Soft demodulation algorithm based on bit interleaving Turbo coding frequency shift chirp keying modulation

Abstract

The invention provides a soft demodulation algorithm based on bit interleaving Turbo coding frequency shift chirp keying modulation, and relates to the technical field of communication. The method comprises the following steps: carrying out Turbo coding on a signal at a sending end, interleaving coded bits, carrying out LoRa modulation mapping on the interleaved coded bits to obtain a corresponding periodic frequency shift chirp signal, obtaining a frequency domain waveform of a despread signal at a receiving end by utilizing an original LoRa demodulation scheme, and obtaining a frequency shift chirp signal according to probability distribution of values of frequency domain sampling points. And calculating a log-likelihood ratio on each bit, sending the obtained log-likelihood ratios to a corresponding de-interleaver, and inputting and transmitting the de-interleaved log-likelihood ratios as soft information to a Turbo decoder for Turbo decoding. The problem of information loss during information transmission between the LoRa demodulator and the Turbo soft input soft output decoder is solved. After Turbo coding is added, the bit error rate performance of the system is obviously improved, and along with the increase of the signal-to-noise ratio, the decline rate of the bit error rate is obviously improved.

Description

Technical Field

[0001] This document relates to the field of communication technology, and in particular to a soft demodulation algorithm based on bit-interleaved Turbo coded frequency-shift chirp keying modulation. Background Art

[0002] Bit-interleaved coded modulation (BICM) is an error control technology that combines channel coding and modulation as a whole to obtain diversity gain and coding gain, thereby improving the spectrum power and power spectrum efficiency of wireless channel transmission. BICM uses a combination of coding and modulation technology to obtain coding gain without sacrificing information rate bandwidth.

[0003] Turbo coding is a channel coding scheme that has been used for more than a decade and is close to the capacity of AWGN (Additive White Gaussian Noise) channels. It has performance that is no less than that of LDPC codes and polar codes, but Turbo codes are more flexible in construction. Turbo codes use an interleaving method to construct random long codes and can use a maximum likelihood ratio iterative decoding algorithm, which can enable the system to obtain good error performance at a low signal-to-noise ratio. It has important practical value for high-order modulation systems. However, its soft-input and soft-output decoder requires accurate likelihood information input, otherwise its performance will deteriorate sharply, which requires obtaining the maximum likelihood ratio of the demodulator output information bit.

[0004] As a low-power wide-area IoT technology, LoRa has been successfully applied to a large number of IoT scenarios, with the advantages of scalable bandwidth, strong anti-interference ability, and high receiving sensitivity. LoRa uses chirp spread spectrum modulation (CSS) signal to modulate data. LoRa modulation makes the receiver have high sensitivity and can accurately recover information under low signal-to-noise ratio conditions.

[0005] One of the fundamental reasons why Turbo codes achieve excellent performance is the use of an iterative decoding algorithm, which improves decoding performance by exchanging soft information between component decoders corresponding to component encoders. For Turbo codes, if the output of the component decoder is a hard decision, it is impossible to achieve the exchange of soft information between component decoders, which limits the further improvement of system performance.

[0006] Therefore, there is an urgent need for a coding method to solve the problem of information loss during information transmission between the LoRa demodulator and the Turbo soft-input soft-output decoder. Summary of the invention

[0007] This specification provides a soft demodulation algorithm based on bit-interleaved Turbo coded frequency-shift chirp keying modulation, which is used to solve the problem of information loss when information is transmitted between the LoRa demodulator and the Turbo soft-input soft-output decoder.

[0008] In a first aspect, this specification provides a soft demodulation algorithm based on bit-interleaved Turbo coded frequency-shift chirp keying modulation, including:

[0009] At the transmitting end, the generated bit information is input into the Turbo encoder for encoding, and the encoded bit information is output;

[0010] The output codeword of the Turbo encoder is sent to the interleaver for interleaving to obtain the interleaved coded bits; the interleaved coded bits are modulated by LoRa and mapped into the corresponding periodic frequency shift chirp signal, and all bits at the transmitting end are coded and modulated into LoRa modulation signals;

[0011] The receiving end demodulates the LoRa signal. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal. The log-likelihood ratio calculation method of the LoRa demodulator is used to obtain the bit log-likelihood ratio.

[0012] The obtained log-likelihood ratio is sent to the Turbo decoder for iterative decoding after passing through the corresponding deinterleaver.

[0013] In the second aspect, this specification provides a soft demodulation device based on bit interleaved Turbo coded frequency shift chirp keying modulation, including: a transmitting end Turbo coding module, a LoRa modulation module, a receiving end despreading and demodulation module, and a deinterleaving decoding module, wherein:

[0014] The transmitting end Turbo encoding module is used to input the generated bit information into the Turbo encoder for encoding at the transmitting end, and output the encoded bit information;

[0015] The LoRa modulation module is used to send the output codeword of the Turbo encoder to the interleaver for interleaving to obtain the interleaved coded bits; perform LoRa modulation mapping on the interleaved coded bits to the corresponding periodic frequency shift chirp signal, and encode and modulate all the bits of the transmitting end into LoRa modulation signals;

[0016] The receiving end despreading and demodulation module is used for demodulating the LoRa signal at the receiving end. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal, and obtains the bit logarithmic likelihood ratio using the LoRa demodulator logarithmic likelihood ratio calculation method;

[0017] The deinterleaving decoding module is used to send the obtained log-likelihood ratio through a corresponding deinterleaver and then send it to a Turbo decoder for iterative deinterleaving.

[0018] In a third aspect, this specification also provides a network device, including: a communication interface, a processor, and a memory;

[0019] The processor calls the program instructions in the memory to perform the following actions:

[0020] At the transmitting end, the generated bit information is input into the Turbo encoder for encoding, and the encoded bit information is output;

[0021] The output codeword of the Turbo encoder is sent to the interleaver for interleaving to obtain the interleaved coded bits; the interleaved coded bits are modulated by LoRa and mapped into the corresponding periodic frequency shift chirp signal, and all bits at the transmitting end are coded and modulated into LoRa modulation signals;

[0022] The receiving end demodulates the LoRa signal. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal. The log-likelihood ratio calculation method of the LoRa demodulator is used to obtain the bit log-likelihood ratio.

