CN117614547A - Delayed serial cascade pulse position modulation system and method and deep space optical communication system - Google Patents

Abstract

The invention relates to a delay serial cascade pulse position modulation system, a delay serial cascade pulse position modulation method and a deep space optical communication system. Relates to the technical field of coding and decoding applicable to deep space optical communication. The delay serial cascade pulse position modulation system comprises a transmitting end and a receiving end. The transmitting end comprises a coding module, a transmitting terminal symbol sequence segmentation module, a sub-symbol interleaving module, a delay module, a sub-symbol sequence merging module and a PPM modulation module which are sequentially connected. The receiving end comprises a PPM sub-symbol demodulation module, a receiving terminal symbol sequence segmentation module, a reverse delay module, a sub-symbol de-interleaving module, a decoding module and a feedback module which are connected with the decoding module and the PPM sub-symbol demodulation module in sequence. With lower computational complexity and better error performance.

Description

Delayed serial cascade pulse position modulation system, method and deep space optical communication system

Technical Field

The present invention relates to the field of coding technology applicable to deep space optical communication, and in particular to a delayed serial cascade pulse position modulation system, method and deep space optical communication system.

Background Art

Serial-Concatenated-Pulse-Position-Modulation (SCPPM) is a coding system for deep space optical communication proposed by NASA in 2005 to support the Mars Demonstration System (MLCD). SCPPM coding is a serial cascade structure. The outer code is a convolutional code, and the inner code is composed of an accumulator and a PPM encoder. This part is also called APPM code. The outer code and the inner code are connected by a bit interleaver. The characteristic of SCPPM is that it uses a relatively simple coding method under PPM modulation to achieve high-reliability deep space optical communication.

When using an M=2 ^m- bit PPM encoder, the SCPPM combines the operations of the accumulator at m consecutive moments and regards it as an equivalent M-bit APPM encoder, referred to as M-APPM. The inventors found that the complexity of M-APPM decoding is positively correlated with the modulation order. Specifically, the computational complexity and the number of grid edges are linearly related to O(2.2 ^m ). For example, the number of grid edges of 16-APPM is 2*16=32, the number of grid edges of 4-APPM is 2*4=8, and the number of accumulator grid edges is 4. At this time, the computational complexity of 16-APPM decoding is about 4 times that of 4-APPM decoding and 8 times that of accumulator decoding.

In order to solve the problem of high computational complexity in M-APPM decoding, the prior art also provides a decoding method based on single bit demodulation, which is a low-complexity decoding method. The decoding method specifically demodulates the received PPM symbol to obtain the bit corresponding to the check bit stream. As the decoder input. After that, the decoder performs iterative decoding between the accumulator and the convolutional code. The disadvantage of the single bit demodulation decoding method is that the error performance is poor. For high-order modulation, demodulating the received PPM symbol to obtain the bit LLR will inevitably bring information loss, resulting in a loss of decoding error performance.

In order to solve the problem of decoding difference performance loss when the single bit demodulation and decoding method is applied to high-order modulation, the prior art also provides a delay-based bit-interleaved coded modulation (BICM), which can be defined as DBICM. DBICM is a general coding modulation method for high-order modulation. DBICM introduces feedback information from the decoder output to the demodulator at the receiving end, so it can reduce the information loss when demodulating the received symbol to obtain the bit LLR to a certain extent.

However, the disadvantage of DBICM is that if the coding part of SCPPM (convolutional code + accumulator) is directly brought into the encoder of the DBICM transmitter, the inner code does not have the APPM structure, so the decoder can only use the bit-level decoding scheme instead of the APPM decoding scheme. Therefore, although DBICM is improved compared with the single-bit demodulation decoding method, it still has a certain gap with the M-APPM encoder.

In view of this, it is urgent to propose a modulation system based on SCPPM, which should have lower computational complexity and better error performance.

Summary of the invention

The object of the present invention is to provide a delayed serial cascade pulse position modulation system, method and deep space optical communication system, which have lower computational complexity and better error performance.

In the first aspect, the present invention provides a delayed serial cascade pulse position modulation system, including a transmitting end and a receiving end. The transmitting end includes a sequentially connected encoding module, a transmitting terminal symbol sequence segmentation module, a sub-symbol interleaving module, a delay module, a sub-symbol sequence merging module and a PPM modulation module. The receiving end includes a sequentially connected PPM sub-symbol demodulation module, a receiving terminal symbol sequence segmentation module, a reverse delay module, a sub-symbol deinterleaving module and a decoding module, and a feedback module connecting the decoding module and the PPM sub-symbol demodulation module.

The encoding module receives the original data sequence and outputs the codeword sequence after encoding. The transmitting terminal symbol sequence segmentation module outputs the sub-symbol sequence based on the codeword sequence. The sub-symbol interleaving module receives the sub-symbol sequence corresponding to each codeword, interleaves the bit sequence in the sub-symbol sequence, and outputs the interleaved sub-symbol sequence. The delay module is configured with a delay parameter, and the interleaved sub-symbol sequence is input into the delay module at a preset time, and a delayed interleaved sub-symbol sequence is output after the delay parameter is reached. The sub-symbol sequence merging module receives the delayed interleaved sub-symbol sequence at a preset time, and merges all delayed interleaved sub-symbol sequences into a delayed codeword sequence. The PPM modulation module receives the delayed codeword sequence and performs M-PPM modulation, and outputs a PPM symbol sequence.

The PPM sub-symbol demodulation module is configured with prior information. The PPM sub-symbol demodulation module receives a received signal corresponding to the PPM symbol sequence. The PPM sub-symbol demodulation module determines the sub-symbol log-likelihood ratio using the prior information and the received signal. The receiving terminal symbol sequence segmentation module receives the sub-symbol log-likelihood ratio, and first segments and then merges it into a sub-symbol log-likelihood ratio sequence. The reverse delay module caches the sub-symbol log-likelihood ratio, and outputs it to the sub-symbol deinterleaving module after all the sub-symbol log-likelihood ratios are collected. The sub-symbol deinterleaving module deinterleaves the sub-symbol log-likelihood ratio sequence separately and outputs the deinterleaved sub-symbol log-likelihood ratio. The decoding module outputs the decoding result and the sub-symbol feedback sequence based on the deinterleaved sub-symbol log-likelihood ratio. The sub-symbol feedback sequence is sent to the PPM sub-symbol demodulation module through the feedback module. In the next demodulation and decoding cycle, the sub-symbol feedback sequence and the received signal participate in the demodulation and decoding process together.

Compared with the prior art, the delayed serial cascade pulse position modulation system provided by the present invention divides each codeword in the codeword sequence into multiple sub-symbol sequences using a transmitting terminal symbol sequence division module, and further interleaves the bit sequence in the sub-symbol sequence to obtain an interleaved sub-symbol sequence. Based on this, part of the interleaved sub-symbol sequence is delayed and modulated together with the sub-symbol sequence corresponding to the subsequent codeword. At this time, each PPM symbol sequence will contain multiple sub-symbol sequences from different codewords.

Moreover, at the receiving end, the received signal received at each moment will use the sub-symbol feedback sequence successfully decoded at the previous moment to assist in demodulation, and obtain the sub-symbol log-likelihood ratio. When the log-likelihood ratios of all sub-symbol sequences corresponding to all codewords are completely received, the receiving end decodes the code block and feeds back the soft decision or hard decision decoding results to the PPM sub-symbol demodulation module to assist in the demodulation of the subsequent received signal. In view of this, the delayed serial cascade pulse position modulation system provided by the present invention has better error performance.

Furthermore, the key point of the present invention is to adopt a joint method of splitting the codeword into sub-symbol sequences, performing delay modulation after interleaving, and performing APPM decoding on the sub-symbols, and the joint method splits a high-order APPM equivalent into low-order APPMs by delay. Without changing the original encoding module of the existing serial cascade pulse position modulation system (SCPPM) and using APPM decoding, the decoding complexity of the inner code is significantly reduced compared with the existing SCPPM.

