Polarization code coding and decoding method based on LoRa modulation

Abstract

The invention belongs to the technical field of communication, and particularly relates to a polar code encoding and decoding method based on LoRa modulation. A multi-layer polarization code modulation structure is adopted to combine polarization codes and LoRa modulation, symbol soft information is calculated to serve as output of LoRa soft demodulation to be sent to a decoder, an ORBGRAND decoder which is low in complexity and capable of being achieved in parallel is adopted as a polarization code decoder, the ORBGRAND decoder inquires error modes in sequence according to bit reliability to conduct decoding, and the decoding efficiency is improved. Soft information updating is assisted by returning an error mode, serial decoding is carried out, and the reliability of the system is effectively improved.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and specifically relates to a polar code encoding method based on LoRa modulation. Background Art

[0002] LoRa modulation is a low-power wide-area IoT technology based on Chirp Spread Spectrum (CSS). It has the characteristics of low power consumption and low complexity, strong anti-interference ability, and can achieve long-distance transmission. There are already many research papers on LoRa technology, but most of the coding schemes used are Hamming codes, which are the channel coding schemes specified in the LoRa physical layer protocol. As a coding based on hard decision decoding, the error performance and coding gain of Hamming codes are not superior to those of the currently widely used polar codes and LDPC codes, so the high reliability of the system cannot be guaranteed. In a paper published by Xu Lang et al., "LoRa Improvement Method Based on Turbo Code and ODPD Judgment Method" (Electronic Measurement Technology, 2020, 43(7): 142-147), a soft demodulation method based on incoherent demodulation of LoRa signals is disclosed, namely, the orthogonal binary peak difference decision method. The soft information output by the demodulation is then sent to the Turbo decoder for iterative decoding. The simulation results show a certain performance improvement compared to the Hamming cascade LoRa system. The shortcomings of this method are: first, the iterative decoding is complex and involves a large number of operations, which increases the complexity and power consumption of the receiver and requires high chip processing capabilities. Second, the decoding delay is too large, which is not suitable for applications with high real-time requirements.

[0003] Polar codes, as the first coding scheme proven to reach channel capacity, have received widespread attention for their simple encoding and decoding algorithms and excellent decoding performance. Compared with Turbo codes designed for long data packets, polar codes have better error correction performance in the case of short data packet transmission, and can achieve fine code rate adjustment, which is suitable for a variety of scenarios and transmission requirements. SEIDL M et al. proposed multi-layer polarization coded modulation in their paper "Multilevel polar-coded modulation" (2013IEEE International Symposium on Information Theory, 2013: 1302-1306), combining polar codes with high-order modulation. This method uses multiple polar code component encoders according to the dimension of high-order modulation, extracts bits from the corresponding positions of the component codes to form vectors, maps them into a modulation symbol according to the modulation mode and sends them to the channel for transmission. At the receiving end, a multi-level demodulation method is used for demodulation and decoding, which has high spectrum efficiency and performance gain. The shortcomings of this method are: First, the channel quality has a great influence on the performance of high-order modulation. When the channel is poor, the bit error rate performance is very poor and even communication interruption may occur. Second, the Euclidean distance of the vector of general high-order modulation is small, which is prone to errors when the interference is large, and has poor anti-interference performance, making it unsuitable for communication scenarios with complex interference. Summary of the invention

[0004] In view of the shortcomings of the above method, the present invention proposes a method for combining LoRa modulation and polarization code for encoding and decoding, which can achieve efficient channel error correction encoding and decoding, and improve the reliability, anti-interference performance and frequency band efficiency of the system.

[0005] The technical solution adopted by the present invention is:

[0006] A polar code encoding method based on LoRa modulation comprises the following steps:

[0007] Step 1: The transmitting end divides the information sequence to be sent into multiple component information sequences of equal length, and the component information sequences do not overlap each other;

[0008] Step 2: Add CRC check bits to each component information sequence and perform polarization code encoding to obtain multiple component codes and sort them in sequence;

[0009] Step 3, sequentially take out 1 bit from the same position of the component code to form a bit vector, and perform Gray encoding on the decimal symbol corresponding to the bit vector;

[0010] Step 4: Perform LoRa modulation on the Gray-coded symbols, and send the obtained LoRa modulated symbols into the channel for transmission;

[0011] Step 5: The receiving end obtains a received signal through the Gaussian channel, and multiplies the received signal with the synchronized original down-chirp signal to perform de-chirping;

[0012] Step 6: Single sampling the dechirped signal and obtain the amplitude through discrete Fourier transform;

[0013] Step 7: Represent all possible symbols in binary form, divide them into sets according to the bits at each position of the binary system, and represent the set symbol K according to whether the i-th bit is 0 or 1. i,0 , K i,1 ;

[0014] Step 8: Find the maximum value of the amplitude of each symbol in the set, then the soft information of the i-th bit of the demodulated symbol is S rx (k)| is the amplitude corresponding to index k;

[0015] Step 9: Calculate symbol soft information to determine symbol reliability. Symbol soft information is the average value of bit soft information within the symbol. N SF is the spreading factor;

[0016] Step 10: Input the symbol soft information as the soft information of the first component code into the ORBGRAND decoder, sort the reliability, gradually find the error mode by increasing the logical weight, and output the decoded codeword and the found error mode after checking with the check matrix; adjust the soft information according to the error mode output by the decoder, so that the soft information S of the error bit is i =min(S i ), and input the updated soft information into the component decoder of the next component code, and so on to achieve serial decoding.

