CN107645359B

CN107645359B - Quick decoding method and device for OvXDM system and OvXDM system

Info

Publication number: CN107645359B
Application number: CN201610584674.XA
Authority: CN
Inventors: 不公告发明人
Original assignee: Shenzhen Kuang Chi Hezhong Technology Ltd
Current assignee: Zhichanbang Beijing Network Technology Co ltd
Priority date: 2016-07-22
Filing date: 2016-07-22
Publication date: 2021-09-03
Anticipated expiration: 2036-07-22
Also published as: CN107645359A

Abstract

The application discloses a quick decoding method and device of an OvXDM system and the OvXDM system, because all state nodes and extension paths thereof do not need to be traversed in the decoding process, only a path corresponding to the minimum measure needs to be selected through measure sequencing for extension, the decoding complexity can be greatly reduced, the decoding efficiency is improved, the decoding complexity does not sharply increase along with the increase of the overlapping multiplexing times K or the coding branch K' like a traditional decoding scheme, and the contradictory requirements of the frequency spectrum efficiency, the decoding complexity and the decoding efficiency are solved.

Description

Quick decoding method and device for OvXDM system and OvXDM system

Technical Field

The present application relates to the field of signal processing, and in particular, to a fast decoding method and apparatus for an OvXDM system, and an OvXDM system.

Background

For an overlapping Multiplexing system, whether it is an overlapping Time Division Multiplexing (OvTDM) system, an overlapping Frequency Division Multiplexing (OvFDM) system, or an overlapping Code Division Multiplexing (OvCDM) system, conventional decoding requires constant access to nodes in a Trellis diagram (Trellis), and two memories are provided for each node, one for storing a relatively optimal path to the node, and one for storing a measure corresponding to the relatively optimal path to the node.

For the OvTDM system and the OvFDM system, since each node in the trellis diagram needs to be expanded in the decoding process, the number of nodes determines the complexity of decoding, and for the system with the number of overlapping times K and the modulation dimension M (M is an integer greater than or equal to 2), the number of nodes in a stable state in the corresponding trellis diagram is M^K-1Therefore, the decoding complexity increases exponentially with the number of overlaps K. In the OvTDM system and the OvFDM system, the spectral efficiency of the system is 2K/symbol, and therefore the greater the number of times of overlap K, the higher the spectral efficiency. Therefore, on one hand, the larger the overlap number K is, the better it is for the requirement of improving the spectral efficiency, and on the other hand, the smaller the overlap number K is, the better it is for the requirement of reducing the decoding complexity, and particularly, when the overlap number K is increased to a certain value, for example, K is greater than 8, the decoding complexity is increased sharply, the existing decoding method is difficult to meet the requirement of real-time decoding, and the spectral efficiency, the decoding complexity and the decoding efficiency form a pair of contradictory requirements.

Similarly, for an M-dimensional modulation OvCDM system with K' encoding branches, the number of nodes in steady state in the corresponding trellis diagram is M^K’-1Therefore, the decoding complexity increases exponentially with the number of encoding branches K'. In the OvCDM system, as large as possible number of coding branches K 'is required to increase the spectral efficiency, but at the same time, the decoding complexity increases sharply with the increase of K', so that the spectral efficiency, the decoding complexity and the decoding efficiency form a pair of contradictory requirements.

Disclosure of Invention

In order to solve the above problems, the present application provides a fast decoding method and apparatus suitable for an OvXDM system, and an OvXDM system.

According to a first aspect of the present application, the present application provides a fast decoding method for an OvXDM system, comprising the following steps:

step one, respectively calculating the measurement between all potential paths of the first r symbols and the first r received symbols in the received symbol sequence;

step two, ordering the measures obtained by calculation and storing the smaller R in the measures_nIndividual measures and their respective corresponding paths;

step three, expanding the path corresponding to the currently stored minimum measure, calculating the instantaneous measure between the expanded path and the corresponding receiving symbol, and adding each instantaneous measure and the accumulated measure corresponding to the previous moment to obtain the accumulated measure of each expanded path after the current moment is added;

step four, accumulating the measurement of each expansion path and the stored R which is not expanded_n-1 measures are ranked and the smaller of them R is stored_nIndividual measures and their respective corresponding paths;

step five, when the path corresponding to the currently stored minimum measure in the step three reaches the depth of the received symbol sequence after being expanded, calculating the instantaneous measure between the expanded path and the corresponding received symbol, comparing all the instantaneous measures, and taking the path corresponding to the minimum instantaneous measure as a decoding path; otherwise, repeating the third step and the fourth step;

wherein R is_nIs a positive integer and is smaller than the number of nodes of the trellis diagram corresponding to the OvXDM system.

Preferably, in step three, when each instantaneous measurement is added to the accumulated measurement corresponding to the previous time, the accumulated measurement is multiplied by the weighting factor and then added to the instantaneous measurement.

Preferably, the value of the weight factor is greater than 0 and less than or equal to 1.

According to a second aspect of the present application, there is provided a fast decoding apparatus suitable for an OvXDM system, comprising:

the first calculation module is used for respectively calculating the measurement between all potential paths of the first r symbols and the first r received symbols in the received symbol sequence;

the first sequencing module is used for sequencing the measures obtained by calculation;

R_ndistance memory and corresponding R_nPath memories for storing smaller R obtained in the first sorting module_nIndividual measures and their respective corresponding paths;

the expansion module is used for expanding the path corresponding to the currently stored minimum measure;

the second calculation module is used for calculating the instantaneous measurement between the path expanded by the expansion module and the corresponding receiving symbol, and adding each instantaneous measurement and the accumulative measurement corresponding to the previous time to obtain the accumulative measurement of each expanded path added at the current time;

a second sorting module for calculating the accumulated measure of each extended path obtained by the second calculation module and the stored R which is not extended_nOrdering by 1 measure, with the smaller R_nEach measure and its corresponding path are used to update the R_nDistance memory and corresponding R_nA value of the individual path memory;

a comparison output module; when the expansion module expands the path corresponding to the currently stored minimum measure to reach the depth of the received symbol sequence, the second calculation module calculates the instantaneous measure between the path expanded by the expansion module and the corresponding received symbol, the comparison output module compares all the instantaneous measures, and the path corresponding to the minimum instantaneous measure is used as a decoding path; otherwise, the expansion module, the second calculation module and the second sequencing module repeatedly work;

Preferably, the fast decoding apparatus of the OvXDM system further comprises a weighting factor module, configured to multiply the accumulated measure at the previous time by a weighting factor when the second calculating module adds the accumulated measure corresponding to the previous time to each instantaneous measure, so that the accumulated measure at the previous time is multiplied by the weighting factor before being added to the instantaneous measure.