In a second aspect, the present invention further provides a delayed serial cascade pulse position modulation method. The delayed serial cascade pulse position modulation system provided in the first aspect performs the delayed serial cascade pulse position modulation method, comprising the following steps:

S10. The transmitter outputs a PPM symbol sequence based on the original data sequence;

S20. The receiving end outputs a decoding result and a sub-symbol feedback sequence based on the PPM symbol sequence, wherein the sub-symbol feedback sequence is fed back to the transmitting end and participates in the demodulation and decoding process together with the received signal in the next demodulation and decoding cycle.

Compared with the prior art, the beneficial effects of the delayed serial cascade pulse position modulation method provided by the present invention are the same as the beneficial effects of the delayed serial cascade pulse position modulation system provided by the first aspect and/or any implementation of the first aspect, and will not be elaborated here.

In a third aspect, the present invention further provides a deep space optical communication system, wherein the encoding and decoding system adopted by the deep space optical communication system is the delayed serial cascade pulse position modulation system provided in the first aspect.

Compared with the prior art, the beneficial effects of the deep space optical communication system provided by the present invention are the same as the beneficial effects of the delayed serial cascade pulse position modulation system provided by the first aspect and/or any implementation method of the first aspect, and will not be elaborated here.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

Figure 1 is a block diagram of a point-to-point laser communication system model;

Figure 2 is a trellis diagram of a (5, 7) convolutional code;

Fig. 3 is a schematic diagram of the structure of the SCPPM encoder;

FIG4 is a schematic diagram of a (5, 7) convolutional code encoder;

Fig. 5 is a schematic diagram of an accumulator encoder;

Fig. 6 is a grid diagram of the accumulator;

Figure 7 is an equivalent 4-APPM grid diagram;

FIG8 is a comparison chart of the performance of APPM decoding and single demodulation decoding in SCPPM;

FIG9 is a schematic diagram of the delayed BICM encoder structure;

FIG10 is a schematic diagram of the delayed BICM decoder structure;

FIG11 is a schematic diagram of a DSCPPM coding and modulation process;

FIG12 is a schematic diagram of a DSCPPM demodulation and decoding process;

FIG13 is a schematic diagram of DSCPPM of {d=(2|2|2), δ=[0, 1, 2]};

Fig. 14 is a grid diagram of an accumulator;

Fig. 15 is a trellis diagram of an accumulator adding 2-bit sub-symbols;

Fig. 16 is a grid diagram of an accumulator;

Fig. 17 is a grid diagram of an accumulator adding 2 and 3 bit sub-symbols;

FIG18 is a performance comparison chart of SCPPM and DSCPPM;

Figure 19 is a performance comparison chart of SCPPM and DSCPPM.

DETAILED DESCRIPTION

In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly on the other element or indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element.

In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined. The meaning of "several" is one or more, unless otherwise clearly and specifically defined.

In the description of the present invention, it is necessary to understand that the directions or positional relationships indicated by terms such as “upper”, “lower”, “front”, “backward”, “left” and “right” are based on the directions or positional relationships shown in the accompanying drawings and are only for the convenience of describing the present invention and simplifying the description. They do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they cannot be understood as limitations on the present invention.

In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

FIG1 is a block diagram of a point-to-point laser communication system model. Referring to FIG1 , the coding principle in the laser communication system can be summarized as follows:

The original data u is channel-coded to generate a coding sequence c, which is then PPM-modulated to obtain an optical pulse x. Channel coding and PPM modulation can be summarized as the transmitter. After the optical pulse passes through the channel (specifically, it can be a Poisson channel, but is not limited to this), it reaches the receiver in the form of a photon. The photon detector included in the receiver counts the arriving photons to obtain the photon count y corresponding to the optical pulse x. The receiver uses y to make a soft decision or a hard decision, calculates the log-likelihood ratio (LLR), obtains the symbol LLR, and then performs PPM demodulation and channel decoding on the symbol LLRΓ(y), and finally obtains the predicted value of the original data. It needs to be further explained that PPM demodulation can also be called PPM decoding. In practical applications, independent PPM demodulators and channel decoders can be selected to perform PPM demodulation respectively according to a specific coding scheme, or an integrated decoder can be selected to perform PPM demodulation and channel decoding simultaneously.

In the point-to-point laser communication system model shown in FIG1 , pulse position modulation (PPM) modulation and PPM demodulation are key units that affect the computational complexity and error performance of coding.

Among them, PPM modulation uses the relative position of a single pulse within a period of time to transmit information. In an M-bit PPM (M-PPM, also known as m-order PPM, m=log ₂ M>1) modulation system, each m-bit information c=[c ₁ , c ₂ , ..., _cm ] is mapped to a PPM symbol x. The transmission of a PPM symbol requires M time slots (unit time), and only one time slot in each symbol transmits an optical pulse, and the position where the optical pulse appears represents the value of c. Therefore, x can be written as an M-bit pulse sequence x=[x ₁ , x ₂ , ..., x _M ], where x _i ={0,1} represents that the i-th time slot of the symbol has/has no optical signal. The one-to-one mapping relationship between the value of c and the position of the optical pulse can use a variety of mapping schemes, such as natural order mapping, Gray mapping, and anti-Gray mapping. In the embodiment of the present invention, unless otherwise specified, PPM modulation uses natural order mapping.

Natural order mapping PPM: Given bit information c = [c ₁ , c ₂ , ..., _cm ], perform bit-to-decimal conversion (c ₁ is the most significant bit) to obtain the symbol value 0≤C≤M-1. The pulse sequence of the PPM symbol corresponding to C is x=[ _x1 , _x2 , ..., _xM ], _xC+1 =1, _{x1≤i≤M, i≠C+1} =0, that is, in the 1st to Mth time slots of a PPM symbol, the C+1th time slot sends an optical pulse, and the remaining time slots do not send pulses.

Consider a PPM symbol. The sequence of photon numbers detected by the receiver in each time slot is recorded as y = [y ₁ , y ₂ , ..., y _M ]. The receiver first uses y to perform soft decision calculation to obtain the time slot transfer probability Pr{y _i | _xi }, and then calculates the symbol transfer probability Pr{y|C} and the symbol log-likelihood ratio (LLR, ), where log represents the natural logarithm if no additional base is specified. The symbol LLR Γ(y)＝{Γ _C (y): 0≤C≤M-1} will be used as the input of the decoder. For an ideal Poisson channel, the accurate symbol LLR can be calculated using Γ _C (y)＝y _C+1 ·log(1+ _ns / _nb ). Note that the symbol LLR calculation formula may be different in an actual system, but the definition of the symbol LLR remains unchanged. For example, the symbol LLR calculation formula can be Γ _C (y)＝y _C+1 or Γ _C (y)＝(y _C+1 -y ₀ )·log(1+ _ns / _nb ).

In addition, for PPM demodulation, the embodiment of the present invention describes the calculation of PPM demodulation or PPM decoding based only on channel information (ie, received signal y). It should be understood that the following description is only for explanation and not for limitation.

Symbol set definition: Let Represents the collection of all PPM symbols. The following defines several Subset of: Let represents the set of symbols whose value at the i-th bit position is 0, represents the set of symbols whose value at the i-th bit position is 1, Example: For 8-PPM,

The symbol transition probability Pr{y|C} is used to calculate the bit transition probability Pr{y|c _i = 0}, Pr{y|c _i = 1} and the bit LLR when the bit c _i (i = 1, 2, ..., m) is 0 and 1 respectively. The detailed definitions are as follows:

Definition of operator max ^* : max ^* represents addition calculation in the natural logarithm domain (hereinafter referred to as logarithm domain). That is, the corresponding calculation of a+b=c in the logarithm domain is max ^* (log(a), log(b))=log(c). It represents the continuous addition of x that meets the conditions.