[0023] The obtained log-likelihood ratio is sent to the Turbo decoder for iterative decoding after passing through the corresponding deinterleaver.

[0024] The beneficial effects of the present invention are as follows:

[0025] This specification provides a bit-interleaved Turbo coded frequency-shift chirp keying modulation algorithm based on soft-input soft-output decoding. The method performs Turbo coding on the signal at the transmitting end, interleaves the coded bits, performs LoRa modulation on the interleaved coded bits and maps them into corresponding periodic frequency-shift chirp signals. The original LoRa demodulation scheme is used at the receiving end to obtain the frequency domain waveform of the demodulated signal, and the log-likelihood ratio on each bit is calculated according to the probability distribution of the values ​​of each frequency domain sampling point. The obtained log-likelihood ratio is sent to the corresponding deinterleaver, and the deinterleaved log-likelihood ratio is passed as soft information input to the Turbo decoder, and Turbo decoding is performed through the Max-Log-MAP iterative decoding algorithm. The problem of information loss during information transmission between the LoRa demodulator and the Turbo soft-input soft-output decoder is solved. Compared with the LoRa signal frequency shift modulation system without Turbo coding, the bit error rate performance of the system has been significantly improved after adding Turbo coding. As the signal-to-noise ratio increases, the rate of decrease of the bit error rate has been significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are used to provide a further understanding of this specification and constitute a part of this specification. The illustrative embodiments and descriptions of this specification are used to explain this specification and do not constitute an improper limitation on this specification. In the drawings:

[0027] Figure 1 It is a schematic diagram of a soft demodulation algorithm based on bit-interleaved Turbo coded frequency-shift chirp keying modulation provided in an embodiment of this specification;

[0028] Figure 2 is a schematic diagram of a transmitting end and a receiving end structure provided in an embodiment of this specification;

[0029] Figure 3 is a schematic diagram of the structure of a Turbo encoder provided in an embodiment of this specification;

[0030] Figure 4 is a schematic diagram of the structure of a component code encoder provided in an embodiment of this specification;

[0031] Figure 5 It is a schematic diagram of a structural frame of a Turbo decoder provided in an embodiment of this specification.

[0032] Figure 6 It is a schematic diagram of a bit error rate performance comparison provided in an embodiment of this specification.

[0033] Figure 7 It is a schematic diagram of a soft demodulation device based on bit-interleaved Turbo coded frequency-shift chirp keying modulation provided in an embodiment of this specification;

[0034] Figure 8 It is a schematic diagram of a network device structure provided in an embodiment of this specification. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solutions and advantages of this specification clearer, the technical solutions of this application will be clearly and completely described below in conjunction with the specific embodiments of this application and their corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this document.

[0036] The technical solutions provided by the embodiments of this specification are described in detail below in conjunction with the accompanying drawings. Specific embodiment one:

[0038] This embodiment is a soft demodulation algorithm based on bit-interleaved Turbo coded frequency-shift chirp keying modulation, which is used for information transmission between the LoRa demodulator and the Turbo soft-input soft-output decoder. Figure 1 , Figure 2 ;

[0039] First of all, it should be noted that the component encoder in the Turbo encoder in this embodiment adopts the Turbo component encoder of the 3GPP LTE standard. The component encoder consists of 3 shift registers, the coding constraint length is 4, has 8 states, and the coding polynomial (octal form) is g(D)=[13,15], where the feedback polynomial is 13 and the forward polynomial is 15. The quadratic permutation polynomial (QPP) interleaver adopted by the LTE standard has the general formula of the interleaving function as follows:

[0040] Π(i)=(f2i 2 +f1i)modK,(i=0,...,K-1);

[0041] The method of this embodiment specifically includes the following steps:

[0042] Step 102: At the transmitting end, the generated bit information is input into a Turbo encoder for encoding, and the encoded bit information is output;

[0043] Specifically, the structural block diagram of the Turbo code encoder is as follows: Figure 3 As shown in Figure 1, it mainly consists of two component encoders, an interleaver, a redundancy detector and a parallel-to-serial conversion unit. The two component encoders are cascaded in parallel and both use the same structure of recursive systematic convolutional code (RSC). The interleaver is located between the two component encoders. One of the RSC encoders processes the original input information bit u k Another RSC encoder encodes the information bit u′ after passing through the interleaver. k The check bits are then input into the parallel-to-serial conversion unit together with the original information bits through the redundancy puncturer to form the codeword c k , the RSC encoder structure used is as follows Figure 4 shown.

[0044] Step 104: Send the output codeword of the Turbo encoder to an interleaver for interleaving to obtain interleaved coded bits; perform LoRa modulation mapping on the interleaved coded bits into corresponding periodic frequency shift chirp signals, and encode and modulate all bits at the transmitting end into LoRa modulation signals;

[0045] Among them, LoRa modulation uses an up-chirp signal to transmit information. Its signal frequency changes linearly within one symbol period. Its frequency change range and speed depend on the bandwidth B and the spreading factor SF. A LoRa signal carries SF bits of information. LoRa modulation carries information through different starting frequencies of the baseband signal.

[0046] Specifically, a specific implementation of step 104 may be:

[0047] S41, sending the encoded bit stream to a random interleaver with a length of 2160 bits for interleaving.

[0048] S42, according to the set parameter bandwidth B and spreading factor SF, determine the signal frequency change range and the corresponding frequency change rate, and the sampling interval of the LoRa signal meets T s =1 / B, a LoRa signal carries SF data bits, each SF of encoded data bits are divided into a group, the nth data bit in each group is defined as d[n], the information carried by the LoRa signal and the corresponding power-normalized periodic frequency shift signal are determined, and the modulated signal is sent to the receiving end.

