CN103595508B - The method and soft de-mapped method of fixed point process are carried out to receiving symbol - Google Patents
The method and soft de-mapped method of fixed point process are carried out to receiving symbol Download PDFInfo
- Publication number
- CN103595508B CN103595508B CN201310586469.3A CN201310586469A CN103595508B CN 103595508 B CN103595508 B CN 103595508B CN 201310586469 A CN201310586469 A CN 201310586469A CN 103595508 B CN103595508 B CN 103595508B
- Authority
- CN
- China
- Prior art keywords
- symbol
- pending
- receiving
- fixed point
- init
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
Abstract
The invention discloses a kind of method and soft de-mapped method that fixed point process is carried out to receiving symbol, wherein the method that fixed point process is carried out to receiving symbol includes：Carry out pretreatment to obtain pending reception symbol and scalar factor to receiving symbol；Object value to be compared is determined based on the pending amplitude for receiving symbol；Judge whether object value to be compared is more than the predetermined threshold of constellation symbols；If the determination result is YES, the pending real part and imaginary part for receiving symbol is all reduced into prearranged multiple, then the pending reception symbol and scalar factor to reducing after the prearranged multiple carries out fixed point process, and the saturation mark of output is set to into 1；If judged result is no, fixed point process is carried out to pending reception symbol and scalar factor, and the saturation mark of output is set to into 0.The technical program simplifies the complexity of fixed point computing during hardware is realized.
Description
Technical field
The present invention relates to information decoding technique field, more particularly to a kind of to receiving the method that symbol carries out fixed point process
And soft de-mapped method.
Background technology
Digital communication system, including typical mobile radio system and received terrestrial digital broadcasting system, its basic task
It is the efficient error free transport for realizing digital information.Constellation mapping will carry finite field " bit " the sequence mapping Cheng Shi of digital information
In " symbol " sequence of transmission.The valued space of these symbols can be one-dimensional real number space, (i.e. plural number is empty for two-dimentional real number space
Between or complex number plane), or more higher-dimension real number space (such as corresponding space of multiple-antenna MIMO system signal transmission).Constellation reflects
Penetrate comprising two key elements, i.e. planisphere and constellation point mapping mode.Planisphere represents all values of constellation mapping output symbol
The set of composition, wherein, a kind of value of each point correspondence output symbol of planisphere.Constellation point mapping mode represents input ratio
The mapped specific relation of special (group) to constellation point, or constellation point is to the mapped specific relation of bit (group), usual each constellation
The bit group that point is constituted with a bit or multiple bits is corresponded.The star of presently the most common and practical complex number space
Seat figure mainly has quadrature amplitude modulation (Quadrature Amplitude Modulation, QAM), phase-shift keying (PSK) (Phase
Shift Keying, PSK), and Amplitude phase shift keying (Amplitude-Phase Shift Keying, APSK) modulation technique.
In receiving terminal, constellation mapping is constellation demapping, abbreviation demapping.Generally, constellation demapping is according to planisphere and star
Seat point mapping mode, obtains the bit soft information that correspondence receives one or more bits of symbol with reference to channel condition information.
Under multiple Gauss channel, input signal is necessary for multiple Gauss distribution and can be only achieved channel capacity.With QAM constellation phase
Than the APSK constellation figure with circular symmetry is distributed closer to multiple Gauss.Thus, it is intended that in additive white Gaussian noise
Under (Additive White Gaussian Noise, AWGN) channel, APSK constellation figure will have more preferable property than QAM constellation
Can, the gain which brings is referred to as Shaping gains.APSK constellation figure is circular in multilayer concentric, and the constellation point on each ring exists
Phase directional is uniformly distributed.European second generation digital video broadcast-satellite standard DVB-S2 employs APSK constellation mapping.Tsing-Hua University
University proposes that one group of M rank has the APSK constellation figure of gray mappings, and the planisphere hasIndividual annulus, each annulus is by uniformPoint composition, equivalent to oneThe set of different ring radiuses composition is special equivalent to oneIts
Middle m=m1+m2=log2M；For the long bit vectors of m, wherein m1Individual bit is only related to phase place, and this m1Individual bit withBetween using PSK gray mappings, another m2Individual bit is only related to amplitude, and this m2Individual bit with
Between using PAM gray mappings.The APSK constellation figure of these parameter optimizations is provided which larger Shaping gains.