Convolutional code is a widely used coding scheme. Its main feature is that the continuous input information sequence u obtains the continuous output encoded sequence v. For a convolutional code with an input length of k bits, an output length of n bits, and a constraint length of α, the register in the encoder can retain some values of the encoding process of α-1 consecutive moments. The encoder sequentially inputs k bits of the information sequence u at time t, and jointly determines the values of the n output bits based on the input at time t and the data stored in the register related to time t-α+1 to time t-1 (the value stored in the register is called the encoder state), and updates the value of the register at the same time. In other words, the output at time t is related to the input information from time t-α+1 to time t.

The encoding process of a convolutional code can be represented by a grid diagram. In the grid diagram, each column of points represents the encoder state at a certain moment, the lines between states represent state transitions, and the numbers on the lines represent the input/output corresponding to the state transition. A common (5, 7) convolutional code grid diagram is shown in Figure 2.

Convolutional codes usually use decoding algorithms based on trellis graphs, among which the most representative ones are the Viterbi algorithm and the BCJR algorithm. The calculation process of these two decoding algorithms is closely related to the trellis graph structure, so their computational complexity is also linearly positively correlated with the number of edges in the trellis graph (the total number of connections between the states at two moments), that is, the decoding complexity is O(2 ^α-1 ·2 ^k ).

SCPPM coding is a scheme for PPM coding. The structure of the SCPPM encoder is shown in Figure 3. First, the SCPPM encoder inputs data u of length K, and uses a convolutional code encoder with a code rate of R to generate a bit sequence v of length N = K/R. After that, the bit interleaver rearranges the bits in v to obtain v'. The accumulator inputs v' to generate a bit sequence c of the same length. Finally, c passes through the PPM encoder to obtain a PPM symbol sequence C of K/(mR) pieces and a corresponding pulse sequence x of MK/(mR) time slots.

The outer code encoder of SCPPM is a convolutional code. Generally speaking, the convolutional code can use any convolutional code with a code rate less than 1. In the embodiment of the present invention, a common convolutional code with a code rate of 1/2 and a generating formula of (5, 7) is considered to be used. The structure block diagram of the convolutional code is shown in Figure 4, where D represents a register, represents binary addition (modulo 2 addition) operation. Given the input bit sequence u = [u ₁ , u ₂ , ..., u _K ] of the (5, 7) convolutional code, at time t, the encoder inputs _ut and outputs two bits and Finally, the total output bit stream of the convolutional code encoder is v = [v _1,1 ,v _1,2 ,v _2,1 ,v _2,2 , ...,v _K,1 ,v _K,2 ].

The coding structure block diagram of the accumulator is shown in Figure 5. The accumulator can be regarded as a special recursive convolutional code with a code rate of 1 (1 bit input and 1 bit output at each time), and its generator can be written as 1/(1+D). Given an input sequence v′＝[v ₁ ,v ₂ ,...,v _N ], the output sequence is c′＝[c ₁ ,c _2, ...,c _N ], where c ₁ ＝u ₁ ,

For M= ^2m- bit PPM (M-PPM), the PPM encoder inputs c _(m-1)t+1 , c _(m-1)t+2 , ..., _cmt at time t and outputs _Ct =bi2de([c _(m-1)t+1 , c _(m-1)t+2 , ..., _cmt ]).

The SCPPM structure is special in that the accumulator and PPM encoder of the inner code part can also be regarded as a whole, called APPM encoder. By regarding the inner code part as two separate structures or a whole, different decoding schemes can be implemented accordingly.

When using an M=2 ^m- bit PPM encoder, the APPM encoder combines the operations of the accumulator at m consecutive moments and regards it as an equivalent M-bit APPM encoder (M-APPM), that is, the M-APPM encoder inputs v _(m-1)t+1 , v _(m-1)t+2 , ..., v _mt at time t, and then directly outputs the PPM symbol C _t . Correspondingly, the receiving end will directly use the PPM symbol LLR Γ _C (y) for decoding based on the grid diagram using M-APPM.

In the M-APPM grid graph, given the encoder state s ₁ = a at the current moment, there are ^2m outputs that can make the encoder transition to the same state s ₂ = b at the next moment. In other words, each pair of state transitions (a→b) contains ^2m parallel edges.

As an example, the accumulator grid diagram and the 4-APPM grid diagram are shown in Figures 6 and 7. The accumulator grid spanning m = 2 time points (t = 1 to t = 3) is merged into a 4-APPM grid spanning 1 time point (t = 1 to t = 2), that is, the state transition a→b→c in the accumulator grid is merged into the state transition a→c in the 4-APPM grid. Note that although a→0→c and a→1→c in the accumulator grid correspond to different inputs/outputs, they can both be merged into a→c in the 4-APPM grid, and these two are called parallel edges of a→c.

The inventors found that the APPM decoding scheme has a disadvantage, that is, the complexity of M-APPM decoding is positively correlated with the modulation order. Specifically, the computational complexity is linearly related to the number of edges of the trellis graph, O(2·2 ^m ). For example, the number of edges of the 16-APPM trellis graph is 2*16=32, the number of edges of the 4-APPM trellis graph is 2*4=8, and the number of edges of the accumulator trellis graph is 4. Therefore, the computational complexity of 16-APPM decoding is about 4 times that of 4-APPM decoding and 8 times that of accumulator decoding.

In view of this, the inventors tried to develop a low-complexity decoding scheme, namely a single bit demodulation scheme. This scheme demodulates the received PPM symbol to obtain the bit LLR corresponding to the check bit stream c As the decoder input. After that, the decoder performs iterative decoding between the accumulator and the convolutional code.

The disadvantage of the single bit demodulation scheme is that the error performance is poor. For high-order modulation, demodulating the received symbol to obtain the bit LLR will inevitably lead to information loss, resulting in a loss of decoding error performance.

Referring to FIG8 , when M=16 and nb=0.1, the single bit demodulation scheme has a 1 dB performance loss relative to the APPM decoding scheme.

Delayed BICM (DBICM) is a general coding modulation scheme for high-order modulation that does not specify a specific coding and modulation format. This scheme introduces feedback information from the decoder output to the demodulator at the receiving end, thereby reducing the information loss when demodulating the received symbol to obtain the bit LLR to a certain extent.

The transmitter of DBICM is shown in FIG9 . The coding here can be any channel coding, and the modulation can also be any type of high-order modulation. At time t, the bit sequence (after interleaving) of the t-th code block c(t) is divided into m subsequences of equal length c(t, 1), ..., c(t, m), of which at most m-1 subsequences are delayed and sent at a subsequent time. The subsequences that are not delayed are modulated and sent together with the subsequences in the previous codeword that are retained to the current time by the delay module. Specifically, the subsequence sent at the t-th time is c(t-δ ₁ , 1), ..., c(t-δ _m , m), and the number of delayed bits δ = {δ ₁ , ..., δ _m } must have at least one value equal to 0 (i.e., the subsequence in c(t) that is not delayed), and the rest of the values are greater than 0 (i.e., the sequence that was delayed in the previous codeword). During modulation, each subsequence will be mapped to a fixed bit position in the modulation symbol.

The processing flow of the DBICM receiver is shown in Figure 10. The signal y(t) received at each moment will use the feedback information of the previously successfully decoded code block to assist in demodulation, and obtain the log-likelihood ratio (LLR) results of the subsequence Γ(c(t-δ ₁ ,1)),...,Γ(c(t-δ _m ,m)). When all the LLR results Γ(c(t)) with subsequences of c(t) are completely received, the receiver decodes the code block and converts the soft decision or hard decision decoding results into Feedback to the demodulator to assist in the subsequent demodulation of the signal. Therefore, the decoding information of each code block can be coupled with other code blocks at the symbol level, and the feedback information to the demodulator can be used to assist the demodulation of other code blocks, thereby improving the overall decoding error performance.

In DBICM, the selection of delayed sub-symbols and delay time has an impact on the decoding error performance. At the same time, the use of hard decision or soft decision feedback information during demodulation at the receiving end and whether iterative demodulation is performed will also affect the overall decoding error performance.