Preferably, the value of the weighting factor module is greater than 0 and less than or equal to 1.

According to a third aspect of the present application, the present application provides an OvXDM system, comprising the above fast decoding apparatus suitable for high-weight OvXDM system.

The beneficial effect of this application is:

according to the quick decoding method and device of the OvXDM system and the OvXDM system, as all state nodes and the expansion paths thereof do not need to be traversed in the decoding process, and only the path corresponding to the minimum measure needs to be selected for expansion through measure sequencing, the decoding complexity can be greatly reduced, and the decoding efficiency is improved;

according to the quick decoding method and device applicable to the OvXDM system and the OvXDM system, the weight factor is introduced in the decoding process, so that the accumulated measure is multiplied by the weight factor and then added with the instantaneous measure, the reference of the node measure far away from the current node is gradually weakened along with the increase of the path depth, and the decoding accuracy is higher.

Drawings

Fig. 1 is a schematic flowchart of a fast decoding method applied to an OvXDM system in an embodiment of the present application;

fig. 2 is a schematic structural diagram of a fast decoding apparatus suitable for an OvXDM system in an embodiment of the present application;

fig. 3 is a schematic structural diagram of a transmitting end of an OvFDM system in the first embodiment of the present application;

FIG. 4 is a schematic diagram of data programming overlay of an OvFDM system in the first embodiment of the present application;

fig. 5(a) and (b) are schematic structural diagrams of a receiving end of an OvFDM system in the first embodiment of the present application;

FIG. 6 is a diagram of the input-output code tree of the OvFDM system in the first embodiment of the present application;

FIG. 7 is a schematic diagram of a trellis diagram extension of an OvFDM system in the first embodiment of the present application;

FIG. 8 is a trellis diagram of the decoding of the OvFDM system in the first embodiment of the present application;

FIG. 9 is a decoding diagram illustrating a fast decoding method according to a first embodiment of the present application;

fig. 10 is a schematic structural diagram of a transmitting end of an OvTDM system in a second embodiment of the present application;

FIG. 11 is a schematic diagram of data programming overlay of an OvTDM system in a second embodiment of the present application;

fig. 12(a) and (b) are schematic structural diagrams of a receiving end of an OvTDM system in a second embodiment of the present application;

FIG. 13 is a code tree diagram of the input-output of an OvTDM system in a second embodiment of the present application;

FIG. 14 is a schematic diagram of a trellis diagram extension of an OvTDM system in a second embodiment of the present application;

FIG. 15 is a decoded trellis diagram of an OvTDM system in a second embodiment of the present application;

FIG. 16 is a decoding diagram illustrating a fast decoding method according to a second embodiment of the present application;

fig. 17 is a schematic structural view of an OvCDM system in a third embodiment of the present application;

fig. 18 is a schematic structural diagram of an encoder of an OvCDM system in a third embodiment of the present application;

fig. 19 is a diagram illustrating a coding matrix of an OvCDM system in a third embodiment of the present application;

fig. 20 is a schematic structural diagram of a decoder of an OvCDM system in a third embodiment of the present application;

FIG. 21 is a trellis diagram of an OvCDM system according to a third embodiment of the present application;

FIG. 22 is a decoding diagram illustrating a fast decoding method according to a third embodiment of the present application.

Detailed Description

The present application will be described in further detail below with reference to the accompanying drawings by way of specific embodiments.

The basic idea of Viterbi decoding is to traverse all paths in a Trellis diagram (Trellis) diagram, and only the path with the minimum distance is reserved by comparing the distances between a plurality of branches reaching each state in the state transition process of the Trellis diagram and the correct path, so that the estimation of the correct path is obtained by comparison and screening, and the decoding is realized. However, as the overlap multiple K increases, the state number M of the Viterbi decoding increases exponentially by 2^ K, which causes a drastic increase in the computational complexity of the algorithm and makes it difficult to meet the requirement of real-time decoding.

The inventive concept of the present application resides in: starting from the origin of the trellis diagram, it has only M arriving nodes for an M-dimensional modulated system, regardless of the code constraint length K' or the number of overlaps K. Respectively calculating the instantaneous measures of the branches reaching the M nodes, selecting the minimum one from the instantaneous path measures of the M reaching nodes for expansion, respectively calculating the instantaneous measures of the expanded M branches, and recording the M reaching nodes. At most only Rn arriving nodes and path instantaneous measures thereof are reserved at each moment, each time, the nodes with the minimum instantaneous path measures are expanded, M-1 arriving nodes are added at each expansion, and the smaller R in the nodes is stored_nIndividual measures and their respective corresponding paths. . This continues until the data frame of the trellis has ended, and the path to the node with the smallest measure is the decision output. In summary, in the decoding process, all state nodes and the expansion paths thereof are not traversed, and the path corresponding to the minimum measure is selected through measure sequencing each time for expansion, so that the decoding complexity can be greatly reduced, and the decoding efficiency is improved.