The method that prior art proposes a kind of APSK soft de-mappeds of simplification, the method are converted into its phase by symbol is received
The amplitude answered and phase place, then calculate the log-likelihood ratio of amplitude and phase place related bits respectively.The present invention from this angle,
Soft de-mapped method to proposing in prior art is improved, i.e., carry out pretreatment to receiving symbol, simplify fixed-point implementation
Complexity；And calculate the judgement of phase bits apart from when, devise new judgement range formula, further simplify and calculate multiple
Miscellaneous degree；In addition, the technology of the present invention gives RTL fixed points and implementation, the performance loss very little after fixed point.
For i-th symbol y that channel is receivedi=Hixi+ni, wherein xiFor the symbol of transmitting, HiSymbol is received for correspondence
Number yiChannel condition information, niFor white multiple Gauss additive noise, monolateral noise power is N0,i.Soft de-mapped algorithm output ratio
Special Soft Inform ation, represents that generally using log-likelihood ratio (Log-Likelihood Ratio, LLR) soft de-mapped algorithm is exported
The Soft Inform ation of bit, it is assumed that transmission symbol etc. is general to be distributed, and the LLR of the jth j bit of i-th reception symbol can be expressed as：
Wherein xiFor i-th transmitting symbol,The constellation symbols set that jth j bit is 1 is represented,Represent the
Jj bit is 0 constellation symbols set.
The LLR value of each bit can accurately be calculated using logarithmic maximum a posteriori (Log-MAP) algorithm, but calculated
In there is logarithm and exponent arithmetic, amount of calculation is quite big.By approximate formula log ∑jzj≈maxjlogzj, maximum can be obtained
Logarithmic maximum a posteriori (Max-Log-MAP) algorithm, under the baseband equivalence model of White Gaussian additive noise, Max-Log-
MAP algorithms can be expressed as：
Intersymbol distance Euclidean distance is calculated now, hard-wired complexity is still higher, and due to channel information
(Hi, N0,i) be included in two parameters, it is unfavorable for hardware fixed point and realization.
The content of the invention
The problem that the present invention is solved is, when soft de-mapped is carried out to reception symbol, to simplify fixed point during hardware is realized
The complexity of computing.
To solve the above problems, a kind of method that fixed point process is carried out to reception symbol is embodiments provided,
Including：Pretreatment is carried out to obtain pending reception symbol and scalar factor to receiving symbol, wherein pending the reception accords with
Number it is to eliminate the reception symbol of channel information, the scalar factor is related to channel information；Accorded with based on pending the reception
Number amplitude determine object value to be compared；Judge whether the object value to be compared is more than the predetermined threshold of constellation symbols；If
Above-mentioned judged result is yes, and the pending real part and imaginary part for receiving symbol is all reduced prearranged multiple, then pre- to reducing this
Determining the pending reception symbol after multiple and scalar factor carries out fixed point process, and the saturation mark of output is set to 1；Its
In, the prearranged multiple is related to the predetermined threshold of the object value to be compared and the constellation symbols；If above-mentioned judgement knot
Fruit is no, carries out fixed point process to the pending reception symbol and scalar factor, and the saturation mark of output is set to
0。
The embodiment of the present invention additionally provides a kind of soft de-mapped method, including：Carried out to receiving symbol using said method
Fixed point process；Based on different saturation marks, to the pending arrow for receiving symbol phase place composition corresponding with constellation point
Amount adopts corresponding processing mode to determine intermediate data；It is based on different saturation marks, true using the first judgement range formula
The judgement distance of the fixed pending phase place related bits for receiving symbol, wherein the first judgement range formula is associated with this
Intermediate data；The judgement distance of the pending amplitude related bits for receiving symbol is determined using the second judgement range formula,
Wherein described second judgement range formula is associated with the maximum of the intermediate data；Respectively according to the described first judgement range formula
To second judgement range formula determine institute it is to be handled reception symbol phase place related bits log-likelihood ratio it is related to amplitude
The log-likelihood ratio of bit.
Compared with prior art, technical solution of the present invention has the advantages that：
Receiving terminal carries out pretreatment by the symbol to receiving to obtain eliminating the reception symbol of channel information and bag
Scalar factor containing channel information, and fixed point process is carried out to the pending reception symbol and scalar factor.As channel is believed
Breath is only related to scalar factor, is more beneficial for during hardware is realized for the selection of bit wide.
Further, after fixed point process, soft de-mapped is carried out to pending reception symbol.According to different saturations
Mark, the vector constituted to pending reception phase place corresponding with constellation point determine mediant using corresponding processing mode
According to；And then using the judgement range formula of the intermediate data is associated with accordingly determining phase place related bits and amplitude correlation ratio
Special judgement distance, so as to simplify computation complexity.