[0049] Specific example:

[0050] The interleaved coded bits are mapped to the corresponding periodic frequency shift chirp signal to realize LoRa modulation. According to the set parameter bandwidth B and spreading factor SF, the signal frequency change range and the corresponding frequency change rate are determined. The sampling interval of the LoRa signal meets T s =1 / B, a LoRa signal can carry SF=6 data bits, and each SF=6 data bits after encoding are divided into a group, and the nth data bit in each group is d[n]. The information carried by the LoRa signal is:

[0051]

[0052] The corresponding power normalized periodic frequency shift signal is:

[0053] Where n=0,1......2 SF -1, send this modulated signal to the receiving end;

[0054] Repeat until all bits at the transmitting end are encoded and modulated into LoRa modulation signals.

[0055] Based on this, the original bits are mapped into bit-interleaved Turbo coded LoRa modulated signals for subsequent transmission processing.

[0056] Step 106: The receiving end demodulates the LoRa signal. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal, and obtains the bit logarithmic likelihood ratio using the LoRa demodulator logarithmic likelihood ratio calculation method;

[0057] Specifically, a specific implementation of step 106 may be:

[0058] S61, at the receiving end, calculate the power spectrum density of the received LoRa signal energy and noise, so as to obtain ζ i =|n i |With The probability density function of and Where n i represents the noise amplitude of the i-th sampling point, P L Indicates signal power;

[0059] S62, multiplying the received signal r(n) by the original down-chirp signal to perform signal despreading to obtain g(n), and performing fast Fourier transform on the despread signal g(n) to obtain a frequency domain waveform f(n) of the despread signal;

[0060] S63. According to the probability distribution of the values ​​of each frequency domain sampling point, the log-likelihood ratio of each bit is calculated using the LoRa demodulator log-likelihood ratio calculation method.

[0061] It should be noted that the frequency domain waveform f(n) is grouped according to the binary number corresponding to each sampling point. When the current bit is 0, the corresponding f(n) frequency sampling point 2 SF-1 points;

[0062] When the current bit is 1, the remaining 2 corresponding to the f(n) frequency sampling point SF-1 points.

[0063] Wherein, the LoRa demodulator log-likelihood ratio calculation method includes:

[0064] According to the values ​​of each frequency sampling point, calculate the probability P(f(n)|u=0) that the current bit is 0:

[0065]

[0066] Where SF represents the spreading factor, Indicates i =|n i |With The probability density function of , u represents the current bit, i, j represent the i-th and j-th sampling points respectively;

[0067] According to the value of each frequency sampling point, calculate the probability P(f(n)|u=1) that the current bit is 1:

[0068]

[0069] Take the logarithm of the ratio of the probability that the current bit is 0 to the probability that the current bit is 1 to get the log-likelihood ratio of the current bit:

[0070] Specific example:

[0071] After synchronization processing at the receiving end, the corresponding TDM-LoRa modulated signal is received: r(n)=s(n)+n i (n);

[0072] Among them, n i (n) means the mean is 0 and the variance is σ 2 =N0 / 2 complex Gaussian noise, N0 is the noise power spectral density of the single sideband.

[0073] At the receiving end, the power spectral density of the received LoRa signal energy and noise is calculated to obtain ζ i =|n i |With The probability density function of and

[0074] The received signal r(n) is combined with the original down-chirp signal Multiply to despread the signal:

[0075]

[0076] g(n)=g s (n)+N(n)s d (n), g s (n) is a sinusoidal signal, and the frequency domain waveform f(n) of g(n) is obtained by performing Fast Fourier Transform (FFT) on g(n).

[0077] Group f(n). Here, taking the first bit as an example, f(n) is divided into two groups. When the first bit is 0, the corresponding f(n) frequency sampling point is 2 SF-1 points, when the first bit is 1, it corresponds to the f(n) frequency sampling point 2 SF-1 When accurately calculating the log-likelihood ratio, first calculate the probability that the first bit is 0, and substitute the value of each frequency sampling point into it, see formula (1);

[0078] Then calculate the probability that the first bit is 1, and substitute the value of each frequency sampling point into it, see formula (2);

[0079] Take the logarithm of the ratio of the probability that the first bit is "0" to the probability that the first bit is "1", that is:

[0080]

[0081] Find the log-likelihood ratio of the first bit.

[0082] Repeat until the log-likelihood ratio soft information of all bits is obtained.

[0083] Based on this, soft demodulation is achieved without substantially losing the information carried by the mapping symbols, and soft information of each bit is obtained, so as to better utilize the advantages of Turbo decoding in the subsequent process.

[0084] Step 108: After passing the obtained log-likelihood ratio through a corresponding deinterleaver, it is sent to a Turbo decoder for iterative decoding.

[0085] Specifically, a specific implementation of step 108 may be:

[0086] S81, sending the obtained log-likelihood ratio to the corresponding deinterleaver, passing the deinterleaved LLR as soft information input to the Turbo decoder, and performing Turbo decoding through the Max-Log-MAP iterative decoding algorithm.

[0087] Specific example:

[0088] The structure of the Turbo decoder used is as follows: Figure 5 As shown in FIG. 1 , its main structure is two soft-input soft-output (SISO) component decoders, a multiplexed interleaver and a deinterleaver, a serial-to-parallel conversion unit and a hard decision unit. The serial-to-parallel conversion unit receives the channel soft information Λ k Input, and restore the soft information Λ corresponding to the three-way bit according to the coded redundancy rule k,s ,Λ k,p1 and Λ k,p2 . It is divided into likelihood information of information bits and check bits, and the prior information of the first iteration is initialized to zero. The SISO component decoder corresponds to the RSC component encoder, which receives the corresponding information bit soft information Λ k,s Sum check bit soft information Λ k,p input, and additionally receives a priori information value Λ from another component decoder k,a, and calculate the branch metric at the same time. According to the zeroing of the component encoder, initialize the forward metric and the backward metric, and then start the recursive calculation of the forward and backward metrics. After the recursive calculation of the forward and backward metrics is completed, calculate the corresponding posterior probability information. Then calculate the posterior information Λ k,apo And extract external information Λ k,e Finally, it is determined whether the iterative decoding is completed. If it is completed, a hard decision is made on the a posteriori information. Otherwise, the external information generated by the component decoder is used as the a priori information of another component decoder after passing through the interleaver or deinterleaver, and the iterative decoding continues.