The application discloses a quick decoding method (hereinafter referred to as a quick decoding method) suitable for a high-weight OvXDM system, and in the invention, high-weight is defined as a system with the superposition times K being more than or equal to 8. In one embodiment, the high-weight OvXDM system may be an OvTDM system, an OvFDM system, an OvCDM system, an OvSDM system, or an OvHDM system, wherein T, F, C, S, H represents a time domain, a frequency domain, a code division domain, a spatial domain, and a hybrid domain, respectively.

As shown in FIG. 1, the fast decoding method of the present application includes steps S01-S17. This will be explained in detail below.

Step S01, respectively calculating the metrics between all paths potentially covered by the first r symbols and the first r received symbols in the received symbol sequence. In a preferred embodiment, r is log_MR_nIs rounded down, wherein M represents the dimension of the system, and the value is an integer greater than or equal to 2. In one embodiment, the measure in this application represents the distance between two signals, defined as:

when p is 2, the euclidean distance is the true distance between two signals, which can truly reflect the distance between the actual signal and the ideal signal, and in one embodiment, the euclidean distance may be selected as the measure of the present application, and is defined as the euclidean distance in the present patent

Step S03 is to sort the measures calculated in step S01.

Step S05, storing the smaller R in the sequence in step S03_nIndividual measures and their respective corresponding paths. Wherein R is_nIs a positive integer and is smaller than the number of nodes of the trellis diagram corresponding to the heavy OvXDM system, e.g., in a preferred embodiment, R_nLess than M^K-1Wherein M represents the dimension of the system, and the value is an integer greater than or equal to 2.

And step S07, expanding the path corresponding to the currently stored minimum measure. For an OvTDM system, an OvFDM system and the like with M-dimensional modulation, M-dimensional expansion is carried out on the last node of a path corresponding to the currently stored minimum measurement.

Step S09, calculating an instantaneous measure between the expanded path and the corresponding received symbol.

And step S11, adding the instantaneous measures and the accumulated measures corresponding to the previous moment to obtain the accumulated measures of the extended paths added at the current moment. In a preferred embodiment, when each instantaneous measurement is added to the accumulated measurement corresponding to the previous time, the accumulated measurement at the previous time is multiplied by the weighting factor and then added to the instantaneous measurement, so that the reference of the node measurement farther from the current node is gradually weakened as the path depth increases, and the decoding accuracy is higher. In a preferred embodiment, the weighting factor is greater than 0 and less than or equal to 1.

Step S13, accumulating the measure of each expansion path and the stored R which is not expanded_n-1 measure to rank.

Step S15, storing the smaller R_nIndividual measures and their respective corresponding paths.

Step S17, when the path corresponding to the minimum measure stored at present is expanded in step S07 and reaches the depth of the received symbol sequence, step S19 is performed to calculate the instantaneous measure between the expanded path and the corresponding received symbol, and compare the instantaneous measures, and the path corresponding to the minimum measure is used as the decoding path; otherwise, repeating the steps S07-S15.

As described above, in the decoding process, only the path corresponding to the currently stored minimum measure is expanded each time, and the corresponding instantaneous measure and accumulated measure are calculated, without traversing all the state nodes and the expansion paths thereof, so that the decoding complexity can be greatly reduced, the decoding efficiency is improved, the decoding complexity does not increase sharply with the increase of the number of overlapping multiplexing times K like the conventional decoding scheme, and the contradictory requirements of the spectrum efficiency, the decoding complexity and the decoding efficiency are solved.

Correspondingly, the application also providesAs shown in fig. 2, the fast decoding apparatus of the present application includes a first calculating module 01, a first ordering module 03, and a first decoding module R_n A distance memory 05, R_nThe path memory 07, the expansion module 09, the second calculation module 11, the second sorting module 13 and the comparison output module 15, in a preferred embodiment, the fast decoding apparatus of the present application may further include a weight factor module 17.

This will be explained in detail below.

The first computation module 01 is configured to compute measures between potentially all paths of the first r symbols and the first r received symbols in the received symbol sequence, respectively. In a preferred embodiment, r is log_MR_nIs rounded down, wherein M represents the dimension of the system, and the value is an integer greater than or equal to 2.

The first sorting module 03 is configured to sort the calculated measures.

R_nA distance memory 05 and corresponding R_nThe path memories 07 are each used to store the smaller R values obtained in the first sorting module 03_nIndividual measures and their respective corresponding paths. In a preferred embodiment, R_nLess than M^K-1Wherein M represents the dimension of the system, and the value is an integer greater than or equal to 2.

The expansion module 09 is configured to expand a path corresponding to a minimum value in the currently stored measurement.

The second calculating module 11 is configured to calculate instantaneous metrics between the path expanded by the expanding module 09 and corresponding received symbols in the received symbol sequence, and add each instantaneous metric and the accumulated metric corresponding to the previous time to obtain the accumulated metric of each expanded path added at the current time. In a preferred embodiment, the weight factor module 17 is configured to multiply the accumulated measure at the previous time by the weight factor when the second calculation module 11 adds the accumulated measure corresponding to the previous time, so that the accumulated measure at the previous time is multiplied by the weight factor and then added to the instantaneous measure. The weight factor module 17 is introduced to gradually weaken the reference of node metrics farther from the current node as the path depth increases, so that the decoding accuracy is higher. In a preferred embodiment, the weighting factor is greater than 0 and less than or equal to 1.

The second sorting module 13 is configured to perform the cumulative measure of each expanded path calculated in the second calculating module 11 and the stored remaining R that is not expanded_nOrdering by 1 measure, with the smaller R_nThe individual measures and their respective corresponding paths are used to update the R_nA distance memory 05 and corresponding R_nThe value of the individual path memory 07.

When the expansion module 09 expands the currently stored path corresponding to the minimum measure to the depth of the received symbol sequence, the second calculation module 11 calculates the instantaneous measure between the path expanded by the expansion module 09 and the corresponding received symbol, the comparison output module 15 compares the instantaneous measures, and takes the path corresponding to the minimum instantaneous measure as a decoding path; otherwise, the expansion module 09, the second calculation module 11 and the second sorting module 13 repeat the work.