Description of the drawings
Fig. 1 is the structural representation of the pending reception symbol that the receiving terminal of the present invention is received；
Fig. 2 is that a kind of flow process of the specific embodiment of method that fixed point process is carried out to reception symbol of the present invention is shown
It is intended to；
Fig. 3 is a kind of schematic flow sheet of the specific embodiment of soft de-mapped method of the present invention；
Fig. 4 A are contrasted using the soft de-mapped method and other existing algorithms of present invention offer to receiving symbol
The simulation result schematic diagram of one example；
Fig. 4 B are contrasted using the soft de-mapped method and other existing algorithms of present invention offer to receiving symbol
The simulation result schematic diagram of another example.
Specific embodiment
Inventor has found that the complexity of the existing soft de-mapped method of receiving terminal is higher, and is unfavorable for realizing process in hardware
In carry out fixed point computing.
For the problems referred to above, inventor is through research, there is provided a kind of to receiving the method that symbol carries out fixed point process
And soft de-mapped method, to simplify the complexity of fixed point computing during hardware is realized.
It is understandable to enable the above objects, features and advantages of the present invention to become apparent from, below in conjunction with the accompanying drawings to the present invention
Specific embodiment be described in detail.
The soft de-mapped method that the technical program is provided is mapped suitable for APSK constellation.What receiving terminal was received is by multiple
Through the symbol stream of APSK constellation mapping, the symbol of each the APSK constellation mapping in symbol stream can be to reply ratio
It is special.
As shown in Fig. 1 the pending reception symbol that receiving terminal of the present invention is received (eliminates the reception of channel information
Symbol) structural representation.With reference to Fig. 1, the symbol of each APSK constellation mapping corresponds to bit group of the length for m, by this m
Individual bit is divided into continuous front m2Individual bitWith continuous rear m1Individual bitIts
In, m1Be with it is pending receive symbol phase place related bits number, m2It is and the pending amplitude related bits for receiving symbol
Number.
As shown in Fig. 2 the present invention's is a kind of to receiving the specific embodiment that symbol carries out the method for fixed point process
Schematic flow sheet.With reference to Fig. 2, described pair receives symbol and carries out the method for fixed point process and comprise the steps：
Step S11：Pretreatment is carried out to obtain pending reception symbol and scalar factor, wherein described treat to receiving symbol
It is the reception symbol for eliminating channel information to process and receive symbol, and the scalar factor is related to channel information；
Step S12：Object value to be compared is determined based on the pending amplitude for receiving symbol；
Step S13：Judge whether the object value to be compared is more than the predetermined threshold of constellation symbols；
Step S14：If above-mentioned judged result is yes, the pending real part and imaginary part for receiving symbol is all reduced into predetermined
Multiple, then the pending reception symbol and scalar factor to reducing after the prearranged multiple carry out fixed point process, and will output
Saturation mark be set to 1；Wherein, the prearranged multiple and the object value to be compared and the predetermined threshold of the constellation symbols
It is related；
Step S15：If above-mentioned judged result is no, fixed point is carried out to the pending reception symbol and scalar factor
Process, and the saturation mark of output is set to into 0.
In a particular embodiment, as described in step S11, carry out pretreatment to obtain pending reception symbol to receiving symbol
And scalar factor.
Unlike the prior art, in the present embodiment, receiving terminal first carries out pretreatment to the reception symbol for receiving
The channel information included in symbol is received to remove, so as to simplify the complexity of fixed-point calculation during hardware is realized.
Specifically, for i-th reception symbol that channel is received, it is set to yi=Hixi+ni, wherein, xiSend out for transmitting terminal
The symbol penetrated, HiSymbol y is received for correspondenceiChannel condition information, niFor white multiple Gauss additive noise.To the reception symbol
The pending reception symbol obtained after carrying out pretreatment is：
The pending reception symbol is the reception symbol for eliminating channel information, wherein,
The channel information includes decay factor and channel noise power.
In addition, also obtaining the scalar factor comprising channel information, it is set toWherein, N0,iRepresent monolateral
Noise power.Thus, channel information only with a parameterIt is relevant, it is more beneficial for the selection to bit wide during hardware is realized.
In other embodiments,Calculating can also complete in prime balance module, regardless of whether willCalculating be placed on before
Calculate during pretreatment module is calculated or be placed in level balance module, the output parameter of pretreatment module isWith
As described in step S12, object value to be compared is determined based on the pending amplitude for receiving symbol.
In the present embodiment, can be so that object value to be compared be determined using the following two kinds mode.
Mode is 1)：Determine respectively it is described it is pending receive symbol real part amplitude and imaginary part amplitude, and with both compared with
Big value is defined as the object value to be compared.