[0089] Furthermore, this embodiment illustrates the effect of the method of this embodiment through data comparison of an example.

[0090] This example uses 3GPP LTE standard Turbo code encoding and decoding combined with LoRa signal frequency shift modulation algorithm, and the system's bit error rate performance under different Doppler frequency offsets and AWGN channels.

[0091] The example tested 10,000 frames, with 1,440 bits of data per frame.

[0092] like Figure 6 It can be seen from the simulation results that the bit interleaved Turbo coded frequency shift chirp keying modulation based on soft input and soft output decoding in this embodiment has a LoRa modulation spreading factor of SF=6 and a bit error rate of 1×10 -5 When compared with LoRa, the modulation performance is improved by about 3.5dB, and the modulation performance is improved by about 2.5dB compared with LoRa with Turbo coding. It is suitable for scenarios that require high information transmission accuracy, low system power consumption, and long information transmission distance.

[0093] In summary, this embodiment performs Turbo encoding on the signal at the transmitting end, interleaves the encoded bits, performs LoRa modulation on the interleaved encoded bits and maps them into corresponding periodic frequency-shifted chirp signals, uses the original LoRa demodulation scheme at the receiving end, obtains the frequency domain waveform of the demodulated signal, calculates the log-likelihood ratio on each bit position according to the probability distribution of the values ​​of each frequency domain sampling point, sends the obtained log-likelihood ratio to the corresponding deinterleaver, passes the deinterleaved log-likelihood ratio as soft information input to the Turbo decoder, and performs Turbo decoding through the Max-Log-MAP iterative decoding algorithm. The problem of information loss during information transmission between the LoRa demodulator and the Turbo soft input and soft output decoder is solved. Compared with the LoRa signal frequency shift modulation system without Turbo coding, after adding Turbo coding, the bit error rate performance of the system has been significantly improved, and the bit error rate decrease rate has been significantly improved as the signal-to-noise ratio increases. Specific embodiment 2:

[0095] like Figure 1 , 2 As shown, a soft demodulation algorithm based on bit-interleaved Turbo coded frequency-shift chirp keying modulation in this specific implementation mode specifically includes:

[0096] This embodiment provides a bit interleaved Turbo coded LoRa frequency shift modulation scheme, and proposes a method for calculating the log likelihood ratio (LLR) of the received bit, which is used for information transmission between the LoRa demodulator and the Turbo soft input soft output decoder. The core idea is:

[0097] At the transmitting end, the signal is Turbo encoded, the encoded bits are interleaved, and the interleaved encoded bits are LoRa modulated and mapped into the corresponding periodic frequency shift chirp signal. At the receiving end, the original LoRa demodulation scheme is used to obtain the frequency domain waveform of the despread signal. According to the probability distribution of the values ​​of each frequency domain sampling point, the LLR on each bit is calculated, and the obtained LLR is sent to the corresponding deinterleaver. The deinterleaved LLR is passed to the Turbo decoder as soft information input, and Turbo decoding is performed through the Max-Log-MAP iterative decoding algorithm.

[0098] It should be noted that from the perspective of information theory, any hard decision will lose some information. The soft output decoder requires that the decoder input and output are both soft information. Therefore, combining LoRa modulation with BICM and Turbo coding requires demodulating the LoRa signal to obtain the corresponding bit log-likelihood ratio (LLR).

[0099] The method provided in this embodiment specifically includes the following steps:

[0100] Assume that the component encoder in the Turbo encoder adopts the Turbo component encoder of the 3GPP LTE standard. The component encoder consists of 3 shift registers, the coding constraint length is 4, has 8 states, and the coding polynomial (octal form) is g(D) = [13, 15], where the feedback polynomial is 13 and the forward polynomial is 15. The quadratic permutation polynomial (QPP) interleaver adopted by the LTE standard has the general formula of the interleaving function as follows:

[0101] Π(i)=(f2i 2 +f1i)modK,(i=0,......,K-1)

[0102] Take the Turbo code with a code length of 1440 bits, a code rate of 2 / 3, 8 iterations, and QPP interleaver parameters of K=1440, f1=149, and f2=60 as an example. Suppose the LoRa modulation bandwidth B=125kHz, the spreading factor SF=6, and the symbol period T=2 SF / B, sampling interval Ts =1 / B, modulation order M = 2 SF =64, assuming that BICM uses a random interleaver with a length of 2160 bits, and Turbo decoding uses the Max-Log-MAP iterative decoding algorithm.

[0103] Step 1: Generate the bit information u k Input to the Turbo encoder to complete the encoding, and output the encoded bit information c k The structural block diagram of the Turbo code encoder is as follows: Figure 3 As shown in Figure 1, it mainly consists of two component encoders, an interleaver, a redundancy detector and a parallel-to-serial conversion unit. The two component encoders are cascaded in parallel and both use the same structure of recursive systematic convolutional code (RSC). The interleaver is located between the two component encoders. One of the RSC encoders processes the original input information bit u k Another RSC encoder encodes the information bit u' after passing through the interleaver. k The check bits are then input into the parallel-to-serial conversion unit together with the original information bits through the redundancy puncturer to form the codeword c k , the RSC encoder structure used is as follows Figure 4 shown.

[0104] Step 2: Send the encoded bit stream to a random interleaver with a length of 2160 bits for interleaving.

[0105] Step 3: Map the interleaved coded bits to the corresponding periodic frequency shift chirp signal to realize LoRa modulation. According to the set parameter bandwidth B and spreading factor SF, determine the signal frequency change range and the corresponding frequency change rate. The sampling interval of the LoRa signal satisfies T s =1 / B, a LoRa signal can carry SF=6 data bits, and each SF=6 encoded data bits are divided into a group, and the nth data bit in each group is d[n]. The LoRa signal carries information:

[0106] The corresponding power normalized periodic frequency shift signal is:

[0107] Where n=0,1......2 SF -1, send this modulated signal to the receiving end.