[0017] In the scheme of the present invention, the polar code is combined with LoRa modulation by using a multi-layer polarization coded modulation structure, and the performance of the combination of polar code and LoRa modulation is achieved without destroying the polarization structure. Since LoRa modulation always makes errors in the form of symbols and there are many error bits in the symbols without obvious correlation, the symbol soft information output by the LoRa soft decision can be used as the reliability information of the corresponding bits of the component code and input into the decoder. The decoder adopts the ORBGRAND decoder with low decoding complexity and excellent performance. The ORBGRAND decoder is suitable for short codes, and the parallel structure is suitable for hardware implementation. The method of decoding by sequentially querying the error mode according to the bit reliability is also suitable for LoRa modulation with errors in the form of symbols.

[0018] The scheme of the present invention adopts serial decoding during decoding, and the error mode output after decoding of each layer of component code is used to assist the soft information update of the next layer of decoding, thereby improving the success rate of subsequent decoding and improving the error performance of the system.

[0019] The beneficial effects of the present invention are as follows: the present invention combines multi-layer polarization coded modulation with LoRa modulation, and the bit error rate of the LoRa modulation system without polarization code encoding and decoding is significantly improved; the symbol soft information calculation method proposed in the present invention and auxiliary serial decoding effectively improve the bit error performance; the adopted ORBGRAND serial decoding is simple to operate, has excellent performance under short code and high code rate, and its complexity is significantly reduced compared with the iterative decoding algorithm of Turbo code. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flowchart of the overall process of the sending end and the receiving end of the present invention.

[0021] Figure 2 This is a structural block diagram of the encoding and decoding algorithm used in the present invention.

[0022] Figure 3 This is a performance simulation diagram of LoRa modulation without coding and the polar code combined with LoRa modulation proposed in the present invention. DETAILED DESCRIPTION

[0023] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings:

[0024] Figure 1 This is the overall flow chart of the transmitting end and the receiving end of the present invention. The component information sequences after the information sequence is evenly divided into serial and parallel conversion are respectively encoded with polar codes, and the bit vectors of the bits at the same position in the component code are Gray mapped and LoRa modulated. The modulated symbols are obtained by the receiver through the Gaussian channel, soft demodulated and Gray decoded, and the obtained symbol soft information is input into the ORBGRAND decoder of the component code, and the soft information is continuously updated for serial decoding. The specific implementation steps of the present invention are as follows:

[0025] Step 1: Figure 2 As shown, the MLC-PCM architecture is used at the transmitting end to evenly divide the information sequence of length K into N SF (Spreading factor) Component information sequence with segment length k.

[0026] Step 2: Perform crc coding and polar coding on the component information sequences to obtain N SF There are component codes of length N, and the construction method of each component code is the same, and can be constructed according to the density evolution (DE) and Gaussian approximation (GA) methods.

[0027] Step 3: From NSF The same position of the coded codewords is taken out one bit at a time to form an N SF The corresponding decimal symbol is Gray-coded to make the difference between adjacent symbols only 1 bit. In this way, if the symbol demodulation is offset by one symbol due to inaccurate synchronization during reception, only a 1-bit error will occur.

[0028] Step 4: Modulate the Gray-coded symbols using LoRa

[0029]

[0030] Where: K is the cyclic shift value, that is, the number of offset chips, which is the decimal symbol encoded by Gray. A symbol contains chips, then there is B is the bandwidth; f c is the carrier frequency; T s is the signal frequency sweep time, i.e., symbol period. The obtained LoRa modulation symbol is sent to the channel for transmission.

[0031] Step 5: At the receiving end of the system, obtain the received signal r(t) through the Gaussian channel

[0032] r(t)=s m (t)+n(t)

[0033] Where n(t) has a mean of 0 and a variance of σ 2 Gaussian white noise.

[0034] The received signal is dechirped by multiplying it with the synchronized original downchirp signal.

[0035]

[0036] Where: is the original downchirp signal, v(t)=n(t)s * (t) is also a Gaussian white noise signal.

[0037] Step 6: After dechirp, the obtained signal is single-sampled and further DFT transformed. The receiving end adopts single-sample mode for receiving sampling, that is, the sampling rate is BHz, then the discrete signal after sampling is

[0038]

[0039] After DFT, the spectrum is obtained

[0040]

[0041] Where V(k) is the noise Then the amplitude signal |S is obtained by taking the modulus. rx (k)|.