For example, ifThe amplitude of real part beThe amplitude of imaginary part isObject value to be compared is s_in.Then
Mode is 2)：Determine the pending amplitude for receiving symbol, and the comparison other value is defined as with the amplitude.
For example, ifThe amplitude of real part beThe amplitude of imaginary part isObject value to be compared is s_in.Then
As described in step S13, judge whether the object value to be compared is more than the predetermined threshold of constellation symbols.Wherein,
The predetermined threshold of the constellation symbols is according to going to push up the different modes of object value to be compared and different in above-mentioned steps S12.
Specifically, object value to be compared 1) is determined according to the mode, then the predetermined threshold of constellation symbols isWherein m1It is the number with the pending phase place related bits for receiving symbol；2) determine according to the mode and treat
Bit object value, then the predetermined threshold of constellation symbols is the maximum radius in the corresponding all annular radiis of planisphere.
As described in step S14, if the judged result of above-mentioned steps S13 is to be that (i.e. described object value to be compared is more than
The predetermined threshold of constellation symbols), the pending real part and imaginary part for receiving symbol is all reduced into prearranged multiple, then to reducing
Pending reception symbol and scalar factor after the prearranged multiple carries out fixed point process, and the saturation mark of output is set to
1。
In the present embodiment, if the prearranged multiple is α, the pending real part and imaginary part for receiving symbol is all reduced
Prearranged multiple α postscripts areWherein, α=log2{ ceil (s_in/thd) }, that is to say, that α and the object to be compared
Value is related to the predetermined threshold of the constellation symbols, and what ceil functions were returned is greater than or equal to specify expression formula (this reality
Apply a middle finger s_in/thd) smallest positive integral.
Further, to reduce prearranged multiple it is described pending reception symbol by the way of (int1, int2, t)
Carry out fixed point process.Wherein, (int1, int2, t) represents pending reception symbol signed number, and first bit is represented
Sign bit；Integer part accounts for (int1-int2-1) position；Fractional part accounts for int2 positions, and low level rounds up, high-order saturation cut position,
Wherein, t represents signed number, and int1 represents that entirely the bit wide of fixed-point data, int2 represent signed number in the case of signed number
In the case of bit wide shared by fractional part.
Fixed point process is carried out by the way of (int3, int4, u) to scalar factor.Wherein, (int3, int4, u) table
Show scalar factor unsigned number, integer part account for (int3-int4) position, fractional part account for int4 positions, low level round up, it is high-order
Saturation cut position, wherein, u represents unsigned number, and int3 represents that entirely the bit wide of fixed-point data, int4 are represented in the case of unsigned number
Bit wide shared by fractional part in the case of unsigned number.
As described in step S15, if the judged result of above-mentioned steps S13 for it is no (i.e. described object value to be compared be less than or
Equal to the predetermined threshold of constellation symbols), fixed point process is carried out to the pending reception symbol and scalar factor, and will
The saturation mark of output is set to 0.
It is similar with the mode that fixed point described in above-mentioned steps S14 is processed, the pending reception symbol is adopted
The mode of (int1, int2, t) carries out fixed point process, carries out fixed point to scalar factor by the way of (int3, int4, u)
Process.Wherein, being specifically defined with reference to the description in above-mentioned steps S14 to (int1, int2, t) and (int3, int4, u),
This repeats no more.
The embodiment of the present invention additionally provides a kind of soft de-mapped method.As shown in Fig. 3 a kind of soft solution of the invention is reflected
The schematic flow sheet of the specific embodiment of shooting method.With reference to Fig. 3, the soft de-mapped method comprises the steps：
Step S21：Fixed point process is carried out to receiving symbol using method provided in an embodiment of the present invention；
Step S22：Based on different saturation marks, the pending symbol phase place corresponding with constellation point that receives is constituted
Vector adopt corresponding processing mode to determine intermediate data；
Step S23：Based on different saturation marks, the pending reception symbol is determined using the first judgement range formula
Phase place related bits judgement distance, wherein it is described first judgement range formula be associated with the intermediate data；
Step S24：Sentencing for the pending amplitude related bits for receiving symbol is determined using the second judgement range formula
Certainly distance, wherein the second judgement range formula is associated with the maximum of the intermediate data；
Step S25：Determine that institute is to be handled according to the described first judgement range formula and the second judgement range formula respectively
Receive the log-likelihood ratio and the log-likelihood ratio of amplitude related bits of the phase place related bits of symbol.