[0108] Step 4: Repeat steps 1 to 3 until all bits at the transmitter are encoded and modulated into LoRa modulation signals.

[0109] Step 5: After synchronization processing at the receiving end, the corresponding TDM-LoRa modulation signal is received: r(n)=s(n)+ni (n);

[0110] Among them, n i (n) means the mean is 0 and the variance is σ 2 =N0 / 2 complex Gaussian noise, N0 is the noise power spectral density of the single sideband.

[0111] Step 6: Calculate the power spectral density of the received LoRa signal energy and noise at the receiving end to obtain ζ i =|n i |With The probability density function P ζi and P γi .

[0112] Step 6: Combine the received signal r(n) with the original down-chirp signal Multiply to despread the signal:

[0113]

[0114] g(n)=g s (n)+N(n)s d (n), g s (n) is a sinusoidal signal, and the frequency domain waveform f(n) of g(n) is obtained by performing Fast Fourier Transform (FFT) on g(n).

[0115] Step 7: Group f(n). Here, taking the first bit as an example, divide f(n) into two groups. When the first bit is "0", the corresponding f(n) frequency sampling point is 2 SF-1 points, when the first bit is "1", it corresponds to the f(n) frequency sampling point 2 SF-1 When accurately calculating the log-likelihood ratio, first calculate the probability that the first bit is "0", and substitute the value of each frequency sampling point into it, see formula (1);

[0116] Then calculate the probability that the first bit is "1" and substitute the value of each frequency sampling point into it, see formula (2);

[0117] Take the logarithm of the ratio of the probability that the first bit is "0" to the probability that the first bit is "1", that is:

[0118]

[0119] Find the log-likelihood ratio of the first bit.

[0120] Step 8: Repeat step 7 until the log-likelihood ratio soft information of all bits is obtained.

[0121] Step 9: Send the received log-likelihood ratio of each bit to the deinterleaver corresponding to the random interleaver used in step 2 to complete deinterleaving.

[0122] Step 9: Input the log-likelihood ratio information output by the demodulator into the Turbo decoder and use the Max-Log-MAP algorithm for Turbo decoding. The structure of the Turbo decoder used is as follows: Figure 5 As shown in FIG. 1 , its main structure is two soft-input soft-output (SISO) component decoders, a multiplexed interleaver and a deinterleaver, a serial-to-parallel conversion unit and a hard decision unit. The serial-to-parallel conversion unit receives the channel soft information Λ k Input, and restore the soft information Λ corresponding to the three-way bit according to the coded redundancy rule k,s ,Λ k,p1 and Λ k,p2 . It is divided into likelihood information of information bits and check bits, and the prior information of the first iteration is initialized to zero. The SISO component decoder corresponds to the RSC component encoder, which receives the corresponding information bit soft information Λ k,s Sum check bit soft information Λ k,p input, and additionally receives a priori information value Λ from another component decoder k,a , and calculate the branch metric at the same time. According to the zeroing of the component encoder, initialize the forward metric and the backward metric, and then start the recursive calculation of the forward and backward metrics. After the recursive calculation of the forward and backward metrics is completed, calculate the corresponding posterior probability information. Then calculate the posterior information Λ k,apo And extract external information Λ k,e Finally, it is determined whether the iterative decoding is completed. If it is completed, a hard decision is made on the a posteriori information. Otherwise, the external information generated by the component decoder is used as the a priori information of another component decoder after passing through the interleaver or deinterleaver, and the iterative decoding continues.

[0123] Furthermore, this embodiment illustrates the effect of the method of this embodiment through data comparison of an example.

[0124] This example uses 3GPP LTE standard Turbo code encoding and decoding combined with LoRa signal frequency shift modulation algorithm, and the system's bit error rate performance under different Doppler frequency offsets and AWGN channels.

[0125] The example tested 10,000 frames, with 1,440 bits of data per frame.

[0126] like Figure 6 It can be seen from the simulation results that the bit interleaved Turbo coded frequency shift chirp keying modulation based on soft input and soft output decoding in this embodiment has a LoRa modulation spreading factor of SF=6 and a bit error rate of 1×10 -5When compared with LoRa, the modulation performance is improved by about 3.5dB, and the modulation performance is improved by about 2.5dB compared with LoRa with Turbo coding. It is suitable for scenarios that require high information transmission accuracy, low system power consumption, and long information transmission distance.

[0127] In summary, this embodiment performs Turbo encoding on the signal at the transmitting end, interleaves the encoded bits, performs LoRa modulation on the interleaved encoded bits and maps them into corresponding periodic frequency-shifted chirp signals, uses the original LoRa demodulation scheme at the receiving end, obtains the frequency domain waveform of the demodulated signal, calculates the log-likelihood ratio on each bit position according to the probability distribution of the values ​​of each frequency domain sampling point, sends the obtained log-likelihood ratio to the corresponding deinterleaver, passes the deinterleaved log-likelihood ratio as soft information input to the Turbo decoder, and performs Turbo decoding through the Max-Log-MAP iterative decoding algorithm. The problem of information loss during information transmission between the LoRa demodulator and the Turbo soft input and soft output decoder is solved. Compared with the LoRa signal frequency shift modulation system without Turbo coding, after adding Turbo coding, the bit error rate performance of the system has been significantly improved, and the bit error rate decrease rate has been significantly improved as the signal-to-noise ratio increases. Specific embodiment three:

[0129] This specification provides a soft demodulation device based on bit interleaved Turbo coded frequency shift chirp keying modulation, see Figure 7 , including: a sending end Turbo encoding module 701, a LoRa modulation module 702, a receiving end despreading and demodulation module 703, and a deinterleaving decoding module 704, wherein:

[0130] The transmitting end Turbo encoding module 701 is used to input the generated bit information into the Turbo encoder for encoding at the transmitting end, and output the encoded bit information;

[0131] The LoRa modulation module 702 is used to send the output codeword of the Turbo encoder to the interleaver for interleaving to obtain interleaved coded bits; perform LoRa modulation on the interleaved coded bits and map them into corresponding periodic frequency shift chirp signals, and encode and modulate all bits of the transmitting end into LoRa modulation signals;

[0132] The receiving end despreading and demodulation module 703 is used for demodulating the LoRa signal at the receiving end. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal, and obtains the bit logarithmic likelihood ratio using the LoRa demodulator logarithmic likelihood ratio calculation method;

[0133] The deinterleaving decoding module 704 is used to send the obtained log-likelihood ratio through a corresponding deinterleaver and then send it to a Turbo decoder for iterative deinterleaving.