The present application also discloses an OvXDM system, which includes the above fast decoding apparatus suitable for the OvXDM system, and in an embodiment, the OvXDM system may be an OvTDM system, an OvFDM system, an OvCDM, an OvSDM, or an OvHDM system.

The present application is further illustrated by the following examples.

Example one

The present embodiment does not take the OvFDM system as an example for explanation.

As shown in fig. 3, the OvFDM system transmitting end encodes a frequency domain signal according to a certain rule, and then converts the frequency domain signal into a time domain signal, i.e., performs inverse fourier transform, and then transmits the signal. Specifically, an initial envelope waveform is generated according to design parameters; then shifting the initial envelope waveform on a frequency domain according to the overlapping multiplexing times according to a preset frequency spectrum interval to obtain each subcarrier envelope waveform; then multiplying the input data sequence with the respective corresponding sub-carrier envelope waveform to obtain the modulation envelope waveform of each sub-carrierShaping; and finally, converting the complex modulation envelope waveform on the frequency domain into a complex modulation envelope waveform on a time domain for sending, wherein the frequency spectrum interval is a subcarrier frequency spectrum interval delta B, the subcarrier frequency spectrum interval delta B is B/K, B is the bandwidth of the initial envelope waveform, and K is the overlapping multiplexing times. The data sequence encoding overlap-add procedure is shown in fig. 4. As shown in fig. 5, the OvFDM system receiving end receives a time domain signal through an antenna, and if the received signal is to be decoded, the time domain signal needs to be converted into a frequency domain signal, that is, the fourier transform is performed before the signal can be processed. Specifically, symbol synchronization is firstly formed on a received signal in a time domain; then sampling and quantizing the received signal of each symbol time interval to convert the received signal into a received digital signal sequence; converting the time domain signal into a frequency domain signal, and segmenting the frequency domain signal at a spectrum interval delta B to form an actual received signal segmented spectrum; then, a one-to-one correspondence relationship between the received signal spectrum and the transmitted data symbol sequence is formed, and finally, the data symbol sequence is detected according to the one-to-one correspondence relationship. The inverse Fourier transform and Fourier transform in the OvFDM system both relate to the setting of the number of sampling points, the number of the sampling points of the inverse Fourier transform and the Fourier transform in the OvFDM system are consistent, and the value of the sampling points is 2ⁿAnd n is a positive integer.

A 2-dimensional modulation OvFDM system with K overlapping times, i.e. M-2, with 2 nodes in the trellis^K-1The distance memory required in the conventional decoding algorithm is 2^K-12, path memory is also^K-1As shown in fig. 6, it is an input-output relationship code tree diagram corresponding to the system when the received signal length N is 4 and the number of times of overlap K is 3, and it can be seen from the diagram that the number of states of the code tree diagram is 2^K-1Correspondingly, as shown in fig. 7, the number of nodes in the corresponding trellis is also 4, and when the decoding state is fully expanded, each decoding step has 8 paths in total. In this embodiment, the distance memory 05 is selected as R_nThe path memory 07 is also R_nWherein R is_nLess than 2^K-1Thereby reducing the complexity of decoding, in one embodiment, let r be log₂R_nRounded down. For a two-dimensional modulated OvFDM system, it has only two arriving nodes, regardless of the overlap factor K, from the origin of the trellis diagram. Respectively calculating the instantaneous measures of the branches reaching the two nodes, selecting the minimum one from the instantaneous path measures of the two reaching nodes for expansion, respectively calculating the instantaneous measures of the two branches for expansion, and recording the two reaching nodes. Retaining at most only R at each moment_nThe strip arriving nodes and their path instantaneous measures are each time extended from the node with the smallest instantaneous path measure, one arriving node is added each time the extension is made, and the paths with larger instantaneous path measures and their measures are all discarded. This continues until the data frame of the trellis has ended, and the path to the node with the smallest instantaneous path measure is the decision output. The decoding step is described in detail below, in this embodiment, R is not allowed to be_n＝2^r。

It should be noted that, in the following embodiments, the description is made by using a low overlap with K being 3, only for the purpose of making clear the specific schemes of the present invention, and the present invention is more suitable for a system with a high number of overlaps, for example, K being 8, etc., and the optimization effect thereof is more obvious as the number of overlaps increases.

(1) The initial first r symbol paths and corresponding measures are determined.

The symbol sequence of length r is represented as

For 2-dimensional modulation, there is a total of 2^rThe possible combination information is in the form of

Thus obtaining 2^rMatrix of dimension r, denoted msg (2)^rR), where each row represents a set of symbol sequences of length r. And respectively calculating the instantaneous measurement of each group of symbol sequence and the first r received symbols, and defining as:

wherein v is_rFor the first r received symbols, x_r,jFor the window function of the OvFDM system,

is msg (2)^rThe (r-j) th symbol in the ith row of r). At this point, the decoding reaches the r-th node, including 2^rDecoding paths U and 2^rAnd (4) instantaneous measure d, wherein the node depth of each decoding path is r. The instantaneous measures and the corresponding decoding paths are stored in a distance memory 05 and a path memory 07, respectively.

(2) Find the minimum and maximum of the measure.

2 obtained in (1) is^rThe instantaneous measures are sequentially ordered, and the indexes idx corresponding to the minimum measure and the maximum measure are respectively found out_min、idx_maxThe instantaneous measure corresponding to the minimum measure index is

And adding 1 to the index node depth corresponding to the minimum measure, wherein the node depth is changed into r + 1.