In the present embodiment, as described in step S22, based on different saturation marks, the pending reception is accorded with
The vector of number corresponding with constellation point phase place composition adopts corresponding processing mode to include to determine intermediate data：
If the saturation is masked as 1,
If the saturation is masked as 0,
Wherein,The pending reception symbol described in i-th；α is prearranged multiple, and α is real number；Wherein, P_init is the vector of the corresponding phase place composition of constellation point；m1It is and pending reception symbol
Phase place related bits number；Re () is treating excess syndrome portion operator；ri,kFor the intermediate data.
As described in step S23, based on different saturation marks, described pending connect is determined using the first judgement range formula
The judgement distance for receiving the phase place related bits of symbol includes：
If the saturation is masked as 1,
If the saturation is masked as 0,
Wherein,The pending reception symbol described in i-th；α is prearranged multiple, and α is real number；ri,kFor the centre
Data；m1It is the number with the pending phase place related bits for receiving symbol；Pi,kIt is related to the pending phase place for receiving symbol
The judgement distance of bit.
Wherein, m2It is the number with the pending amplitude related bits for receiving symbol；A_init for planisphere annulus half
The vector of footpath composition.
In step S24, the second judgement range formula is：
Wherein, riFor the maximum of the intermediate data；Vectors of the A_init for the radius composition of planisphere annulus；Ai,k
It is the judgement distance with the pending amplitude related bits for receiving symbol.
As described in step S25, determined according to the described first judgement range formula and the second judgement range formula respectively all
The log-likelihood ratio and the log-likelihood ratio of amplitude related bits of the pending phase place related bits for receiving symbol.
Specifically, this step includes：
1) according to described first judgement range formula determine institute it is to be handled reception symbol phase place related bits logarithm
Likelihood ratio.
If i-th pending jth j for receiving symbol is LLR with the log-likelihood ratio of phase place related bitsi,jj
Wherein,It is the scalar factor related to channel information；P_bin is oneRow m1Matrix in two codomain of row,
It is gray mappings between the row bit sequence of its matrix.
2) according to described second judgement range formula determine institute it is to be handled reception symbol amplitude related bits logarithm
Likelihood ratio.
If i-th pending jth j for receiving symbol is LLR' with the log-likelihood ratio of amplitude related bitsi,jj
Wherein,It is the scalar factor related to channel information；A_bin is oneRow m2Matrix in two codomain of row,
It is gray mappings between the row bit sequence of its matrix.
In sum, the complexity of fixed point computing during hardware is realized is simplified using the technical program.
Inventor also to receive symbol adopt soft de-mapped method of the present invention and other existing algorithms to carry out emulation experiment with
Carry out Contrast on effect.
Example 1：
Using M=256 rank APSK constellation figures, m1=5, m2=3；The LDPC code of 9/13 code check, 59904 bit of code length；LDPC
Interweaved using ranks between code and Gray-APSK mappings；De-mapping method is respectively adopted Log-MAP algorithms, existing simplified solution
Mapping method (Simplified v1 are labeled as in Fig. 4 A) and its input fixed point algorithm (are labeled as Simplified in Fig. 4 A
V1, input fixed) and floating-point arithmetic (Simplified v2 are labeled as in Fig. 4 A) of the invention and its input fixed point calculation
Method (is labeled as Simplified v2, input fixed) in Fig. 4 A；LDPC decoding algorithms adopt BP algorithm, maximum iteration time
For 50 times；Channel is awgn channel；Under above-mentioned 256-APSK constellations, the phase vectors of each constellation point areCorresponding P_bin is the two values matrix of the row of 32 row 5, is expressed as [0 0000；0 0
0 0 1；0 0 0 1 1；0 0 0 1 0；0 0 1 1 0；0 0 1 1 1；0 0 1 0 1；0 0 1 0 0；0 1 1 0 0；0
1 1 0 1；0 1 1 1 1；0 1 1 1 0；0 1 0 1 0；0 1 0 1 1；0 1 0 0 1；0 1 0 0 0；1 1 0 0
0；1 1 0 0 1；1 1 0 1 1；1 1 0 1 0；1 1 1 1 0；1 1 1 1 1；1 1 1 0 1；1 1 1 0 0；1 0 1
0 0；1 0 1 0 1；1 0 1 1 1；1 0 1 1 0；1 0 0 1 0；1 0 0 1 1；1 0 0 0 1；1 0 0 0 0]；Star
Seat schemes the radius vectors of each annulus
A_init=[0.2639,0.4750,0.6333,0.7916,0.9499,1.1346,1.3457,1.6360] is right
The A_bin for answering is the two values matrix of the row of 8 row 3, is expressed as [0 00；0 0 1；0 1 1；0 1 0；1 1 0；1 1 1；1 0 1；
1 0 0]。
In the case where above-mentioned parameter is arranged, 1000 LDPC blocks of simulation run, the BER (bit error rate) of the code modulation system of gained
Simulation result is as shown in Figure 4 A.With reference to Fig. 4 A, it is 10 in BER-6When, the method for the present invention coordinates BP algorithm to match somebody with somebody with respect to Log-MAP
BP algorithm is closed, performance loss is 0.05dB, fixed point performance is almost lossless compared with floating-point performance, hence it is evident that better than existing simplification
The fixed point algorithm of de-mapping method, and any trigonometric function operation is not needed, amount of calculation has obtained great simplification.