[0134] Optionally, the Turbo encoder is composed of two component encoders, an interleaver, a redundancy puncturer and a parallel-to-serial conversion unit;

[0135] Among them, two component encoders are cascaded in parallel, and both use recursive systematic convolutional codes with the same structure;

[0136] The interleaver is located between two component encoders; one RSC encoder encodes the original input information bits, and the other RSC encoder encodes the information bits after passing through the interleaver. The check bits they output are input into the parallel-to-serial conversion unit together with the original information bits through the redundancy remover to combine into a codeword.

[0137] Optionally, LoRa modulation uses an up-chirp signal to transmit information, and its signal frequency changes linearly within one symbol period. The frequency change range and speed depend on the bandwidth B and the spreading factor SF. A LoRa signal carries SF bits of information. LoRa modulation carries information through different starting frequencies of the baseband signal.

[0138] Optionally, the LoRa modulation module 702 is specifically used for:

[0139] According to the set parameter bandwidth B and spreading factor SF, the signal frequency change range and the corresponding frequency change rate are determined, and the sampling interval of the LoRa signal meets T s =1 / B, a LoRa signal carries SF data bits, each SF of encoded data bits are divided into a group, the nth data bit in each group is defined as d[n], the information carried by the LoRa signal and the corresponding power-normalized periodic frequency shift signal are determined, and the modulated signal is sent to the receiving end.

[0140] Optionally, the receiving end despreading and demodulation module 703 is specifically used to:

[0141] At the receiving end, the power spectral density of the received LoRa signal energy and noise is calculated to obtain ζ i =|n i |With The probability density function of and

[0142] The received signal r(n) is multiplied by the original down-chirp signal to perform signal despreading, and the despread signal g(n) is subjected to fast Fourier transform to obtain the frequency domain waveform f(n) of the despread signal;

[0143] According to the probability distribution of the values ​​of each frequency domain sampling point, the log-likelihood ratio calculation method of the LoRa demodulator is used to calculate the bit log-likelihood ratio of each bit.

[0144] Optionally, the LoRa demodulator log-likelihood ratio calculation method includes:

[0145] According to the values ​​of each frequency sampling point, calculate the probability P(f(n)|u=0) that the current bit is 0:

[0146]

[0147] Where SF represents the spreading factor, Indicates i =|n i |with The probability density function of

[0148] According to the value of each frequency sampling point, calculate the probability P(f(n)|u=1) that the current bit is 1:

[0149]

[0150] Take the logarithm of the ratio of the probability that the current bit is 0 to the probability that the current bit is 1 to get the log-likelihood ratio of the current bit:

[0151] Optionally, the frequency domain waveform f(n) is grouped. When the first bit is 0, the corresponding f(n) frequency sampling point is 2 SF-1 points;

[0152] When the first bit is 1, the corresponding f(n) frequency sampling point is 2 SF-1 points.

[0153] Optionally, the deinterleaving and decoding module 704 is specifically configured to:

[0154] The obtained log-likelihood ratio is sent to the corresponding deinterleaver, and the deinterleaved LLR is passed to the Turbo decoder as soft information input, and Turbo decoding is performed through the Max-Log-MAP iterative decoding algorithm.

[0155] In summary, this embodiment performs Turbo encoding on the signal at the transmitting end, interleaves the encoded bits, performs LoRa modulation on the interleaved encoded bits and maps them into corresponding periodic frequency-shifted chirp signals, uses the original LoRa demodulation scheme at the receiving end, obtains the frequency domain waveform of the demodulated signal, calculates the log-likelihood ratio on each bit position according to the probability distribution of the values ​​of each frequency domain sampling point, sends the obtained log-likelihood ratio to the corresponding deinterleaver, passes the deinterleaved log-likelihood ratio as soft information input to the Turbo decoder, and performs Turbo decoding through the Max-Log-MAP iterative decoding algorithm. The problem of information loss during information transmission between the LoRa demodulator and the Turbo soft input and soft output decoder is solved. Compared with the LoRa signal frequency shift modulation system without Turbo coding, after adding Turbo coding, the bit error rate performance of the system has been significantly improved, and the bit error rate decrease rate has been significantly improved as the signal-to-noise ratio increases. Specific embodiment four:

[0157] This manual also provides a network device, see Figure 8 , the network device can implement the details of the method described in the above embodiment and achieve the same effect. Figure 8 As shown, the network device 800 includes: a processor 801, a transceiver 802, a memory 803, a user interface 804 and a bus interface, wherein:

[0158] In the embodiment of this specification, the network device 800 further includes: a computer program stored in the memory 803 and executable by the processor 801, and the computer program is executed by the processor 801 to implement the following steps:

[0159] At the transmitting end, the generated bit information is input into the Turbo encoder for encoding, and the encoded bit information is output;

[0160] The output codeword of the Turbo encoder is sent to the interleaver for interleaving to obtain the interleaved coded bits; the interleaved coded bits are modulated by LoRa and mapped into the corresponding periodic frequency shift chirp signal, and all bits at the transmitting end are coded and modulated into LoRa modulation signals;

[0161] The receiving end demodulates the LoRa signal. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal. The log-likelihood ratio calculation method of the LoRa demodulator is used to obtain the bit log-likelihood ratio.