(3) Extension node

The decoding path corresponding to the minimum measure index is expanded to be 1 branch and-1 branch, namely the decoding path is changed into

And

the extended decoding path has 2^r+1, but we only need to keep 2^rA path, therefore, the decoding path corresponding to the maximum measure index found in (2) is deleted, and the decoding branch is extended to 1

Stored in index idx_maxIn the memory (may also be replaced by a-1 branch), a-1 branch

Continue to store at index idx_minIn the memory of (2). When only idx exists in the decoding path_min、idx_maxThe depth of the node corresponding to the index is r +1, and the depths of the nodes of the other indexes are still r. Node state transition reference is made to figure 8.

(4) Computing cumulative measures

Respectively solving instantaneous measures of the two expanded paths and the received symbols, and calculating a measure formula in the formula (2) to obtain measures d 'of the two paths'_r+1、d”_r+1And adding the instantaneous metrics to the instantaneous metrics corresponding to the r nodes to obtain an accumulated metric. In the addition process, a weight factor alpha can be introduced, the value of the weight factor alpha is a number which is greater than 0 and less than or equal to 1, the specific numerical value depends on the system requirement, the purpose of doing so is to gradually weaken the node measure reference which is farther away from the current node along with the increase of the node depth, and the addition process can be expressed as follows:

the obtained accumulated measures are respectively stored in the corresponding idx in (3)_min、idx_maxIn a memory.

And (4) screening and expanding the rest symbol sequences in the same way as the ways from (2) to (4) until the depth of the node reaches the depth N of the symbol sequence, and comparing the instantaneous measures after the last symbol expansion, wherein the smaller is the decoding path finally output.

From the above steps, it can be seen that the fast decoding method proposed in this patent only needs to perform R for each time_nThe extension is carried out on each node, and the classical method needs to carry out the extension on 2^K-1The next node is expanded, when K is larger, R_nCan take relatively small value and satisfy R_n<2^K-1Therefore, for the decoding sequence with the length of N, the decoding complexity is greatly reduced.

The present embodiment will be described below by way of a practical example.

Since the decoding complexity is high when the number of overlapping times K is large, and it is difficult to completely express the data frame length N by using characters if the data frame length N is too large, the embodiment selects a slightly smaller value of K and N as an example, and aims to describe a specific implementation scheme of fast decoding.

Taking the OvFDM system of two-dimensional modulation with the overlap number K being 5 as an example, it performs modulation using a rectangular multiplexing window H being { 11111 }. It should be noted that the fast decoding grammar of the present application is applicable to various multiplexing window functions. Since the number of overlapping times K is 5, the trellis is totally 2 when completely expanded^K-1For 16 nodes, R is selected in this example_nIs 4, so r is log₂R_nThe value of (a) is rounded down to 2, it should be noted that in the fast decoding method of the present application, R is_nIs only required to be less than 2^K-1The decoding complexity can be reduced.

The transmission code sequence is not made x_iAfter being modulated and coded by an OvFDM system, a receiving end is subjected to synchronization and channel estimation equalization processing, and is converted into a frequency domain signal to obtain a receiving sequence y { + 1-1-1 + 1-1 +1 +1 + 1-1 }, wherein the receiving sequence y is obtained_i{ + 10-10-1-1 +1 +1 +1 +1 }. Received sequence y according to the fast decoding method described above_iDecoding is performed, and the decoding path is as shown in fig. 9, and finally, a correct decoding result is obtained.

In the embodiment, by adopting a rapid decoding method, access nodes in a trellis diagram are screened in the decoding process, the accumulated measure is calculated by combining with a weight factor, the path with the minimum measure is selected for expansion, and the optimal decoding path is screened finally, so that the complexity and the calculated amount of system design are reduced by using the method in the decoding process of a high-weight OvFDM system, and the method has a low error rate. In addition, in the ordinary OvFDM system, when the accumulated measure is calculated, the measure before the current node and the instantaneous measure are directly added, and the accumulated measure is larger and larger with the increase of the decoding depth, which has higher requirements for the distance memory. In practice, however, in an OvFDM system with the overlap number of K, the mutual overlap of symbols only exists in the first K-1 symbols and the last K-1 symbols associated with the current symbol, and when the overlap number exceeds the range, the overlap of the current symbol will not be affected, so that the measurement of the overlap has smaller and smaller weight for judging the path of the current node. Based on the reason, when the accumulated measure is calculated, the weight results of the instantaneous measure and the previous node measure are added, the accumulated measure is always kept in a certain range along with the increase of the decoding depth, the requirement on a distance memory is reduced, and the final decoding path has higher reliability.

It should be noted that, in this embodiment, only the path with the smallest measurement is expanded each time, that is, only one path is expanded each time, so that the depths of the paths stored at each time are different, and therefore, when the signal-to-noise ratio is low, a node and a path repeatedly fall back in the process of screening and expanding. In addition, because the length of the data to be decoded is longer, and the accumulated distance is larger and larger along with the deepening of the decoding depth, if the system decodes and outputs all the data after the decoding is completed, system resources are consumed, and therefore a better processing method is adopted for the storage capacity and the distance of the path. Generally, the storage length of the selected path is 4K-5K, and if the path memory is full and the decoding decision output is not forced to be output, the initial nodes with the same path are output first. In addition, the accumulated measure is larger and larger as the decoding depth is deeper, the accumulated measure can be stored as a relative measure, namely, a reference measure is defined, the value of the reference measure is determined according to different systems, and the distance storage records the relative value of the second measure of each path relative to the reference measure, and the relative measure is compared when the optimal path is screened.

Example two

The present embodiment will not be described by taking the OvTDM system as an example.

As shown in fig. 10, for the transmitting end of the OvTDM system, an initial envelope waveform in a time domain is generated according to design parameters; then shifting the initial envelope waveform on the time domain according to the overlapping multiplexing times and preset time intervals to obtain the offset envelope waveform at each moment; multiplying the input data sequence by the offset envelope waveform at each moment to obtain a modulation envelope waveform at each moment; and then, the modulation envelope waveforms at all the moments are superposed on a time domain to obtain complex modulation envelope waveforms on the time domain to be sent, wherein the time interval is delta T, the delta T is T/K, T is the time domain width of the initial envelope waveform, and K is the overlapping multiplexing times. The data sequence encoding and superimposing process is shown in fig. 11. As shown in fig. 12, the OvTDM system receiving end forms a received digital signal sequence for the received signal in each frame, and then performs detection on the formed received digital signal sequence to obtain the decision of the modulated data modulated on all symbols in the frame length.