Example 2：
With simulation parameter in example 1, but demapping, only with Log-MAP algorithms, (Fig. 4 B's floating-point arithmetic of the present invention get the bid
Be designated as Simplified) and its population parameter (include inner parameter) fixed point algorithm (in Fig. 4 B, be labeled as Simplified, all
Fixed), from Fig. 4 B, even if population parameter all fixed-point implementations, the performance of inventive algorithm is only poor than Log-MAP algorithm
0.1dB, but complexity is much lower.
Performance loss very little of the method for the present invention in the case where computational complexity is simplified to system, to reducing hardware
Realize the resource for consuming and improve calculating speed being very helpful.
Although the present invention is disclosed as above with preferred embodiment, which is not for limiting the present invention, any this area
Technical staff without departing from the spirit and scope of the present invention, may be by the methods and techniques content of the disclosure above to this
Bright technical scheme makes possible variation and modification, therefore, every content without departing from technical solution of the present invention, according to the present invention
Technical spirit any simple modification, equivalent variations and modification that above example is made, belong to technical solution of the present invention
Protection domain.
Claims (11)
1. it is a kind of to receiving the method that symbol carries out fixed point process, it is characterised in that to include：
Pretreatment is carried out to obtain pending reception symbol and scalar factor, wherein the pending reception symbol to receiving symbol
To eliminate the reception symbol of channel information, the scalar factor is related to channel information；
Object value to be compared is determined based on the pending amplitude for receiving symbol；
Judge whether the object value to be compared is more than the predetermined threshold of constellation symbols；
If above-mentioned judged result is yes, the pending real part and imaginary part for receiving symbol is all reduced into prearranged multiple, then to contracting
Pending reception symbol and scalar factor after the little prearranged multiple carries out fixed point process, and the saturation mark of output is put
For 1；Wherein, the prearranged multiple is related to the predetermined threshold of the object value to be compared and the constellation symbols；
If above-mentioned judged result is no, fixed point process is carried out to the pending reception symbol and scalar factor, and will be defeated
The saturation mark for going out is set to 0,
Wherein, it is described to determine that object value to be compared includes following any one mode based on the pending amplitude for receiving symbol：
Mode is 1)：The pending real part amplitude and imaginary part amplitude for receiving symbol is determined respectively, and with the higher value in both
It is defined as the object value to be compared；
Mode is 2)：Determine the pending amplitude for receiving symbol, and the object value to be compared is defined as with the amplitude.
2. as claimed in claim 1 to receiving the method that symbol carries out fixed point process, it is characterised in that
1) object value to be compared is determined according to the mode, then the predetermined threshold of constellation symbols isWherein m1
It is the number with the pending phase place related bits for receiving symbol；
2) determine according to the mode and treat bit object value, then the predetermined threshold of constellation symbols is that planisphere is corresponding all
Maximum radius in annular radii.
3. it is as claimed in claim 1 to receiving the method that symbol carries out fixed point process, it is characterised in that to described pending
Pending reception symbol after receiving symbol or reducing prearranged multiple carries out fixed point process to be included：
To the pending pending reception symbol for receiving symbol or reducing after prearranged multiple using (int1, int2, t)
Mode carries out fixed point process；Wherein, (int1, int2, t) represents pending reception symbol signed number, and first bit
Represent sign bit；Integer part accounts for (int1-int2-1) position；Fractional part accounts for int2 positions, and low level rounds up, high-order saturation is cut
Position,
Wherein, t represents signed number, and int1 represents that entirely the bit wide of fixed-point data, int2 indicate symbol in the case of signed number
Bit wide shared by fractional part in the case of count.
4. it is as claimed in claim 1 to receiving the method that symbol carries out fixed point process, it is characterised in that scalar factor is entered
The process of row fixed point includes：
Fixed point process is carried out by the way of (int3, int4, u) to scalar factor；Wherein, (int3, int4, u) represents mark
Amount factor unsigned number, integer part account for (int3-int4) position, fractional part account for int4 positions, low level round up, high-order saturation
Cut position,
Wherein, u represents unsigned number, and int3 represents that entirely the bit wide of fixed-point data, int4 are indicated without symbol in the case of unsigned number
Bit wide shared by fractional part in the case of count.