[0162] The obtained log-likelihood ratio is sent to the Turbo decoder for iterative decoding after passing through the corresponding deinterleaver.

[0163] exist Figure 8 In the embodiment, the bus framework may include any number of interconnected buses and bridges, specifically one or more processors represented by processor 801 and various circuits of memory represented by memory 803 are linked together. The bus framework may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described in this document. The bus interface provides an interface. The transceiver 802 may be a plurality of components, namely, a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium. For different user devices, the user interface 804 may also be an interface capable of externally and internally connecting required devices, and the connected devices include but are not limited to a keypad, a display, a speaker, a microphone, a joystick, and the like.

[0164] The processor 801 manages the bus architecture and the usual processing, and the memory 803 can store data used by the processor 801 when performing operations. Optionally, when the computer program is executed by the processor 801, the following steps can also be implemented:

[0165] Optionally, the Turbo encoder is composed of two component encoders, an interleaver, a redundancy puncturer and a parallel-to-serial conversion unit;

[0166] Among them, two component encoders are cascaded in parallel, and both use recursive systematic convolutional codes with the same structure;

[0167] The interleaver is located between two component encoders; one RSC encoder encodes the original input information bits, and the other RSC encoder encodes the information bits after passing through the interleaver. The check bits they output are input into the parallel-to-serial conversion unit together with the original information bits through the redundancy remover to combine into a codeword.

[0168] Optionally, LoRa modulation uses an up-chirp signal to transmit information, and its signal frequency changes linearly within one symbol period. The frequency change range and speed depend on the bandwidth B and the spreading factor SF. A LoRa signal carries SF bits of information. LoRa modulation carries information through different starting frequencies of the baseband signal.

[0169] Optionally, the interleaved coded bits are subjected to LoRa modulation mapping to be converted into corresponding periodic frequency shift chirp signals, and all bits at the transmitting end are coded and modulated into LoRa modulation signals, including:

[0170] According to the set parameter bandwidth B and spreading factor SF, the signal frequency change range and the corresponding frequency change rate are determined, and the sampling interval of the LoRa signal meets T s=1 / B, a LoRa signal carries SF data bits, each SF of encoded data bits are divided into a group, the nth data bit in each group is defined as d[n], the information carried by the LoRa signal and the corresponding power-normalized periodic frequency shift signal are determined, and the modulated signal is sent to the receiving end.

[0171] Optionally, the receiving end demodulates the LoRa signal, the receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs a discrete Fourier transform on the multiplied received signal, and uses the LoRa demodulator log-likelihood ratio calculation method to obtain the bit log-likelihood ratio, including:

[0172] At the receiving end, the power spectral density of the received LoRa signal energy and noise is calculated to obtain ζ i =|n i |with The probability density function of and

[0173] The received signal r(n) is multiplied by the original down-chirp signal to perform signal despreading, and the despread signal g(n) is subjected to fast Fourier transform to obtain the frequency domain waveform f(n) of the despread signal;

[0174] According to the probability distribution of the values ​​of each frequency domain sampling point, the log-likelihood ratio calculation method of the LoRa demodulator is used to calculate the bit log-likelihood ratio of each bit.

[0175] Optionally, the LoRa demodulator log-likelihood ratio calculation method includes:

[0176] According to the values ​​of each frequency sampling point, calculate the probability P(f(n)|u=0) that the current bit is 0:

[0177]

[0178] Where SF represents the spreading factor, Indicates i =|n i |with The probability density function of

[0179] According to the value of each frequency sampling point, calculate the probability P(f(n)|u=1) that the current bit is 1:

[0180]

[0181] Take the logarithm of the ratio of the probability that the current bit is 0 to the probability that the current bit is 1 to get the log-likelihood ratio of the current bit:

[0182] Optionally, the frequency domain waveform f(n) is grouped. When the first bit is 0, the corresponding f(n) frequency sampling point is 2 SF-1 points;

[0183] When the first bit is 1, the corresponding f(n) frequency sampling point is 2 SF-1 points.

[0184] Optionally, the step of sending the obtained log-likelihood ratio through a corresponding deinterleaver and then sending it to a Turbo decoder for iterative decoding includes:

[0185] The obtained log-likelihood ratio is sent to the corresponding deinterleaver, and the deinterleaved LLR is passed to the Turbo decoder as soft information input, and Turbo decoding is performed through the Max-Log-MAP iterative decoding algorithm.

[0186] In summary, this embodiment performs Turbo encoding on the signal at the transmitting end, interleaves the encoded bits, performs LoRa modulation on the interleaved encoded bits and maps them into corresponding periodic frequency-shifted chirp signals, uses the original LoRa demodulation scheme at the receiving end, obtains the frequency domain waveform of the demodulated signal, calculates the log-likelihood ratio on each bit position according to the probability distribution of the values ​​of each frequency domain sampling point, sends the obtained log-likelihood ratio to the corresponding deinterleaver, passes the deinterleaved log-likelihood ratio as soft information input to the Turbo decoder, and performs Turbo decoding through the Max-Log-MAP iterative decoding algorithm. The problem of information loss during information transmission between the LoRa demodulator and the Turbo soft input and soft output decoder is solved. Compared with the LoRa signal frequency shift modulation system without Turbo coding, after adding Turbo coding, the bit error rate performance of the system has been significantly improved, and the bit error rate decrease rate has been significantly improved as the signal-to-noise ratio increases.

[0187] The above description is only a preferred embodiment of this specification and is not intended to limit this specification. For those skilled in the art, this specification may have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification shall be included in the protection scope of this specification.

Claims

1. A soft demodulation algorithm based on bit-interleaved turbo coded frequency-shift chirp keying modulation, characterized in that: include: At the transmitting end, the generated bit information is input into the Turbo encoder for encoding, and the encoded bit information is output; The output codeword of the Turbo encoder is sent to the interleaver for interleaving to obtain the interleaved coded bits; the interleaved coded bits are modulated by LoRa and mapped into the corresponding periodic frequency shift chirp signal, and all bits at the transmitting end are coded and modulated into LoRa modulation signals; The receiving end demodulates the LoRa signal. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal. The log-likelihood ratio calculation method of the LoRa demodulator is used to obtain the bit log-likelihood ratio. The obtained log-likelihood ratio is sent to the Turbo decoder for iterative decoding after passing through the corresponding deinterleaver.