The disadvantage of this scheme is that if the coding part of SCPPM (convolutional code + accumulator) is directly brought into the encoder of the DBICM transmitter, the inner code does not have the APPM structure, so the decoder can only use the bit-level decoding scheme instead of the APPM decoding scheme. Therefore, although the performance of DBICM is improved compared with the single-bit demodulation scheme, it still lags behind SCPPM coding.

In summary, in high-order modulation, treating demodulation and decoding as two completely independent steps will inevitably lead to error performance loss, and the complexity of (demodulation + decoding) integrated processing is high. For SCPPM, the APPM decoding scheme with high error performance is highly complex, but the low-complexity single-bit demodulation scheme has too large an error performance loss.

In view of the above problems, the embodiment of the present invention proposes an improved coding modulation scheme based on SCPPM, called delayed SCPPM (hereinafter abbreviated as DSCPPM), and its corresponding demodulation and decoding scheme. In general, the DSCPPM achieves an error performance that is better than the SCPPM single bit demodulation scheme when the inner code decoding complexity is significantly smaller than the APPM decoding of SCPPM, and in some cases can achieve an error performance close to that of the APPM decoding of SCPPM.

Consider the case where multiple codewords {c(t)|t＝1，2，...，T} need to be sent continuously. For SCPPM, c(t) is directly PPM modulated to obtain the symbol sequence x(t). For DSCPPM, we divide each codeword c(t) into multiple sub-symbol sequences {S(t，i)|i＝1，2，...，m′}, and delay some of the sub-sequences and modulate them together with the subsequent sub-sequences of the codewords. Therefore, each symbol sequence x(t) will contain multiple sub-symbol sequences from different codewords.

At the receiving end, the signal y(t) received at each moment will use the feedback information of the previously successfully decoded code block to assist in demodulation and obtain the log-likelihood ratio (LLR) result of the subsequence. When all the LLRs with subsequences of c(t) are completely received, the receiving end decodes the code block and feeds back the soft decision or hard decision decoding results to the demodulator to assist in the demodulation of subsequent signals.

Referring to FIG. 11 and FIG. 12 , the present invention provides a delayed serial cascade pulse position modulation system, including a transmitting end and a receiving end. The transmitting end includes a sequentially connected encoding module, a transmitting terminal symbol sequence segmentation module, a sub-symbol interleaving module, a delay module, a sub-symbol sequence merging module and a PPM modulation module. The receiving end includes a sequentially connected PPM sub-symbol demodulation module, a receiving terminal symbol sequence segmentation module, a reverse delay module, a sub-symbol deinterleaving module and a decoding module, and a feedback module connecting the decoding module and the PPM sub-symbol demodulation module.

The encoding module receives the original data sequence and outputs the codeword sequence after encoding. The transmitting terminal symbol sequence segmentation module outputs the sub-symbol sequence based on the codeword sequence. The sub-symbol interleaving module receives the sub-symbol sequence corresponding to each codeword, interleaves the bit sequence in the sub-symbol sequence, and outputs the interleaved sub-symbol sequence. The delay module is configured with a delay parameter, and the interleaved sub-symbol sequence is input into the delay module at a preset time, and a delayed interleaved sub-symbol sequence is output after the delay parameter is reached. The sub-symbol sequence merging module receives the delayed interleaved sub-symbol sequence at a preset time, and merges all delayed interleaved sub-symbol sequences into a delayed codeword sequence. The PPM modulation module receives the delayed codeword sequence and performs M-PPM modulation, and outputs a PPM symbol sequence.

The PPM sub-symbol demodulation module is configured with prior information. The PPM sub-symbol demodulation module receives a received signal corresponding to the PPM symbol sequence. The PPM sub-symbol demodulation module determines the sub-symbol log-likelihood ratio using the prior information and the received signal. The receiving terminal symbol sequence segmentation module receives the sub-symbol log-likelihood ratio, and first segments and then merges it into a sub-symbol log-likelihood ratio sequence. The reverse delay module caches the sub-symbol log-likelihood ratio, and outputs it to the sub-symbol deinterleaving module after all the sub-symbol log-likelihood ratios are collected. The sub-symbol deinterleaving module deinterleaves the sub-symbol log-likelihood ratio sequence separately and outputs the deinterleaved sub-symbol log-likelihood ratio. The decoding module outputs the decoding result and the sub-symbol feedback sequence based on the deinterleaved sub-symbol log-likelihood ratio. The sub-symbol feedback sequence is sent to the PPM sub-symbol demodulation module through the feedback module. In the next demodulation and decoding cycle, the sub-symbol feedback sequence and the received signal participate in the demodulation and decoding process together.

As a possible implementation, the encoding module in this embodiment may specifically include the convolutional code encoding, bit interleaving and accumulator shown in Figure 3. The encoding module receives the original data sequence and outputs the codeword sequence after encoding. The specific execution is as follows: all data u(1), u(2), ..., u(T) are encoded respectively to obtain codewords c(1), c(2), ..., c(T) respectively.

As a possible implementation method, the sending terminal symbol sequence segmentation module outputs a sub-symbol sequence based on the codeword sequence, specifically including: the sending terminal symbol sequence segmentation module receives the codeword sequence, and first segments each codeword in the codeword sequence into a bit sequence, and then merges the bit sequence into a sub-symbol sequence.

As an example, the transmitting terminal symbol sequence segmentation module (abbreviated as S/P), for a codeword c(t), 1≤t≤T, first segments c(t) into m bit sequences c(t, 1), ..., c(t, m), and then merges them into sub-symbol sequences S(t, 1), ..., S(t, m′). The bit sequence segmentation method is: Assuming c(t) = [c ₁ , c ₂ , ..., c _N ], then N is an integer multiple of m. The sub-symbol sequence merging method is: given the sub-symbol segmentation scheme d, the i-th sub-symbol sequence is S(t, i) = [c(t, _{di-1, sum} +1); ...; c(t, di _{, sum} )].

For example, for 16-PPM, d=(2|2), c(t)=[ _c1 , _c2 , ..., _cN ] is first segmented into 4 bit sequences c(t, 1)=[ _c1 , _c5 , ..., cN _-3 ], c(t, 2)=[ _c2 , _c6 , ..., _cN-2 ], c(t, 3)=[ _c3 , _c7 , ..., cN _-1 ], c(t, 4)=[ _c4 , _c8 , ..., _cN ]. The sub-symbol sequence is:

That is, the i-th bit in c(t, 1) and the i-th bit in c(t, 2) correspond to the i-th sub-symbol in S(t, 1), that is, the i-th bit in c(t, 3) and the i-th bit in c(t, 4) correspond to the i-th sub-symbol in S(t, 2).

As a possible implementation, the sub-symbol interleaving module interleaves the bit sequence in the sub-symbol sequence according to the following interleaving principle to output an interleaved sub-symbol sequence:

The interleaving coefficients used for the bit sequences belonging to different sub-symbol sequences are different;

Bit sequences belonging to the same sub-symbol sequence use the same interleaving coefficient.

As an example, given a sub-symbol segmentation scheme d, the bit sequences are interleaved respectively. The bit sequences belonging to different sub-symbol sequences use different interleaving sequences, and the bit sequences belonging to the same sub-symbol sequence use the same interleaving sequence. The sequence obtained after interleaving c(t, i) is recorded as c _π (t, i), S _π (t, i) = [c _π (t, di _{-1, sum} +1); ...; c _π (t, di _{, sum} )].

For example, for 16-PPM, d = (2|2). Use interleaving coefficient 1 to perform bit interleaving on c(t, 1) and c(t, 2) to obtain _cπ (t, 1) and _cπ (t, 2); use interleaving coefficient 2 to perform bit interleaving on c(t, 3) and c(t, 4) to obtain _cπ (t, 3) and _cπ (t, 4) (interleaving coefficient 1 and interleaving coefficient 2 are different). This is equivalent to using interleaving sequence 1 to interleave S(t, 1) in units of sub-symbols to obtain _Sπ (t, 1); and using interleaving sequence 2 to interleave S(t, 2) in units of sub-symbols to obtain _Sπ (t, 2).