5. a kind of soft de-mapped method, it is characterised in that comprise the steps：
Fixed point process is carried out to receiving symbol using the method described in claim 1；
Based on different saturation marks, to the pending vector for receiving symbol phase place composition corresponding with constellation point using right
The processing mode answered is determining intermediate data；
Based on different saturation marks, the pending phase place correlation ratio for receiving symbol is determined using the first judgement range formula
Special judgement distance, wherein the first judgement range formula is associated with the intermediate data；
The judgement distance of the pending amplitude related bits for receiving symbol, wherein institute are determined using the second judgement range formula
State the maximum that the second judgement range formula is associated with the intermediate data；
Respectively according to described first judgement range formula and second judgement range formula determine institute it is to be handled reception symbol phase
The log-likelihood ratio and the log-likelihood ratio of amplitude related bits of position related bits.
6. soft de-mapped method as claimed in claim 5, it is characterised in that described based on different saturation marks, to described
The pending vector for receiving symbol phase place composition corresponding with constellation point adopts corresponding processing mode to determine intermediate data bag
Include：
If the saturation is masked as 1,
If the saturation is masked as 0,Wherein,
The pending reception symbol described in i-th；
α is prearranged multiple, and α is real number；
Wherein, P_init is the vector of the corresponding phase place composition of constellation point；
m1It is the number with the pending phase place related bits for receiving symbol；
Re () is treating excess syndrome portion operator；
ri,kFor the intermediate data.
7. soft de-mapped method as claimed in claim 6, it is characterised in that described based on different saturation marks, utilizes
One judgement range formula determines that the judgement distance of the pending phase place related bits for receiving symbol includes：
If the saturation is masked as 1,
If the saturation is masked as 0,Wherein,
The pending reception symbol described in i-th；
α is prearranged multiple, and α is real number；
ri,kFor the intermediate data；
m1It is the number with the pending phase place related bits for receiving symbol；
Pi,kIt is the judgement distance with the pending phase place related bits for receiving symbol；
Wherein, m2It is the number with the pending amplitude related bits for receiving symbol；
Vectors of the A_init for the radius composition of planisphere annulus.
8. soft de-mapped method as claimed in claim 5, it is characterised in that the second judgement range formula is：
Ai,k=(ri-A_initk)2,Wherein,
riFor the maximum of the intermediate data；
Vectors of the A_init for the radius composition of planisphere annulus；
Ai,kIt is the judgement distance with the pending amplitude related bits for receiving symbol.
9. soft de-mapped method as claimed in claim 7, it is characterised in that institute is determined according to the described first judgement range formula
The log-likelihood ratio of the phase place related bits for receiving symbol to be handled includes：
Wherein,It is the scalar factor related to channel information；
P_bin is oneRow m1Matrix in two codomain of row, is gray mappings between the row bit sequence of its matrix；
LLRi,jjRepresent log-likelihood ratio of jth j of i-th pending reception symbol with phase place related bits.
10. soft de-mapped method as claimed in claim 8, it is characterised in that determined according to the described second judgement range formula
The log-likelihood ratio of amplitude related bits for receiving symbol to be handled include：
Wherein,It is the scalar factor related to channel information；
A_bin is oneRow m2Matrix in two codomain of row, is gray mappings between the row bit sequence of its matrix；
LLR'i,jjRepresent log-likelihood ratio of jth j of i-th pending reception symbol with amplitude related bits.