2. The method according to claim 1, characterized in that The turbo encoder is composed of two component encoders, an interleaver, a puncturer and a parallel-to-serial conversion unit; Among them, two component encoders are cascaded in parallel, and both use recursive systematic convolutional codes with the same structure; The interleaver is located between two component encoders; one RSC encoder encodes the original input information bits, and the other RSC encoder encodes the information bits after passing through the interleaver. The check bits they output are input into the parallel-to-serial conversion unit together with the original information bits through the redundancy remover to combine into a codeword.

3. The method according to claim 2, characterized in that LoRa modulation uses an up-chirp signal to transmit information. Its signal frequency changes linearly within one symbol period. The frequency change range and speed depend on the bandwidth B and the spreading factor SF. A LoRa signal carries SF bits of information. LoRa modulation carries information through different starting frequencies of the baseband signal.

4. The method according to claim 3, characterized in that The interleaved coded bits are subjected to LoRa modulation mapping to be corresponding periodic frequency shift chirp signals, and all bits of the transmitting end are coded and modulated to be LoRa modulation signals, including: According to the set parameter bandwidth B and spreading factor SF, the signal frequency change range and the corresponding frequency change rate are determined, and the sampling interval of the LoRa signal meets T s =1 / B, a LoRa signal carries SF data bits, each SF of encoded data bits are divided into a group, the nth data bit in each group is defined as d[n], the information carried by the LoRa signal and the corresponding power-normalized periodic frequency shift signal are determined, and the modulated signal is sent to the receiving end.

5. The method according to claim 4, characterized in that The receiving end demodulates the LoRa signal, the receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal, and uses the LoRa demodulator log-likelihood ratio calculation method to obtain the bit log-likelihood ratio, including: At the receiving end, the power spectral density of the received LoRa signal energy and noise is calculated to obtain ζ i = ｜ n i | and The probability density function of and Where n i represents the noise amplitude of the i-th sampling point, P L Indicates signal power; The received signal r(n) is multiplied by the original down-chirp signal to perform signal despreading, and the despread signal g(n) is subjected to fast Fourier transform to obtain the frequency domain waveform f(n) of the despread signal; According to the probability distribution of the values ​​of each frequency domain sampling point, the log-likelihood ratio calculation method of the LoRa demodulator is used to calculate the bit log-likelihood ratio of each bit.

6. The method according to claim 5, characterized in that The LoRa demodulator log-likelihood ratio calculation method comprises: According to the values ​​of each frequency sampling point, calculate the probability P(f(n)|u=0) that the current bit is 0: Where SF represents the spreading factor, Indicates i = | n i| and The probability density function of , where i, j represent the sampling point numbers; According to the value of each frequency sampling point, calculate the probability P(f(n)|u=1) that the current bit is 1: Take the logarithm of the ratio of the probability that the current bit is 1 to the probability that the current bit is 1, and get the log-likelihood ratio of the current bit:

7. The method according to claim 6, characterized in that The frequency domain waveform f(n) is grouped according to the binary number corresponding to each sampling point. When the first bit is 0, the corresponding f(n) frequency sampling point before 2 SF-1 points; when the first bit is 1, it corresponds to 2 points after the f(n) frequency sampling point SF-1 points.

8. The method according to claim 7, characterized in that The obtained log-likelihood ratio is sent to a Turbo decoder for iterative decoding after passing through a corresponding deinterleaver, including: The obtained log-likelihood ratio is sent to the corresponding deinterleaver, and the deinterleaved LLR is passed to the Turbo decoder as soft information input, and Turbo decoding is performed through the Max-Log-MAP iterative decoding algorithm.

9. A soft demodulation device based on bit-interleaved turbo coding frequency shift chirp keying modulation, used to implement the method according to any one of claims 1 to 8, characterized in that: include: The transmitting end Turbo coding module, LoRa modulation module, receiving end despreading and demodulation module, deinterleaving and decoding module, among which: The transmitting end Turbo encoding module is used to input the generated bit information into the Turbo encoder for encoding at the transmitting end, and output the encoded bit information; The LoRa modulation module is used to send the output codeword of the Turbo encoder to the interleaver for interleaving to obtain the interleaved coded bits; perform LoRa modulation mapping on the interleaved coded bits to the corresponding periodic frequency shift chirp signal, and encode and modulate all the bits of the transmitting end into LoRa modulation signals; The receiving end despreading and demodulation module is used for demodulating the LoRa signal at the receiving end. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal, and obtains the bit logarithmic likelihood ratio using the LoRa demodulator logarithmic likelihood ratio calculation method; The deinterleaving decoding module is used to send the obtained log-likelihood ratio through a corresponding deinterleaver and then send it to a Turbo decoder for iterative deinterleaving.

10. A network device, characterized in that: include: Communications interface, processor and memory; The processor calls the program instructions in the memory to perform the following actions: At the transmitting end, the generated bit information is input into the Turbo encoder for encoding, and the encoded bit information is output; The output codeword of the Turbo encoder is sent to the interleaver for interleaving to obtain the interleaved coded bits; the interleaved coded bits are modulated by LoRa and mapped into the corresponding periodic frequency shift chirp signal, and all bits at the transmitting end are coded and modulated into LoRa modulation signals; The receiving end demodulates the LoRa signal. The receiver multiplies the received signal with the original down-chirp signal synchronized with the time to perform signal despreading, and performs discrete Fourier transform on the multiplied received signal. The log-likelihood ratio calculation method of the LoRa demodulator is used to obtain the bit log-likelihood ratio. The obtained log-likelihood ratio is sent to the Turbo decoder for iterative decoding after passing through the corresponding deinterleaver.

Images