[0042] Step 7: Convert 0 to All symbols are represented by The set is divided according to the bits at each position of the binary system, and the set symbol is K according to whether the i-th bit is 0 or 1. i,0 ={k|x i =0}, K i,1 ={k|x i =1}.

[0043] Step 8: Find the set K separately i,0 and K i,1 The maximum value of the amplitude of the symbol in , then the soft information of the i-th bit of the demodulated symbol is

[0044]

[0045] Step 9: Calculate symbol soft information to determine symbol reliability. Since LoRa modulation always makes errors in the form of symbols and there are many error bits in the symbol without obvious correlation, the symbol reliability can be determined to find the position of the error-prone bits in the component code. Symbol soft information is the average value of the bit soft information in the symbol.

[0046]

[0047] Where N SF is the spreading factor.

[0048] Step 10: The decoding process is as follows: Figure 2 As shown in the figure, the symbol soft information is input into the ORBGRAND decoder as the reliability information of the first component code for sorting, and the error pattern is found by gradually increasing the logical weight. The decoded sequence and the error pattern are output after verification by the check matrix. The soft information is adjusted according to the error pattern output by the decoder, so that the soft information S of the error bit i =min(S i ), so that the error mode in which it is located can be quickly found in the subsequent decoding, and the updated soft information is sequentially input into the subsequent component decoders to realize serial decoding and improve the decoding accuracy.

[0049] The effect of the present invention is further illustrated by simulation experiments below.

[0050] The simulation experiment uses Matlab simulation software. The polarization component code length is N = 128, the code rate is R = 4 / 5, and a crc check bit with a length of 6 is added. The simulation parameters of LoRa are: bandwidth BW = 125000 Hz, sampling rate Fs = 125000 Hz, and center frequency fc = 0 Hz.

[0051] The simulation experiment of the present invention simulates the performance of LoRa modulation without coding and the polar code combined with LoRa modulation proposed by the present invention when the spreading factor is 8, 9, and 10, and obtains 6 bit error rate performance curves as shown in FIG. Figure 3 shown. Figure 3 The horizontal axis represents the signal-to-noise ratio, and the vertical axis represents the bit error rate. Figure 3 The solid line curve marked with diamonds in the figure represents the bit error rate of LoRa hard demodulation when the spreading factor is 8; Figure 3 The dotted curve marked with diamonds in the figure represents the bit error rate of polar code combined with LoRa modulation when the spreading factor is 8. Figure 3 The solid line curve marked with a circle in the figure represents the bit error rate of LoRa hard demodulation when the spreading factor is 9; Figure 3 The dotted curve marked with a circle in the figure represents the bit error rate of polar code combined with LoRa modulation when the spreading factor is 9. Figure 3 The solid line curve marked with a cross in the figure represents the bit error rate of LoRa hard demodulation when the spreading factor is 10; Figure 3 The dotted curve marked with a cross in the figure represents the bit error rate of polar code combined with LoRa modulation when the spreading factor is 10.

[0052] Depend on Figure 3 It can be seen that the polar code encoding method based on LoRa modulation proposed in the present invention has a bit error rate of 10 -3 There is a coding gain of nearly 2dB, which significantly improves the bit error performance of LoRa modulation without coding, effectively improving the reliability of the system.

Claims

1. A polar code encoding method based on LoRa modulation, characterized in that: The steps include: Step 1: The transmitting end divides the information sequence to be sent into multiple component information sequences of equal length, and the component information sequences do not overlap each other; Step 2: Add CRC check bits to each component information sequence and perform polarization code encoding to obtain multiple component codes and sort them in sequence; Step 3, sequentially take out 1 bit from the same position of the component code to form a bit vector, and perform Gray encoding on the decimal symbol corresponding to the bit vector; Step 4: Perform LoRa modulation on the Gray-coded symbols, and send the obtained LoRa modulated symbols into the channel for transmission; Step 5: The receiving end obtains a received signal through the Gaussian channel, and multiplies the received signal with the synchronized original down-chirp signal to perform de-chirping; Step 6: Single sampling the dechirped signal and obtain the amplitude through discrete Fourier transform; Step 7: Represent all possible symbols in binary form, divide them into sets according to the bits at each position of the binary system, and represent the set symbol K according to whether the i-th bit is 0 or 1. i,0 , K i,1 ; Step 8: Find the maximum value of the amplitude of each symbol in the set, then the soft information of the i-th bit of the demodulated symbol is S rx (k)| is the amplitude corresponding to index k; Step 9: Calculate symbol soft information to determine symbol reliability. Symbol soft information is the average value of bit soft information within the symbol. N SF is the spreading factor; Step 10: Input the symbol soft information as the soft information of the first component code into the ORBGRAND decoder, sort the reliability, gradually find the error mode by increasing the logical weight, and output the decoded codeword and the found error mode after checking with the check matrix; adjust the soft information according to the error mode output by the decoder, so that the soft information S of the error bit is i =min(S i ), and input the updated soft information into the component decoder of the next component code, and so on to achieve serial decoding.

Images