11. soft de-mapped methods as claimed in claim 5, it is characterised in that the soft de-mapped mode is applied to APSK constellation
Mapping.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310586469.3A CN103595508B (en) | 2013-11-20 | 2013-11-20 | The method and soft de-mapped method of fixed point process are carried out to receiving symbol |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310586469.3A CN103595508B (en) | 2013-11-20 | 2013-11-20 | The method and soft de-mapped method of fixed point process are carried out to receiving symbol |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103595508A CN103595508A (en) | 2014-02-19 |
| CN103595508B true CN103595508B (en) | 2017-03-29 |
Family
ID=50085505
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310586469.3A Active CN103595508B (en) | 2013-11-20 | 2013-11-20 | The method and soft de-mapped method of fixed point process are carried out to receiving symbol |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103595508B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105162545B (en) * | 2015-10-16 | 2018-06-29 | 中国科学院上海高等研究院 | Realize the QAM soft-decision system and methods of low complex degree high power efficiency |
| CN107204954B (en) * | 2016-03-16 | 2019-12-03 | 中国科学院上海高等研究院 | A kind of optimization method and system of fixed point Soft Inform ation |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1984100A (en) * | 2006-04-05 | 2007-06-20 | 威盛电子股份有限公司 | Method of equalizing signal in orthogonal frequency-division multiplexing fundamental frequency receiver |
| CN101138212A (en) * | 2005-03-08 | 2008-03-05 | 法国原子能委员会 | Method for the flexible demodulation of estimated complex symbols |
| CN102150377A (en) * | 2006-12-01 | 2011-08-10 | 北方电讯网络有限公司 | Antenna selection and soft demapping for MIMO decoding |
| CN102202035A (en) * | 2010-03-25 | 2011-09-28 | 卓胜微电子（上海）有限公司 | Irregular symbol constellation point soft-demapping realization method for communication system |
| CN102710581A (en) * | 2012-05-07 | 2012-10-03 | 清华大学 | Soft solution mapping method |
-
2013
- 2013-11-20 CN CN201310586469.3A patent/CN103595508B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101138212A (en) * | 2005-03-08 | 2008-03-05 | 法国原子能委员会 | Method for the flexible demodulation of estimated complex symbols |
| CN1984100A (en) * | 2006-04-05 | 2007-06-20 | 威盛电子股份有限公司 | Method of equalizing signal in orthogonal frequency-division multiplexing fundamental frequency receiver |
| CN102150377A (en) * | 2006-12-01 | 2011-08-10 | 北方电讯网络有限公司 | Antenna selection and soft demapping for MIMO decoding |
| CN102202035A (en) * | 2010-03-25 | 2011-09-28 | 卓胜微电子（上海）有限公司 | Irregular symbol constellation point soft-demapping realization method for communication system |
| CN102710581A (en) * | 2012-05-07 | 2012-10-03 | 清华大学 | Soft solution mapping method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103595508A (en) | 2014-02-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gan et al. | Complex lattice reduction algorithm for low-complexity full-diversity MIMO detection | |
| US20080049863A1 (en) | Apparatus, method and computer program product providing soft decision generation with lattice reduction aided MIMO detection | |
| JP5326976B2 (en) | Wireless communication apparatus, error correction method, and error correction program | |
| US8243860B1 (en) | Co-channel interference cancellation with multiple receive antennas for BICM | |
| US8737540B1 (en) | System and method for providing reduced complexity maximum likelihood MIMO detection | |
| CN106911374A (en) | A kind of low complexity soft exports spatial modulation detection method | |
| US20090154600A1 (en) | QRD-QLD searching based sphere detector for MIMO receiver | |
| CN103581103B (en) | It is applicable to the soft de-mapped method of APSK constellation | |
| CN102457470B (en) | Method and apparatus for soft-decision sphere decoding | |
| CN104301267A (en) | Multi-stage iterative detection method and device of MIMO wireless communication receiver | |
| CN110417515B (en) | Large-scale MIMO signal detection method based on discrete iterative estimation | |
| KR102010562B1 (en) | Apparatus and method for generating decoding metric for frequency-quadrature amplitude modulated signals in wireless communication system | |
| CN105099974B (en) | A kind of spatial modulation soft decision detection method of low complex degree | |
| EP2745405B1 (en) | Method and apparatus for soft symbol processing in a communication receiver | |
| CN103595508B (en) | The method and soft de-mapped method of fixed point process are carried out to receiving symbol | |
| CN102739576B (en) | Soft bit digital demodulating method and device of planisphere based on complex number space | |
| CN101197603B (en) | Low complexity step-by-step detecting system and method of multi-antenna system based on spherical decoding | |
| CN103023622A (en) | Signal detection method and signal detection equipment | |
| CN108199752A (en) | A kind of dual polarization satellite MIMO signal detection method of low complex degree | |
| CN113364535B (en) | Method, system, device and storage medium for mathematical form multiple-input multiple-output detection | |
| WO2016143863A1 (en) | Communication device, demodulation method, and program | |
| CN103580808A (en) | Soft demapping method | |
| CN105553899A (en) | Signal detection method and device based on approximate solution solving of linear equation group | |
| WO2020046620A1 (en) | Communication device and method for processing a received signal | |
| Gou et al. | Low-complexity sphere decoding for quadrature spatial modulation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: Room 1018, block B, No. three East Bridge Road, Pudong New Area, Shanghai, 200125, China Applicant after: Shanghai NERC-DTV National Engineering Research Center Co., Ltd. Address before: 200125 Shanghai East Road, Pudong New Area, No. three, No. 1018 Applicant before: Shanghai NERC-DTV National Engineering Research Center Co., Ltd. |
|
| COR | Change of bibliographic data | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |