CN102263563B - A kind of processing method of the Soft Inform ation for decoding and device - Google Patents
A kind of processing method of the Soft Inform ation for decoding and device Download PDFInfo
- Publication number
- CN102263563B CN102263563B CN201010187352.4A CN201010187352A CN102263563B CN 102263563 B CN102263563 B CN 102263563B CN 201010187352 A CN201010187352 A CN 201010187352A CN 102263563 B CN102263563 B CN 102263563B
- Authority
- CN
- China
- Prior art keywords
- inform ation
- soft inform
- average
- decoder
- ation
- 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
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/3723—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35 using means or methods for the initialisation of the decoder
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/45—Soft decoding, i.e. using symbol reliability information
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/65—Purpose and implementation aspects
- H03M13/6577—Representation or format of variables, register sizes or word-lengths and quantization
- H03M13/6583—Normalization other than scaling, e.g. by subtraction
Abstract
The invention discloses a kind of processing method and device of the Soft Inform ation for decoding.The method comprises: obtain the Soft Inform ation needed for decoding according to the signal received, and calculates the average of described Soft Inform ation; The average of described Soft Inform ation and default Soft Inform ation threshold value are compared, according to comparative result, described Soft Inform ation is moved to right; Whether the numerical value of the Soft Inform ation after judging to move to right belongs to the input range of decoder, if do not belong to the input range of decoder, then the numerical value of the Soft Inform ation after described moving to right is set to the maximum in the input range of decoder or minimum value, as the numerical value after the Soft Inform ation process before moving to right; If belong to the input range of decoder, then the Soft Inform ation after described moving to right is intercepted from low level the number of bits that decoder allows input, as the numerical value after the Soft Inform ation process before moving to right; Numerical value after described Soft Inform ation process is supplied to decoder for decoding.Carry out decoding with the Soft Inform ation of the method and device process and can improve decoding performance.
Description
Technical field
The present invention relates to the communications field, particularly relate to a kind of processing method and device of the Soft Inform ation for decoding.
Background technology
Over the communication channels during transmission of digital signals, because channel transfer characteristic pays no attention to the impact thinking of additive noise, the digital signal that receiving terminal receives inevitably makes a mistake.In order to obtain good error rate index in certain SNR ranges, the error rate can be reduced as far as possible by technology such as appropriate design baseband signal, selection modulation demodulation system, the equilibriums of employing time-domain and frequency-domain.In addition, also can the performance of influential system to the restriction of the quantification of signal and precision.
When demodulation is carried out to signal, the method for soft-decision can be adopted.Soft-decision does not need to judge that the bit received is 0 or 1 immediately, to the decoder based on log-likelihood ratio, receiver assigns one to represent to each bit, and this bit is the multilevel scale value of the probability of 1 or 0, i.e. log-likelihood ratio probability, with an integer of certain scope, namely Soft Inform ation represents each bit.Here, the amplitude of Soft Inform ation does not represent the numerical value of this bit, and only represent a probability, when Soft Inform ation is a positive number, the amplitude of this positive number is larger, represent bit representated by Soft Inform ation be 0 probability larger, when Soft Inform ation is a negative, the amplitude of this negative is larger, represent bit representated by Soft Inform ation be 1 probability larger, when Soft Inform ation is zero, represent that the logical bits representated by Soft Inform ation be can not determine.This shows, the number range of the Soft Inform ation of input decoder is larger, and the performance of decoder is better, and the error rate is lower.But for fixed-point implementation, if wish that Soft Inform ation retains all data messages, namely number range is enough large, then need many number of bits, but decoding time will be long like this, thus affects decoding efficiency.So in order to ensure decoding efficiency, generally can limit to some extent the input range of decoder, to improve the decoding efficiency of decoder.The input range of decoder refers to that decoder allows the number range corresponding to maximal bit figure place of input.In actual conditions, the number range of the Soft Inform ation that demodulation obtains generally all is greater than the input range of decoder, namely the number of bits of Soft Inform ation that demodulation obtains is greater than the number of bits that decoder allows input, so need the Soft Inform ation to demodulation obtains to carry out bit truncation, to meet the requirement of decoding efficiency.Such as, the Soft Inform ation that demodulation obtains is 16bit signed number, and its number range is [-32768,32767].The input range of decoder is then [-128,127], namely input 8bit signed number is only allowed, the data of the input range exceeding permission are then needed to give up, Soft Inform ation by 16bit intercepts as 8bit, to improve the decoding efficiency of decoder, but so, the precision of Soft Inform ation will certainly be reduced, the error rate index of influential system.Implementation at present when carrying out bit truncation to Soft Inform ation is all generally directly intercept higher bit position.Such as, for the Soft Inform ation 0000000011000001 of a 16bit, if decoder allows the maximal bit figure place of input to be 8bit, then will be intercepted is the data of a 8bit, current implementation is generally directly from a high position, intercept 8 bits, namely intercept high-order 8bit data, therefore, this Soft Inform ation is 00000000 after intercepting.Again for the Soft Inform ation 0000000000000011 of another one 16bit, if decoder allows the maximal bit figure place of input to be also 8bit, so according to current interception way, result after this Soft Inform ation intercepts and first Soft Inform ation come to the same thing, be all 00000000, but in fact the amplitude of these two Soft Inform ation still has very big difference, adopt the direct mode intercepting the number of bits of decoder permission input from a high position in prior art then well can not embody difference therebetween, thus affect decoding performance.
This shows, adopt in prior art and directly intercept from a high position method that decoder allows the number of bits of input, there is certain defect: when the amplitude of most of data is all less than normal, especially the value when higher bit position is all zero, time real effective information is all positioned at low bit, an amplitude own is intercepted high-order 8bit data again with regard to very little data, the precision of these data will be caused seriously to reduce, thus loss bulk information, reduce decoding performance.