A 2-dimensional modulation OvTDM system with overlap number K, i.e. M2, and the number of nodes of the trellis is 2^K-1The distance memory required in the conventional decoding algorithm is 2^K-12, path memory is also^K-1As shown in fig. 13, the code tree diagram corresponds to input and output when the length N of one received signal is 4 and the number of times of overlap K is 3, and it can be seen that the number of states of the code tree diagram is 2^K-1Correspondingly, as shown in fig. 14, the number of nodes in the corresponding trellis diagram is also 4, and when the decoding state is fully expanded, each decoding step has 8 paths in total. In this embodiment, the distance memory 05 is selected as R_nThe path memory 07 is also R_nWherein R is_nLess than 2^K-1Thereby reducing the complexity of decoding, in one embodiment, let r be log₂R_nRounded down. For a two-dimensional modulated OvTDM system, it has only two arriving nodes, regardless of the overlap factor K, starting from the origin of the trellis diagram. Respectively calculating the instantaneous measures of the branches reaching the two nodes, selecting the minimum one from the instantaneous path measures of the two reaching nodes for expansion, respectively calculating the instantaneous measures of the two branches for expansion, and recording the two reaching nodes. Retaining at most only R at each moment_nThe strip arriving nodes and their path instantaneous measures are each time extended from the node with the smallest instantaneous path measure, one arriving node is added each time the extension is made, and the paths with larger instantaneous path measures and their measures are all discarded. This continues until the data frame of the trellis has ended, and the path to the node with the smallest instantaneous path measure is the decision output. The decoding step is described in detail below, in this embodiment, R is not allowed to be_n＝2^r。

(1) The initial first r symbol paths and corresponding measures are determined.

The symbol sequence of length r is represented as

wherein v is_rFor the first r received symbols, x_r,jFor the window function of the OvTDM system,

(2) Find the minimum and maximum of the measure.

(3) Extension node

And

Continue to store at index idx_minIn the memory of (2). When only idx exists in the decoding path_min、idx_maxThe depth of the node corresponding to the index is r +1, and the depths of the nodes of the other indexes are still r. Node state transition refer to fig. 15.

(4) Computing cumulative measures

Respectively solving instantaneous measures of the two expanded paths and the received symbols, and calculating a measure formula in the formula (2) to obtain measures d 'of the two paths'_r+1、d”_r+1And adding the instantaneous metrics to the instantaneous metrics corresponding to the r nodes to obtain an accumulated metric. In the addition process, a weight factor alpha can be introduced, the value of the weight factor alpha is 0-1, the specific numerical value depends on the system requirement, the purpose of doing so is to gradually weaken the node measurement reference far away from the current node along with the increase of the node depth, and the addition process can be expressed as follows:

The present embodiment will be described below by way of a practical example.

Since the decoding complexity is high when the number of overlapping times K is large, and it is difficult to express the data frame length N by using a specific embodiment if the data frame length N is too large, the embodiment takes a slightly smaller value of K and N as an example, and aims to describe a specific implementation scheme of fast decoding.

Taking the OvTDM system of two-dimensional modulation with the number of overlapping times K being 5 as an example, it performs modulation with a rectangular multiplexing window H being { 11111 }. It should be noted that the fast decoding grammar of the present application is applicable to various multiplexing window functions. Since the number of overlapping times K is 5, the trellis is totally 2 when completely expanded^K-1For 16 nodes, R is selected in this example_nIs 4, so r is log₂R_nThe value of (a) is rounded down to 2, it should be noted that in the fast decoding method of the present application, R is_nIs only required to be less than 2^K-1The decoding complexity can be reduced.

The transmission code sequence is not made x_iAfter OvTDM system modulation coding, the receiving end obtains a receiving sequence y after synchronization and channel estimation equalization processing_i{ + 10-10-1-1 +1 +1 +1 +1 }. Received sequence y according to the fast decoding method described above_iDecoding is performed, and the decoding path is as shown in fig. 16, and a correct decoding result is finally obtained.

In the embodiment, by adopting a rapid decoding method, access nodes in a trellis diagram are screened in the decoding process, the accumulated measure is calculated by combining with a weight factor, the path with the minimum measure is selected for expansion, and the optimal decoding path is screened finally, so that the complexity and the calculated amount of system design are reduced by using the method in the decoding process of a high-weight OvTDM system, and the method has a low error rate. In addition, in the common OvTDM system, when the accumulated measure is calculated, the measure in front of the current node and the instantaneous measure are directly added, and the accumulated measure is larger and larger along with the increase of the decoding depth, so that the common OvTDM system has higher requirements on a distance memory. In practice, however, in an OvTDM system with the overlapping number of K, the mutual overlapping of symbols only exists in the first K-1 symbols and the last K-1 symbols associated with the current symbol, and the overlapping of the current symbol after exceeding the range does not affect the overlapping of the current symbol, so that the measurement of the overlap has smaller and smaller weight for judging the path of the current node. Based on the reason, when the accumulated measure is calculated, the weight results of the instantaneous measure and the previous node measure are added, the accumulated measure is always kept in a certain range along with the increase of the decoding depth, the requirement on a distance memory is reduced, and the final decoding path has higher reliability.

