unchanged

Then, the value of the point is reset to 0 to obtain a new preamble signal B, wherein the value of P is taken according to the condition of the channel,

(3) carrying out corresponding point multiplication on each symbol in the load signal after down sampling and the original down-chirp signal, and then sequentially carrying out FFT (fast Fourier transform) and taking an absolute value of each symbol to obtain a load signal D after direct demodulation;

(4) carrying out matched filtering operation on each symbol in the load signal D after direct demodulation and the new preamble signal B to obtain a new load signal E;

(5) and searching a position corresponding to the maximum value of each symbol in the new load signal E, and sequentially carrying out decimal to binary conversion and gray inverse mapping on the position to obtain modulation information bits.

Compared with the prior art, the invention has the following advantages:

1. improving the reliability of the system

In the invention, the channel information contained in the preamble signal is fully utilized through the matched filtering operation of the preamble signal and the load signal at the receiving end, thereby reducing the influence of the channel; meanwhile, in the process of obtaining the preamble signal, noise is reduced by intercepting the preamble signal and resetting data; compared with the traditional direct demodulation method, the method and the device have the advantages that the received load signals are directly demodulated, the influence of channels and noise is less, the obtained demodulation performance is higher, and the reliability of the LoRa modulation signal demodulation system can be improved.

2. Good bit error performance

Experiments prove that under the condition that the channel is set to be two fixed paths and SF is 7, the error rate is 10^-3Compared with the traditional direct demodulation method, the performance of the invention is improved by 1.5dB, and compared with the performance under an additive white Gaussian noise channel, the performance of the invention is only 0.5dB away.

Drawings

FIG. 1 is a plot of LoRa signal frequency modulation used in the present invention;

FIG. 2 is a LoRa signal modulation plot used in the present invention;

FIG. 3 is a block diagram of an implementation of the present invention;

FIG. 4 is a flow chart of the implementation of the preamble signal and the payload signal during demodulation in the present invention;

FIG. 5 is a schematic diagram of the preamble signal interception and data reset during demodulation according to the present invention;

fig. 6 is a schematic diagram of the conventional demodulation method for demodulating the LoRa signal;

fig. 7 is a comparison graph of bit error performance for a fixed two-path channel at SF 7, using the four methods of the present invention, the conventional direct demodulation method, and under the condition of using matched filtering, without interception and zero-setting of the preamble signal;

fig. 8 is a comparison graph of bit error performance for a fixed two-path channel at SF 8, using the four methods of the present invention, the conventional direct demodulation method, and the non-clipping and non-zero-setting of the preamble signal in the case of matched filtering.

Detailed Description

The objects, aspects and advantages of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, all other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without any inventive work belong to the protection scope of the present invention.

The invention relates to a demodulation method of a LoRa signal, wherein the LoRa signal uses an up-chirp or down-chirp frequency sweep signal to encode signal information.

Referring to fig. 1, the frequency modulation of the LoRa signal includes a frequency curve 1 of an unmodulated LoRa signal, i.e., an up-chirp signal, and a frequency curve 2 of a modulated LoRa signal.

Referring to fig. 2, the LoRa signal modulation includes an unmodulated LoRa signal diagram a and a modulated LoRa signal diagram b.

At present, most of the demodulation methods related to the LoRa modulation signal are traditional direct demodulation methods, and the method can obtain better performance under a single-path channel, but under a multi-path channel, the bit error performance is obviously reduced, and the reliability of the system is deteriorated. The invention solves the problem of demodulation of the LoRa modulation signal under a multipath channel.

Referring to fig. 3, the implementation steps of the present invention are as follows:

step 1, acquiring a preamble signal and a load signal.

Respectively carrying out down-sampling on the received signal and the unmodulated LoRa signal, namely the original up-chirp and the original down-chirp;

and dividing the down-sampled received signal according to the symbol numbers of the preamble signal and the load signal known by the receiving end to obtain a preamble signal F and a load signal P1.

And 2, acquiring a new preamble signal B and a new load signal D.

Referring to fig. 4, the specific implementation of this step is as follows:

(2a) acquiring a new preamble signal B:

(2a1) performing dot multiplication on the preamble signal F and the unmodulated LoRa signal after down sampling, namely the original down-chirp, and then sequentially performing FFT and solving an absolute value to obtain the length of 2^SFThe preamble signal a of (a);

(2a2) for length of 2^SFThe preamble signal a of (a) is intercepted:

referring to fig. 5, the implementation of this step is: storing M points around the maximum value to obtain a signal C with the length of (2M +1) points, finding out the maximum value from the directly demodulated preamble signal A, determining a subscript MAX of a position corresponding to the maximum value, and comparing the size of MAX with M:

when MAX is less than or equal to M, subscript of preamble signal A is 2^SF- (MAX-M) to subscript 2^SFThe section of signal is taken as the first half section of the intercepted signal, the section of signal with the subscript of the leading signal A ranging from 1 to the subscript MAX + M is taken as the second half section of the intercepted signal, and the front section and the rear section form a leading signal C;

when MAX is more than or equal to 2^SFwhen-M, index of preamble signal A is MAX-M to index is 2^SFThe signal of the section is taken as the first half section of the intercepted signal, and the subscript of the leading signal A is 1 to the subscript MAX + M-2^SFThe section of signal is used as the second half section of the intercepted signal, and the front section and the rear section form a preamble signal C;

when the MAX does not meet the two conditions, the subscript of the preamble signal A is MAX-M to MAX + M, and the preamble signal C is formed by the signals;

(2a3) resetting the intercepted preamble signal C:

the value of each point in the intercepted preamble signal C is compared with the maximum value

Making a comparison of magnitude, when the value at a certain point is less than the maximum value

When the point data is reset to 0, the value of a certain point is greater than or equal to the maximum value

Keeping the value of the point unchanged to obtain a new preamble signal B, wherein the value of P is taken according to the condition of the channel,

(2b) acquiring a new load signal D:

and performing corresponding point multiplication on each symbol in the load signal P1 subjected to down sampling and an unmodulated LoRa signal, namely the original down-chirp signal, and performing FFT (fast Fourier transform) and absolute value extraction on each symbol in sequence to obtain a load signal D subjected to direct demodulation.

And 3, acquiring a new load signal E.

And performing matched filtering on each symbol in the new preamble signal B and the load signal D, performing point-by-point sliding point multiplication on the new preamble signal B and the load signal D under the condition of keeping each symbol in the load signal D fixed, when the new preamble signal B is overlapped with the last point of each symbol in the load signal D, sequentially moving each symbol in the load signal D from the starting point and placing the symbol behind the last point until the starting points of the new preamble signal B and the load signal D are overlapped again, and terminating the operation, wherein the obtained result is the new load signal E.

And 4, acquiring modulation information bits.

(5a) Finding the subscript MAX of the position corresponding to the maximum value of each symbol in the new load signal E, calculating the value of MAX + M-MAX, and comparing the calculation result with 2^SFPerforming remainder operation to eliminate the influence of the initial position MAX-M and obtain the actual position P of the maximum value of each symbol in the new load signal E;

(5b) dividing P obtained in the step (5a) by 2, taking the rest numbers as the coefficient d0 of the lowest bit of the binary gray code, then continuously dividing the quotient by 2, taking the remainder as the coefficient d1 of the lowest bit of the binary gray code, and repeating the steps until the quotient is 0 to obtain the binary gray code b;

(5c) reserving the highest bit of the binary gray code b as the highest bit of the natural binary code, and taking the result of the exclusive OR of the highest bit of the natural binary code and the second highest bit of the gray code as the second highest bit of the natural binary code;

(5d) and (5) sequentially solving the rest bits of the natural binary code according to the same method as the method (5c) to obtain modulation information bits.

The technical effects of the invention are further explained in detail by combining simulation experiments as follows:

in the case of the experiment 1,

1.1) experimental conditions: let SF equal to 7, i.e. symbol length 128 points, sampling frequency Fs equal to 10MHz, up-sampling multiple 8, signal bandwidth B equal to Fs/8, Byte size of input modulation information equal to 64, channel as two fixed paths, and delay between two pathsThe delay time is 0.6us, 16 points around the maximum value of the preamble signal A are reserved, and the maximum value is taken

1.2) contents of the experiment

Under the condition of 1.1) above, the LoRa signal is demodulated by the method of the present invention and the conventional demodulation method, and the result is shown in fig. 7:

as can be seen from FIG. 7, the method of the present invention has a bit error rate of 10 in comparison with the conventional demodulation method^-3The bit error performance is improved by 1.5dB, the bit error performance of the 33-point preamble signal without 0 is improved compared with the 128-point preamble signal, and the bit error performance of the 33-point preamble signal with 0 is also improved compared with the 33-point preamble signal without 0. From the viewpoint of reliability, the lower the bit error rate is for the same SNR, the better the reliability of the system is.

It can also be seen from the 5 curves in fig. 7 that the performance of the method of the present invention is close to that of the theoretical case of additive white gaussian noise AWGN, i.e. there is only a 0.5dB difference.

In the case of the experiment 2, the reaction was,

2.1) experimental conditions: let SF be 256 points, i.e. symbol length, sampling frequency Fs be 10MHz, up-sampling multiple be 8, signal bandwidth be Fs/8, Byte size of input modulation information be 64, channel be two fixed paths, delay time between two paths be 0.6us, retain 16 points around maximum value of preamble signal a, take

2.2) contents of the experiment

Under the condition of 2.1) above, the LoRa signal is demodulated by the method of the present invention and the conventional demodulation method, and the result is shown in fig. 8:

as can be seen from FIG. 8, the method of the present invention has a bit error rate of 10 in comparison with the conventional demodulation method^-3The bit error performance is improved by 1.5dB, and compared with the lead signal with 256 points, the lead signal with 33 points not set with 0 is improved, and 33 points are longThe bit error performance of the preamble signal with 0 set is also improved compared with the preamble signal with 33 points with 0 set.

From the viewpoint of reliability, the lower the bit error rate is for the same SNR, the better the reliability of the system is.

In summary, compared with the conventional method for directly demodulating the LoRa modulated signal under the multipath channel, the method for demodulating the LoRa modulated signal under the multipath channel of the present invention has the same SF and the same fixed two paths, and the error rate is 10^-3The error bit performance is improved by 1.5dB, and compared with the method of not matching, the performance of the preamble after 0 is simultaneously arranged, the preamble without 0 is not intercepted, and the performance is the best, and the performance difference with the performance under the additive white Gaussian noise AWGN under the theoretical condition is only 0.5dB, which shows that the invention has better demodulation performance under the multipath channel and higher system reliability.

Claims

1. A LoRa modulation signal demodulation method under a multipath channel is characterized by comprising the following steps:

(2b) The value of each point of the preamble signal C is compared with the maximum value

And (3) comparison: when the value at a certain point is greater than or equal to the maximum value

Then, the value of the point is reset to 0 to obtain a new preamble signal B, wherein the value of Z is taken according to the condition of the channel,

(5) and searching a position P corresponding to the maximum value of each symbol in the new load signal E, and sequentially carrying out decimal to binary conversion and Gray code inverse mapping on the position P to obtain modulation information bits.

2. The method of claim 1, wherein step (2a) is performed for a length of 2^SFIntercepting the leading signal A, storing M points around the maximum value to obtain a signal C with the length of (2M +1) points, finding out the maximum value from the leading signal A after direct demodulation, determining a subscript MAX of a position corresponding to the maximum value, and comparing the size of the MAX with that of the M:

when MAX is more than or equal to 2^SFwhen-M, index of preamble signal A is MAX-M to index is 2^SFThe signal of the section is taken as the first half section of the intercepted signal, and the subscript of the leading signal A is 1 to the subscript MAX + M-2^SFThis segment of the signalAs the second half section of the intercepted signal, the front section and the rear section form a preamble signal C;

when MAX is not more than MAX and is not less than MAX and not less than 2^SFIn both cases-M, the preamble signal A is indexed MAX-M to MAX + M to form the preamble signal C.

3. The method of claim 1, wherein the step (4) of performing matched filtering on each symbol in the new preamble signal B and the payload signal D is to perform point-by-point sliding dot multiplication on the new preamble signal B while keeping each symbol in the payload signal D fixed, and when the new preamble signal B coincides with the last point of each symbol in the payload signal D, the symbols in the payload signal D are shifted from the starting point to the end of the last point one by one until the starting points of the new preamble signal B and the payload signal D coincide again, and the operation is terminated.

4. The method of claim 1, wherein the step (5) of maximum value detection is performed on the new load signal E after matching, and gray code inverse mapping is performed on the finally generated binary system, and the steps are implemented as follows:

(5b) dividing P obtained in the step (5a) by 2, taking the rest numbers as the coefficient d0 of the lowest bit of the binary gray code, then continuously dividing the quotient by 2, taking the remainder as the coefficient d1 of the lowest bit of the binary gray code, and repeating the steps until the quotient is 0, so as to obtain the binary gray code b;