As a possible implementation method, the delay parameters are configured for the delay module based on the following principles:

After determining the number of sub-symbol sequences contained in the PPM modulation module and the codeword, the delay parameter satisfies the following conditions:

The delay parameter corresponding to the i-th sub-symbol sequence is a non-negative integer, and the minimum delay parameter among all delay parameters corresponding to all sub-symbol sequences is zero, and the maximum delay parameter is greater than zero and less than the number of sub-symbol sequences of the codeword, where 1≤i≤the number of sub-symbol sequences. For example, if a codeword contains 3 sub-symbol sequences, the value of the delay parameter can be 0, 1, 2, and the maximum value does not exceed 2.

The number of delay parameters is equal to the number of sub-symbol sequences included in the codeword, that is, one sub-symbol sequence corresponds to one delay parameter.

As an example, given a delay parameter δ of the delay module = [δ ₁ , δ ₂ , ..., δ _m′ ]. At time t, the delay module caches the sub-symbol sequence S _π (t, 1), ..., S _π (t, m′), and thereafter, the delay module outputs the sub-sequence S _π (t-δ ₁ , 1), ..., S _π (t-δ _m′ , m′). In other words, S _π (t, i) is input into the delay module at time t and is output by the delay module at time t+δ _i .

As an example, the method for determining the value of the delay parameter δ is as follows:

Given M-PPM and a sub-symbol partitioning scheme d=(d ₁ |d ₂ | ... |d _m′ ), the delay parameter δ=[δ ₁ , δ ₂ , ..., δ _m ] needs to satisfy the following conditions.

For 1≤i≤m, δ _i is a non-negative integer, and at least one of δ ₁ , δ ₂ , ..., δ _m must be 0 (min(δ)=0), but not all values can be zero (max(δ)>0).

The maximum delay parameter must be greater than zero and less than the number of sub-symbol sequences, 0＜max(δ)＜m′.

For example, for 16-PPM, d = (2|2), the delay parameter δ can be [0, 1] or [1, 0], but cannot be [0, 2] (it does not meet condition 2, consumes more cache but does not bring additional improvement to error performance). For 64-PPM, you can use {d = (2|2|2), δ = [0, 1, 2]}, {d = (2|2|2), δ = [0, 1, 1]}, {d = (3|3), δ = [0, 1]}, {d = (2|4), δ = [0, 1]} and other solutions.

For the convenience of expression, given a pair of sub-symbols + delay parameters {d, δ}, the transmission sequence can be abbreviated as Where D _i represents the sequence number of the subsequence with delay value i. If D _i contains subsequences belonging to different sub-symbols, square brackets [] are used to separate the sub-symbols. For example: {(2|2), [0, 1]} can be written as Example: {(2|2|2), [0, 1, 1]} can be written as

As a possible implementation method, the sub-symbol sequence merging module merges the delay module outputs c _π (t-δ ₁ , 1), ..., c _π (t-δ _m , m) at the t-th moment into a sequence c _δ (t). The segmentation method is: assuming c _π (t-δ _i , i) = [c _{i, 1} , c _{i, 2} , ...], the merged sequence is c _δ (t) = [c ₁ , 1 , c ₂ , 1 , ..., _{cm, 1} , c 1, ₂ , c _{2, 2} , ..., _{cm, 2} , ...].

As a possible implementation manner, the PPM modulation module performs M-PPM modulation on the sequence c _δ (t) to generate a PPM symbol sequence x(t).

It needs to be further explained that when the transmitter needs to transmit T groups of data u(1), u(2), ..., u(T), it actually needs to transmit T+δ _max groups of symbol sequences x(1), x(2) ..., x(T+δ _max ). _{δ max} =max(δ).

As a possible implementation, the PPM sub-symbol demodulation module determines the sub-symbol log-likelihood ratio using prior information and the received signal, specifically including:

For a sub-symbol sequence satisfying that a delay parameter is equal to a maximum delay parameter, a sub-symbol log-likelihood ratio is determined using a received signal;

For a sub-symbol sequence satisfying that the delay parameter is less than the maximum delay parameter, the sub-symbol log-likelihood ratio is determined jointly by using the received signal and the prior information.

For example, the receiving end needs to transmit T groups of data u(1), u(2), ..., u(T), and the detailed coding scheme of DSCPPM with a modulation bit number of M is as follows:

PPM sub-symbol demodulation module: When the transmitter transmits x(t), the receiver receives the corresponding signal y(t). Given the delay parameter δ = [δ ₁ , δ ₂ , ..., δ _m ], δ _max = max(δ). The PPM sub-symbol demodulator uses y(t) and prior information Γ ^a (S _π (t-δ _i , i)) to calculate the sub-symbol LLR Γ(S _π (t-δ _i , i)), i = 1, 2, ..., m′.

For a sub-symbol sequence S _π (t-δ _i , i) satisfying δ _i =δ _max , its sub-symbol LLR Γ(S _π (t-δ _i , i)) is calculated using only the received signal y(t) (without prior information).

For a subsymbol sequence S _π (t-δ _i , i) satisfying δ _i ＜δ _max , its subsymbol LLR Γ(S _π (t-δ _i , i)) is calculated using the received signal y(t) and prior information {Γ ^a (S _π (t-δ _j , j))|δ _j ＞δ _i }.

As a possible implementation, the reverse delay module stores all received sub-symbol LLRs in a buffer, and when {Γ(S _π (t, i))|i=1, 2, ..., m′} is collected in the buffer, it is output to the deinterleaver.

As a possible implementation manner, the sub-symbol deinterleaving module deinterleaves Γ(S _π (t, 1)), ..., Γ _π (S(t, m′)) respectively to obtain Γ(S(t, 1)), ..., Γ(S(t, m′)).

As a possible implementation, the decoding module inputs Γ(S(t, 1)), ..., Γ(S(t, m′)) and outputs a hard decision decoding result And sub-symbol feedback Γ ^a (S(t, 1)), ..., Γ ^a (S(t, m′)). Note: There can be many feedback methods, and several examples are listed below:

Feedback sub-symbol soft decision LLR (decoder external information LLR).

Feedback sub-symbol hard decision

In the case where u(t) includes CRC check bits: if the CRC check is passed (the system determines that the decoding is successful), a sub-symbol hard decision is sent; if it is not passed, a soft decision LLR is sent.

In the case where u(t) includes CRC check bits: if the CRC check is passed, a sub-symbol hard decision is sent; if it is not passed, all feedback values are set to 0 (no valid feedback).

The feedback LLRs are interleaved to obtain Γ ^a (S _π (t, 1)), ..., Γ ^a (S _π (t, m′)), which are buffered in the delay module.

As a possible implementation, the method for PPM sub-symbol demodulation is as follows:

No feedback information (only sub-symbol demodulation using received symbol y):

Sub-symbol set definition: Let Represents the collection of all PPM symbols. represents the set of symbols in all PPM symbols that satisfy the value of the i-th sub-symbol (S _i ) is j, that is:

Assuming that the PPM sub-symbol demodulator only has the symbol LLR Γ _C (y) from the channel, the calculation formula of the sub-symbol LLR is: The number of LLRs corresponding to a sub-symbol is 2 ^d (d is the sub-symbol bit length), that is, the value range of the sub-symbol is j∈{0, 1, ..., 2 ^d -1}. Example: Assume that the sub-symbol S _i contains 2 bits, then S _i has 4 LLR values: Γ(S _i ) = {Γ _{S(i), 0} , Γ _{S(i), 1} , Γ _{S(i), 2} , Γ _{S(i), 3} }.

Soft decision feedback information, assuming that the PPM sub-symbol demodulator has both the channel LLR Γ _C (y) and the prior sub-symbol The calculation formula of sub-symbol LLR (posterior) is: in The calculation formula of sub-symbol LLR (external information) is:

Example: For 16-PPM, d = (2|2), given C = 13, then

Hard decision feedback information, assuming that the PPM sub-symbol demodulator has both the channel LLR Pr{y|C} and the hard decision of the sub-symbol S _i The calculation formula of sub-symbol LLR (posterior) is: Note: When the sub-symbol demodulator is known Then the decoder no longer recalculates Γ _S(i),j , that is, k≠i.

As a possible implementation method, the feedback module includes a feedback sub-symbol interleaving unit and a feedback delay unit connected in sequence; the feedback sub-symbol interleaving unit interleaves the sub-symbol feedback sequence after receiving it, and outputs the interleaved sub-symbol feedback sequence; the delay parameter is configured in the feedback delay unit, the interleaved sub-symbol feedback sequence is input into the feedback delay unit at a preset time, and the delayed interleaved sub-symbol feedback sequence is output after the delay parameter is reached; in the next decoding cycle, the delayed interleaved sub-symbol feedback sequence and the received signal participate in the demodulation and decoding process together.

The following is a detailed description of the execution process of the receiving end with specific examples. It should be understood that the following examples are only examples and are not intended to be limiting.

Referring to FIG13, consider {d = (2|2|2), δ = [0, 1, 2]}. As shown in FIG13, the three sub-symbol sequences S(t, 1), S(t, 2), and S(t, 3) corresponding to c(t) are sent in x(t), x(t+1), and x(t+2), respectively. In other words, x(t) contains S(t, 1), S(t-1, 2), and S(t-2, 3). Therefore, when x(t+2) is received, the LLR of S(t, 3) can be directly demodulated; after x(t+1) is received and c(t-1) is decoded to obtain the feedback LLR of S(t-1, 3), the LLR of S(t, 2) can be demodulated; after x(t) is received and c(t-1) and c(t-2) are decoded to obtain the feedback LLR of S(t-1, 2) and S(t-2, 3), the LLR of S(t, 1) can be demodulated. The following are more detailed calculation steps:

Time t: At this time, the reverse delay module already has {Γ(S _π (t-2, i))|i＝1,2}; the feedback delay module already has Γ ^a (S _π (t-3,3)). Receive y(t). Calculate Γ(S _π (t-2,3)) using y(t). The reverse delay module collects {Γ(S _π (t-2, i))|i＝1,2,3}. Decoding yields {Γ ^a (S(t-2, i))|i=2, 3}. Note: Γ ^a (S(t-2, 1)) does not participate in the decoding of other sub-symbols, so the decoder does not calculate it. Calculate Γ(S _π (t-1, 2)) using y(t) and Γ ^a (S(t-2, 3)); calculate Γ(S _π (t-1, 1)) using y(t-1) and Γ ^a (S(t-2, 2)).

Time t+1: Receive y(t+1). Use y(t+1) to calculate Γ(S _π (t-1, 3)). The reverse delay module already has {Γ(S _π (t-1, i))|t＝1，2，3}. Decoding yields {Γ ^a (S(t-1,i))|i=2,3}. Use y(t), Γ ^a (S _π (t-1, 2)), Γ ^a (S _π (t-2, 3)) to calculate Γ (S _π (t, 1)); use y (t+1) and Γ ^a (S _π (t-1, 3)) to calculate Γ (S _π (t, 2)).

At time t+2: y(t+2) is received. Calculate Γ(S _π (t, 3)) using y(t+2). The reverse delay module already has {Γ(S _π (t, i))|t=1, 2, 3}. Decoding yields {Γ ^a (S(t,i))|i=2,3}. Use y(t+1), Γ ^a (S _π (t, 2)), Γ ^a (S _π (t-1, 3)) to calculate Γ (S _π (t+1, 1)); use y (t+2) and Γ ^a (S _π (t, 3)) to calculate Γ (S _π (t+1, 2)).

As an example, a PPM symbol is divided into sub-symbols of equal bit length, such as d = (2|2), d = (3|3), d = (2|2|2), etc. In this case, the grid diagram of (accumulator + sub-symbol with bit length d) is consistent with the grid of ^2d- APPM, so the decoding process of (convolutional code + accumulator + sub-symbol with bit length d) is equivalent to the decoding process of ^2d- SCPPM. Example: For 16-PPM, d = (2|2). The length of the codeword c(t) is N bits, which is equivalent to N/4 symbols and N/2 sub-symbols. The decoder combines the sub-symbols LLR Γ(S(t,1))=[Γ(S _1,1 ), Γ(S _1,2 ),..., Γ(S _1,N/2 )] and Γ(S(t,2))=[Γ(S _2,1 ), Γ(S _2,2 ),..., Γ(S _2,N/2 )] into Γ(S(t))=[Γ(S _1,1 ), Γ(S _2,1 ), Γ(S _1,2 ), Γ(S _2,2 ),..., Γ(S _1,N/2 ), Γ(S _2,N/2 )]. Where S _1,i = bi2de([c _1+4(i-1) , c _2+4(i-1) ]), S _2,i = bi2de([c _3+4(i-1) , c _4+4(i-1) ]). Then Γ(S(t)) is used as input for iterative decoding of (4-APPM+convolutional code) (see the grid diagrams in Figures 14 and 15).

Consider a more complex case: the PPM symbol is divided into sub-symbols of unequal bit lengths. In this case, the decoder needs to use a different APPM grid diagram. Example: for 32-PPM, d = (2|3). The length of the codeword c(t) is N bits, which is equivalent to N/5 symbols and 2N/5 sub-symbols. The decoder combines the sub-symbol LLRs Γ(S(t,1))=[Γ( _S1,1 ),Γ( _S1,2 ), ...,Γ( _S1,2N/5 )] and Γ(S(t,2))=[Γ( _S2,1 ),Γ( _S2,2 ), ...,Γ( _S2,2N/5 )] into Γ(S(t))=[Γ( _S1,1 ),Γ( _S2,1 ),Γ( _S1,2 ),Γ( _S2,2 ), ...,Γ( _S1,2N/5 ),Γ( _S2,2N/5 )], where S1 _,i =bi2de([ _c1+5(i-1) , _c2+5(i-1) ]), S2 _,i =bi2de([ _c3+5(i-1) ,c _4+5(i-1) , c _5+5(i-1) ]). Then, Γ(S(t)) is used as input for iterative decoding of (4/8-APPM+convolutional code). 4/8-APPM refers to that the APPM decoder uses a hybrid trellis diagram, using a 4-APPM trellis diagram for odd sub-symbols S _1,i and an 8-APPM trellis diagram for even sub-symbols S _2,i (see Figures 16 and 17).

Compared with the prior art, the delayed serial cascade pulse position modulation system provided by the present invention divides each codeword in the codeword sequence into multiple sub-symbol sequences using a transmitting terminal symbol sequence division module, and further interleaves the bit sequence in the sub-symbol sequence to obtain an interleaved sub-symbol sequence. Based on this, part of the interleaved sub-symbol sequence is delayed and modulated together with the sub-symbol sequence corresponding to the subsequent codeword. At this time, each PPM symbol sequence will contain multiple sub-symbol sequences from different codewords.

Moreover, at the receiving end, the received signal received at each moment will use the sub-symbol feedback sequence successfully decoded at the previous moment to assist in demodulation, and obtain the sub-symbol log-likelihood ratio. When the log-likelihood ratios of all sub-symbol sequences corresponding to all codewords are completely received, the receiving end decodes the code block and feeds back the soft decision or hard decision decoding results to the PPM sub-symbol demodulation module to assist in the demodulation of the subsequent received signal. In view of this, the delayed serial cascade pulse position modulation system provided by the present invention has better error performance.

The decoding error performance of DSCPPM is significantly better than that of SCPPM+single demodulation decoding scheme, and in the best case it can approach that of SCPPM+APPM decoding scheme. See Figures 18 and 19 for the reliability performance simulation results. See Figure 18, when M=16 and nb=0.1, SCPPM and APPM decoding are ideal. nb is the background noise, nb=0.1 indicates that 0.1 photons are received in each time slot; M=16 indicates that the modulation mode is 16PPM; there are 4 bits in each symbol, and these 4 bits are divided into 2 groups of 2-bit data streams by d(2,2); delta[0,1] represents the delay, 0 indicates no delay; 1 indicates that the delay is 1.

Referring to Figure 19, when M=64 and nb=0.1, nb is the background noise, and nb=0.1 indicates that O.1 photons are received in each time slot; M=64 indicates that the modulation mode is 64PPM; there are 6 bits in each symbol, and these 6 bits are divided into 2 groups of 3-bit data streams through d(3,3); delta[0,1] represents delay, 0 indicates no delay; 1 indicates that the delay is 1, and these 6 bits are divided into 3 groups of 2-bit data streams through d(2,2,2); delta[0,0,1] represents delay, 0 indicates no delay; 1 indicates that the delay is 1.

Furthermore, the key point of the present invention is to adopt a joint method of splitting the codeword into sub-symbol sequences, performing delay modulation after interleaving, and performing APPM decoding on the sub-symbols, and the joint method splits a high-order APPM equivalent into low-order APPMs by delay. Without changing the original encoding module of the existing serial cascade pulse position modulation system (SCPPM) and using APPM decoding, the decoding complexity of the inner code is significantly reduced compared with the existing SCPPM.

It needs to be further explained that the core idea of the DSCPPM scheme is to divide the symbol sequence of a codeword into sub-symbol sequences, and disperse the sub-symbol sequences in different time periods for PPM modulation, so that the PPM signal sent in each time period contains sub-symbol sequences of multiple codewords at the same time. Under this structure, the receiving end uses the decoding result of a codeword to assist in the demodulation of the sub-symbols of the subsequent codewords.

The embodiment of the present invention aims to propose a coding modulation structure and its corresponding sending and receiving process. Some of the designs can be modified, such as: the interleaving coefficient of the interleaving module, the sub-symbol segmentation scheme, and the delay scheme. Some of the calculation details can be replaced by different algorithms: such as the specific decoding algorithm used inside the decoder (from inputting sub-symbol LLR to giving and decoding results) and its hardware implementation, the calculation of feedback information, and the specific calculation of sub-symbol demodulation.

If you want to further reduce the computational complexity of the receiving end, you can also directly demodulate the PPM signal (directly calculate the bit LLR), and the decoder uses the bit LLR for direct decoding. DSCPPM + single bit demodulation and decoding has a loss in error performance compared to sub-symbol decoding, but it is still better than SCPPM + single bit demodulation and decoding. For multiple bit sequences with the same delay, SCPPM + multiple bit demodulation and decoding can also be used, that is, after using the PPM bit demodulation LLR of c(t) for decoding, use the feedback LLR of c(t) to re-demodulate the PPM bit, update the PPM bit demodulation LLR of c(t) and decode c(t) again, and repeat the process of c(t) PPM bit demodulation + c(t) decoding multiple times.

The error performance can be roughly ranked as follows: SCPPM+APPM>DSCPPM+subsymbol APPM>DBICM+SCPPM+single bit demodulation and decoding. For DSCPPM, the fewer splits, the better the error performance; when the number of subsymbols is the same, the more delayed moments, the better the performance. Example: For 64PPM, d=(3|3) is better than d=(2|2|2). For d=(2|2|2), δ=[0,1,2] is better than δ=[0,1,1].

For an ideal Poisson channel, the delay order of the subsequences does not affect the error performance. (For example, for 16-PPM, d = (2|2), δ = [0, 1] or δ = [1, 0], the error performance is consistent.) However, the actual system may not be an ideal Poisson channel, and the delay order of the subsequences may affect the error performance. Therefore, the embodiment of the present invention does not make specific provisions for the delay order.

In a second aspect, the present invention further provides a delayed serial cascade pulse position modulation method. The delayed serial cascade pulse position modulation system provided in the first aspect performs the delayed serial cascade pulse position modulation method, comprising the following steps:

S10. The transmitter outputs a PPM symbol sequence based on the original data sequence;

S20. The receiving end outputs a decoding result and a sub-symbol feedback sequence based on the PPM symbol sequence, wherein the sub-symbol feedback sequence is fed back to the transmitting end and participates in the demodulation and decoding process together with the received signal in the next demodulation and decoding cycle.

As a possible implementation manner, the transmitting end outputs a PPM symbol sequence based on the original data sequence, including:

S100. The encoding module receives the original data sequence and outputs a codeword sequence after encoding;

S101. The sending terminal symbol sequence segmentation module outputs a sub-symbol sequence based on the codeword sequence;

S102. The sub-symbol interleaving module receives the sub-symbol sequence corresponding to each codeword, interleaves the bit sequence in the sub-symbol sequence, and outputs the interleaved sub-symbol sequence;

S103. The delay module is configured with a delay parameter, the interleaved sub-symbol sequence is input into the delay module at a preset time, and the delayed interleaved sub-symbol sequence is output after the delay parameter is reached;

S104. The sub-symbol sequence merging module receives the delayed interleaved sub-symbol sequence at a preset time, and merges all the delayed interleaved sub-symbol sequences into a delayed codeword sequence;

S105. The PPM modulation module receives the delayed codeword sequence and performs M-PPM modulation, and outputs a PPM symbol sequence.

As a possible implementation manner, the receiving end outputs a decoding result and a sub-symbol feedback sequence based on the PPM symbol sequence, including:

S200. The PPM sub-symbol demodulation module is configured with prior information, the PPM sub-symbol demodulation module receives a received signal corresponding to the PPM symbol sequence, and the PPM sub-symbol demodulation module determines a sub-symbol log-likelihood ratio using the prior information and the received signal;

S201. The receiving terminal symbol sequence segmentation module receives the sub-symbol log-likelihood ratios, and first segments and then merges them into a sub-symbol log-likelihood ratio sequence; the reverse delay module caches the sub-symbol log-likelihood ratios, and when all the sub-symbol log-likelihood ratios are collected, they are output to the sub-symbol deinterleaving module;

S202. The sub-symbol deinterleaving module deinterleaves the sub-symbol log-likelihood ratio sequences respectively, and outputs the deinterleaved sub-symbol log-likelihood ratios;

S203. The decoding module outputs a decoding result and a sub-symbol feedback sequence based on the deinterleaved sub-symbol log-likelihood ratio;

S204. The sub-symbol feedback sequence is sent to the PPM sub-symbol demodulation module through the feedback module. In the next demodulation and decoding cycle, the sub-symbol feedback sequence and the received signal participate in the demodulation and decoding process together.

Compared with the prior art, the beneficial effects of the delayed serial cascade pulse position modulation method provided by the present invention are the same as the beneficial effects of the delayed serial cascade pulse position modulation system provided by the first aspect and/or any implementation of the first aspect, and will not be elaborated here.

In a third aspect, the present invention further provides a deep space optical communication system, wherein the encoding and decoding system adopted by the deep space optical communication system is the delayed serial cascade pulse position modulation system provided in the first aspect.

Compared with the prior art, the beneficial effects of the deep space optical communication system provided by the present invention are the same as the beneficial effects of the delayed serial cascade pulse position modulation system provided by the first aspect and/or any implementation method of the first aspect, and will not be elaborated here.

In the description of the above embodiments, specific features, structures, materials or characteristics may be combined in a suitable manner in any one or more embodiments or examples.

The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. The delay serial cascade pulse position modulation system is characterized by comprising a transmitting end and a receiving end;

the transmitting end comprises a coding module, a transmitting terminal symbol sequence segmentation module, a sub-symbol interleaving module, a delay module, a sub-symbol sequence merging module and a PPM modulation module which are sequentially connected;

the receiving end comprises a PPM sub-symbol demodulation module, a receiving terminal symbol sequence segmentation module, a reverse delay module, a sub-symbol de-interleaving module and a decoding module which are sequentially connected, and a feedback module which is connected with the decoding module and the PPM sub-symbol demodulation module;

the coding module receives an original data sequence, and outputs a codeword sequence after coding; the transmission terminal symbol sequence segmentation module outputs a sub-symbol sequence based on the codeword sequence; the sub-symbol interleaving module receives the sub-symbol sequence corresponding to each codeword, interleaves the bit sequences in the sub-symbol sequences, and outputs an interleaved sub-symbol sequence; the delay module is configured with delay parameters, the interleaving sub-symbol sequence is input into the delay module at preset time, and the delay interleaving sub-symbol sequence is output after the delay parameters are reached; the sub-symbol sequence combining module receives the delay interleaving sub-symbol sequences at the preset time and combines all the delay interleaving sub-symbol sequences into a delay codeword sequence; the PPM modulation module receives the delayed codeword sequence, performs M-PPM modulation and outputs a PPM symbol sequence;

The PPM sub-symbol demodulation module is configured with prior information, receives a received signal corresponding to the PPM symbol sequence, and determines a sub-symbol log-likelihood ratio by using the prior information and the received signal; the receiving terminal symbol sequence segmentation module receives the sub-symbol log-likelihood ratio, and segments and then merges the sub-symbol log-likelihood ratio into a sub-symbol log-likelihood ratio sequence; the reverse delay module caches the sub-symbol log-likelihood ratios, and outputs the sub-symbol log-likelihood ratios to the sub-symbol de-interleaving module after collecting all the sub-symbol log-likelihood ratios; the sub-symbol de-interleaving module is used for respectively de-interleaving the sub-symbol log-likelihood ratio sequences and outputting de-interleaved sub-symbol log-likelihood ratios; the decoding module outputs a decoding result and a sub-symbol feedback sequence based on the de-interleaving sub-symbol log-likelihood ratio; and the sub-symbol feedback sequence is sent to the PPM sub-symbol demodulation module through the feedback module, and the sub-symbol feedback sequence and the received signal participate in the demodulation and decoding process together in the next demodulation and decoding period.

2. The delayed serial concatenated pulse position modulation system of claim 1, wherein the feedback module comprises a feedback sub-symbol interleaving unit and a feedback delay unit connected in sequence; the feedback sub-symbol interleaving unit receives the sub-symbol feedback sequence and then interleaves the sub-symbol feedback sequence to output an interleaved sub-symbol feedback sequence; the feedback delay unit is configured with delay parameters, the interleaving sub-symbol feedback sequence is input to the feedback delay unit at preset time, and the interleaving sub-symbol feedback sequence is output after the delay parameters are reached; and in the next decoding period, the delay interleaving sub-symbol feedback sequence and the received signal participate in a demodulation and decoding process together.

3. The delayed serial concatenated pulse position modulation system of claim 1, wherein the transmit terminal symbol sequence segmentation module outputs a sequence of sub-symbols based on the sequence of codewords comprises:

the transmitting terminal symbol sequence segmentation module receives the codeword sequences, segments each codeword in the codeword sequences into bit sequences, and then merges the bit sequences into sub-symbol sequences.

4. The delayed serial concatenated pulse position modulation system of claim 1, wherein said sub-symbol interleaving module interleaves said bit sequences in said sub-symbol sequence in accordance with the following interleaving principle to output an interleaved sub-symbol sequence:

the interleaving coefficients used for the bit sequences belonging to different sub-symbol sequences are different;

the bit sequences belonging to the same sub-symbol sequence use the same interleaving coefficient.

5. A delayed serial concatenated pulse position modulation system as claimed in claim 3 wherein the delay module is configured with delay parameters based on the following principle:

after determining the number of sub-symbol sequences contained in the PPM modulation module and the code word, the delay parameter satisfies the following conditions:

For the i-th sub-symbol sequence, the corresponding delay parameter is a non-negative integer, the minimum delay parameter in all delay parameters corresponding to all sub-symbol sequences is zero, and the maximum delay parameter is greater than zero and less than the number of the sub-symbol sequences of the code word, wherein i is greater than or equal to 1 and less than or equal to the number of the sub-symbol sequences;

the number of the delay parameters is equal to the number of the sub-symbol sequences contained in the code word, namely one sub-symbol sequence corresponds to one delay parameter.

6. The delayed serial concatenated pulse position modulation system of claim 1, wherein the PPM sub-symbol demodulation module utilizes the prior information and the received signal to determine a sub-symbol log likelihood ratio, comprising:

for the sequence of sub-symbols satisfying a delay parameter equal to a maximum delay parameter, the sub-symbol log likelihood ratios are determined using the received signal;

for the sequence of sub-symbols satisfying a delay parameter less than a maximum delay parameter, the sub-symbol log likelihood ratios are jointly determined using the received signal and a priori information.

7. The delayed serial concatenated pulse position modulation method of claim 1, wherein the delayed serial concatenated pulse position modulation system of any of claims 1 to 6 performs a delayed serial concatenated pulse position modulation method comprising the steps of:

S10, the transmitting end outputs a PPM symbol sequence based on the original data sequence;

s20, the receiving end outputs a decoding result and a sub-symbol feedback sequence based on a received signal corresponding to the PPM symbol sequence, wherein the sub-symbol feedback sequence is fed back to the receiving end and participates in a demodulation decoding process together with the received signal in a next demodulation decoding period.

8. The serial concatenated pulse position modulation method of claim 7, wherein the transmitting side outputting the PPM symbol sequence based on the original data sequence comprises:

s100, an encoding module receives an original data sequence, and outputs a codeword sequence after encoding;

s101, a sending terminal symbol sequence segmentation module outputs a sub-symbol sequence based on the code word sequence;

s102, receiving the sub-symbol sequence corresponding to each code word by a sub-symbol interleaving module, interleaving the bit sequences in the sub-symbol sequences, and outputting an interleaved sub-symbol sequence;

s103, configuring delay parameters in a delay module, inputting the interleaving sub-symbol sequence into the delay module at preset time, and outputting the delay interleaving sub-symbol sequence after the delay parameters are reached;

s104, a sub-symbol sequence merging module receives the delay interleaving sub-symbol sequences at the preset moment and merges all the delay interleaving sub-symbol sequences into a delay codeword sequence;

S105, a PPM modulation module receives the delayed codeword sequence, carries out M-PPM modulation and outputs a PPM symbol sequence.

9. The serial concatenated pulse position modulation method of claim 7, wherein the receiving end outputting the decoding result and the sub-symbol feedback sequence based on the PPM symbol sequence comprises:

s200, configuring prior information in a PPM sub-symbol demodulation module, wherein the PPM sub-symbol demodulation module receives a received signal corresponding to the PPM symbol sequence, and determining a sub-symbol log-likelihood ratio by using the prior information and the received signal;

s201, a receiving terminal symbol sequence segmentation module receives the sub-symbol log-likelihood ratio, and segments and then merges the sub-symbol log-likelihood ratio into a sub-symbol log-likelihood ratio sequence; the reverse delay module caches the sub-symbol log-likelihood ratios, and outputs the sub-symbol log-likelihood ratios to the sub-symbol de-interleaving module after collecting all the sub-symbol log-likelihood ratios;

s202, respectively de-interleaving the sub-symbol log-likelihood ratio sequences by a sub-symbol de-interleaving module, and outputting de-interleaved sub-symbol log-likelihood ratios;

s203, the decoding module outputs a decoding result and a sub-symbol feedback sequence based on the de-interleaving sub-symbol log-likelihood ratio;

S204, the sub-symbol feedback sequence is sent to the PPM sub-symbol demodulation module through the feedback module, and the sub-symbol feedback sequence and the received signal participate in the demodulation and decoding process together in the next demodulation and decoding period.

10. A deep space optical communication system, characterized in that a coding and decoding system adopted by the deep space optical communication system is the delay serial cascade pulse position modulation system as claimed in any one of claims 1 to 6.