EXAMPLE III

The present embodiment will not be described by taking the OvCDM system as an example. The core of overlapping code division multiplexing of the OvCDM system is overlapping and multiplexing, with the aim of improving the spectral efficiency of the communication system. The OvCDM system generalizes a convolutional coding coefficient to a generalized convolutional coding model of a complex field, generates a constraint relationship through symbol overlapping, and has a system structure diagram as shown in fig. 17 and a corresponding encoder structure as shown in fig. 18, wherein main parameters include a coding branch number K' and a coding constraint length L. The key of the OvCDM system is that the coding matrix, i.e. the convolutional spreading factor, is required to satisfy the linear relationship, and at this time, the input sequence and the output sequence are in one-to-one correspondence, so that decoding without error can be performed theoretically, and generally, a computer searches all matrixes with larger measurement values as the coding matrix, and the arrangement of the coding matrix is as shown in fig. 19.

To illustrate the fast decoding method of the present embodiment, the encoding process of the OvCDM system is first given.

(1) Converting the data to be transmitted into K' path sub-data stream through serial-parallel conversion, and recording the data stream on the ith path as u_i＝u_i,0u_i,1u_i,2… are provided. E.g. when K' is 2, u₀＝u_0,0u_0,1u_0,2…，u₁＝u_1,0u_1,1u_1,2…，

(2) Sending each path of data into a shift register for weighted superposition, wherein the weighting coefficient of the ith path is b_i＝b_i,0b_i,1b_i,2…, which is a complex vector.

(3) Adding and outputting the signals to obtain the final output c ═ c of the OvCDM encoder₀c₁c₂…，

The code rate of OvCDM is

Where n is the sub-stream lengthAnd (4) degree. When n is long, the code rate loss caused by shift register tailing is negligible, and r is obtained_OVCDM≈k。

The code rate of the traditional binary domain convolutional coding model is generally less than 1, which can cause the loss of spectral efficiency. While the rate of convolutional coding in the complex domain of OvCDM is equal to 1, single-path convolutional coding expansion does not cause loss of spectral efficiency, and extra coding gain is also added.

After receiving the signal, the receiving end performs synchronization, channel estimation and digitization processing on the signal, and then decodes the processed data. The core of the decoding algorithm is to determine the optimal decoding path by calculating the measure of the received signal and the ideal state, and using the path memory and the measure to obtain the final detection sequence, and the sequence detection process block diagram is shown in fig. 20.

The above is the case of the OvCDM system in the prior art. The fast decoding method in this embodiment includes the following specific steps.

The number of coding branches of the OvCDM system is K' branches and the coding constraint length L, the dimension of a coding output vector is N, and two-dimensional modulation is adopted, so that the code elements are +1 and-1. Since the number of encoding branches is K' and the length of encoding constraint is L, the number m of nodes reachable by each state is 2^K’In total comprise 2^K’A number of states S, so that the number of nodes reachable in all states is 2^2K’. In this embodiment, the distance memory 05 is selected as R_nThe path memory 07 is also R_nWherein R is_nLess than 2^K’-1Thereby reducing the complexity of decoding, in one embodiment, let r be log₂R_nRounded down.

(1) Initial state:

let the path metric of the initial node state (l ═ 0) be d_0,0＝0。

(2) Determining initial first r symbol paths and corresponding measures

The symbol sequence of length r is represented as

For 2 dimensionsModulation, in total 2^rThe possible combination information is in the form of

wherein v is_rFor the first r received symbols, x_r,jIs a coding matrix of the OvCDM, and,

(3) Calculating node measure:

each node comprises S states in total, and the measurement of the I-th node is calculated by calculating all m ideal signal symbols and received signal sequences transferred from the previous state to the state

Measure d between_s,m(l, l +1) represented by

(4) Calculating an accumulation measure:

the Euclidean distance d of each state S of the current node_s,mMeasure d of (l, l +1) and their respective starting states S_s',l-1And adding to form the cumulative Euclidean distance of the new m paths. In the addition process, a weight factor alpha can be introduced, the value is a number of 0-1, the specific numerical value depends on the system requirement, and the purpose of doing so isThe node measure references farther from the current node are gradually weakened as the depth of the node increases.

(5) Path screening:

sequencing the accumulated Euclidean distances, selecting the path with the minimum accumulated Euclidean distance to expand the path, expanding m branches, sequentially calculating the current node measurement of the m branches, reserving the path with the minimum measurement, and discarding the rest (m-1) paths, wherein the memory contains R in common at the moment_nThe strip path and its corresponding measure.

(6) Final path determination:

repeating the steps (3) to (5) until the decoding is finished, wherein R is reserved in the memory_nAnd the Euclidean distance of the strip path and the path corresponding to the strip path, wherein the path with the minimum Euclidean distance is the decoding result.

In this embodiment, the number of reserved paths R_nDetermining the information retained in the decoding process, the number of paths to be discarded is 2 for the OvCDM system with the number of encoding branches K^2K’-R_nThus R_nThe smaller the decoding complexity. But R is_nCannot be infinitely small, R_nThe smaller the decoding performance loss, the higher the signal-to-noise ratio under the same bit error rate condition. Therefore, it is necessary to select an appropriate R according to the actual system and channel_nAnd the decoding complexity is reduced, and meanwhile, the performance loss of decoding is ensured to be small. General selection of R_nHas a value of 2 or more^L(K’-4)And is less than or equal to 2^L(K’-2)At the moment, the decoding performance can be ensured, and the decoding complexity is greatly reduced.

The following is a description of an example.

Because the coding complexity is high when the number of coding branches K 'and the coding constraint length L are large, and it is difficult to express the data frame length N by a specific embodiment if the data frame length N is too large, the embodiment selects slightly smaller values of K', L, and N as an example, and aims to describe a specific implementation scheme of fast coding.

In this case, the input data stream is u { +1, -1, -1, -1, -1, +1, -1, +1, +1, -1, -1, -1, -1, +1, -1, and 1, K ═ 2, L { [ 2 ], R { [ 1 ], [ 2 ], and [ 2 ], [ R ], [ 1 ], [ 2 ]_n4, coding matrix

For example, the decoding process of the OvCDM system is described, and the trellis diagram corresponding to the OvCDM system under the parameter design is shown in FIG. 21.

(1) Encoding process

First, for K' ═ 2, the input data stream u is converted into two paths, corresponding to:

u₁＝{+1,-1,-1,-1,+1,+1,-1,-1}

u₂＝{-1,-1,+1,+1,-1,-1,-1,+1}

the convolution coefficient for each way is expressed as: b₀＝[1 j],b₁＝[j 1]Referring to the coding structure of fig. 18 and the trellis diagram of fig. 21, the coded output is c ═ 1-j, -2, -2,0,2-2j,0, -2, -2 }. The encoding implementation process of the OvCDM system can also adopt other modes, and the invention of the present application lies in the decoding process rather than the encoding process.

(2) Decoding

After signal synchronization, channel estimation and digitization processing, the receiving end obtains a signal sequence, for convenience of description, an ideal state is assumed, at this time, the received signal sequence is {1-j, -2, -2,0,2-2j,0, -2, -2}, and decoding is performed by using the received signal:

the first symbol received is 1-j, which is measured separately from the ideal four states (1, 1), (1, -1), (-1,1), (-1, -1), and ordered to find the minimum measure, which corresponds to a path of (1, -1). And storing the four measures and the corresponding paths into a memory.

And then processing the 2 nd symbol-2, expanding the path (1, -1) with the minimum measure, expanding four branches from the current state, wherein the four branches are respectively (1, -1,1, 1), (1, -1,1, -1), (1, -1, -1, -1,1) and (1, -1, -1, -1), evaluating the current symbol with the four ideal paths respectively, reserving the path (1, -1, -1, -1) corresponding to the minimum measure, and discarding the other three branches. The four paths stored in the path memory at this time are (1, 1), (1, -1, -1, -1), (-1,1), (-1, -1) in this order.

And the latter symbols adopt the same method, each time, the path with the minimum measurement is expanded, the path with the minimum measurement is reserved, other expanded paths are abandoned at the same time until the last symbol is reached, 4 paths and the corresponding measurements thereof are obtained, the sequence corresponding to the path with the minimum measurement is the final decoding output sequence which is (1, -1, -1, -1, -1,1, -1,1, 1, -1,1, -1, -1, -1,1), and the decoding is finished. The decoding process is shown in fig. 22.

The fast decoding method and the fast decoding device provided by the application can be widely applied to practical mobile communication systems such as TD-LTE, TD-SCDMA and the like, besides being applied to an OvXDM system, and can also be widely applied to any wireless communication systems such as satellite communication, microwave line-of-sight communication, scattering communication, atmospheric optical communication, infrared communication, aquatic communication and the like. The quick decoding method, the quick decoding device and the OvXDM system can be applied to large-capacity wireless transmission and can also be applied to a small-capacity light radio system.

The foregoing is a more detailed description of the present application in connection with specific embodiments thereof, and it is not intended that the present application be limited to the specific embodiments thereof. It will be apparent to those skilled in the art from this disclosure that many more simple derivations or substitutions can be made without departing from the inventive concepts herein.

Claims

1. A quick decoding method of an OvXDM system is characterized by comprising the following steps:

wherein R is_nIs a positive integer and is less than the number of nodes of the trellis diagram corresponding to the high-weight OvXDM system; a system in which the high redefinition is that the number of overlays K is 8 or more.

2. The fast decoding method for OvXDM system as claimed in claim 1, wherein in step three, when each instantaneous measurement is added to the accumulated measurement corresponding to the previous time, the accumulated measurement is multiplied by the weighting factor and then added to the instantaneous measurement.

3. The fast decoding method for OvXDM system as claimed in claim 2, wherein the weighting factor is greater than 0 and less than or equal to 1.

4. The fast decoding method for OvXDM system as claimed in claim 1, wherein r is log_MR_nIs rounded down, wherein M represents the dimension of the system, and the value is an integer greater than or equal to 2.

5. The fast decoding method for OvXDM system as claimed in any one of claims 1 to 4, wherein the high-repetition OvXDM system is an OvTDM system, an OvFDM system, an OvCDM system, an OvSDM system or an OvHDM system.

6. A fast decoding apparatus of OvXDM system, comprising:

a second sorting module for calculating the accumulated measure of each extended path obtained by the second calculation module and the stored R which is not extended_nOrdering by 1 measure, with the smaller R_nThe individual measures and their respective corresponding paths are used to update the R_nDistance memory and corresponding R_nA value of the individual path memory;

a comparison output module; when the expansion module expands the path corresponding to the currently stored minimum measure to reach the depth of the received symbol sequence, the second calculation module calculates the instantaneous measure between the path expanded by the expansion module and the corresponding received symbol, the comparison output module compares all the instantaneous measures, and takes the path corresponding to the minimum instantaneous measure as a decoding path; otherwise, the expansion module, the second calculation module and the second sequencing module repeatedly work;

7. The fast decoding apparatus for OvXDM system as defined in claim 6, further comprising a weighting factor module, for multiplying the accumulated measure at the previous time by a weighting factor when the second calculating module adds each instantaneous measure to the accumulated measure corresponding to the previous time, so that the accumulated measure at the previous time is multiplied by the weighting factor and then added to the instantaneous measure.

8. The fast decoding apparatus for OvXDM system in accordance with claim 7, wherein the value of the weighting factor module is greater than 0 and less than or equal to 1.

9. The rapid decoding apparatus for OvXDM system in accordance with claim 6, wherein r is log_MR_nIs rounded down, wherein M represents the dimension of the system, and the value is an integer greater than or equal to 2.

10. OvXDM system, characterized in that it comprises a fast decoding device suitable for use in an OvXDM system, as claimed in any one of claims 6 to 9.

11. The OvXDM system in accordance with claim 10, wherein the OvXDM system is an OvTDM system, an OvFDM system, an OvCDM system, an OvSDM system, or an OvHDM system.