Summary of the invention
The embodiment of the present invention provides a kind of method and device of Soft Inform ation process, for improving decoding performance.For a processing method for the Soft Inform ation of decoding, the method comprises the following steps:
Obtain the Soft Inform ation needed for decoding according to the signal received, and calculate the average of described Soft Inform ation; The average of described Soft Inform ation and default Soft Inform ation threshold value are compared, according to comparative result, described Soft Inform ation is moved to right;
Whether the numerical value of the Soft Inform ation after judging to move to right belongs to the input range of decoder, if do not belong to the input range of decoder, then the numerical value of the Soft Inform ation after described moving to right is set to the maximum in the input range of decoder or minimum value, as the numerical value after the Soft Inform ation process before moving to right; If belong to the input range of decoder, then the Soft Inform ation after described moving to right is intercepted from low level the number of bits that decoder allows input, as the numerical value after the Soft Inform ation process before moving to right;
Numerical value after described Soft Inform ation process is supplied to decoder for decoding.
For a processing unit for the Soft Inform ation of decoding, comprising:
Average calculation unit, for obtaining the Soft Inform ation needed for decoding according to the signal received, and calculates the average of described Soft Inform ation;
Relatively shift unit, for the average of described Soft Inform ation and default Soft Inform ation threshold value being compared, moves to right described Soft Inform ation according to comparative result;
Judge processing unit, whether the numerical value for the Soft Inform ation after judging to move to right belongs to the input range of decoder, if do not belong to the input range of decoder, then the numerical value of the Soft Inform ation after described moving to right is set to the maximum in the input range of decoder or minimum value, as the numerical value after the Soft Inform ation process before moving to right; If belong to the input range of decoder, then the Soft Inform ation after described moving to right is intercepted from low level the number of bits that decoder allows input, as the numerical value after the Soft Inform ation process before moving to right;
Data transmission unit, for being supplied to decoder for decoding by the numerical value after described Soft Inform ation process.
The processing method of the Soft Inform ation in the embodiment of the present invention and device, by the relation of softer information average with the Soft Inform ation threshold value preset, and according to comparative result, Soft Inform ation is moved to right, Soft Inform ation is intercepted again according to the relation between the numerical value of the Soft Inform ation after moving to right and decoder input range, thus the maximum numerical value of Soft Inform ation larger for amplitude is directly taken as in decoder input range or minimum value, and Soft Inform ation less for amplitude is retained more bit, thus improve decoding performance.
Accompanying drawing explanation
Fig. 1 is the process flow figure of Soft Inform ation in the embodiment of the present invention;
Fig. 2 is the concrete process flow figure of Soft Inform ation in the embodiment of the present invention;
Fig. 3 is the processing unit structure chart of Soft Inform ation in the embodiment of the present invention;
Fig. 4 is the decoding performance simulation comparison figure of the processing method in the embodiment of the present invention.
Embodiment
The embodiment of the present invention gives a kind of processing method and device of the Soft Inform ation for decoding, uses the Soft Inform ation of the method and device process to carry out decoding, can improve decoding performance.Below in conjunction with Figure of description, the preferred embodiments of the present invention are described, should be appreciated that preferred embodiment described herein is only for instruction and explanation of the present invention, is not intended to limit the present invention.
Embodiments provide a kind of processing method of the Soft Inform ation for decoding, as shown in Figure 1, the method comprises the following steps:
S101: obtain the Soft Inform ation needed for decoding according to the signal received, and calculate the average of described Soft Inform ation;
S102: the average of described Soft Inform ation and default Soft Inform ation threshold value are compared, according to comparative result, described Soft Inform ation is moved to right;
S103: whether the numerical value of the Soft Inform ation after judging to move to right belongs to the input range of decoder, if do not belong to the input range of decoder, then the numerical value of the Soft Inform ation after described moving to right is set to the maximum in the input range of decoder or minimum value, as the numerical value after the Soft Inform ation process before moving to right; If belong to the input range of decoder, then the Soft Inform ation after described moving to right is intercepted from low level the number of bits that decoder allows input, as the numerical value after the Soft Inform ation process before moving to right;
S104: the numerical value after described Soft Inform ation process is supplied to decoder for decoding.
Preferably, according to comparative result, the implementation that described Soft Inform ation moves to right is comprised in described step S102:
Determine the number of bits that described Soft Inform ation moves to right; According to the number of bits that described Soft Inform ation moves to right, described Soft Inform ation is moved to right.
Preferably, the number of bits that in described step S102, Soft Inform ation moves to right is determined according to such as under type:
As E≤t
0time, move to right Soft Inform ation 0 bit;
Work as t
0≤ E≤t
1time, move to right Soft Inform ation 1 bit;
Work as t
1≤ E≤t
2time, move to right Soft Inform ation 2 bits;
Work as t
n-2≤ E≤t
n-1time, move to right Soft Inform ation n-1 bit;
Wherein, t0, t1, t2...tn-2, tn-1 represent Soft Inform ation threshold value, and n represents the number of Soft Inform ation threshold value, and the value of n is positive integer, and E represents the average of Soft Inform ation.
Preferably, the implementation of the average of described calculating Soft Inform ation comprises: the computing cycle of the average of setting Soft Inform ation, calculate the average of the Soft Inform ation in described computing cycle, then the account form of the average of described Soft Inform ation comprises: by the amplitude sum of the Soft Inform ation in the described computing cycle number divided by the Soft Inform ation in described computing cycle.
Preferably, the computing cycle of the average of described Soft Inform ation is the integral multiple in the processing time of a frame.
Preferably, the implementation of the step of the average of described Soft Inform ation and the relation of described Soft Inform ation threshold value comprises:
The average of the Soft Inform ation of the computing cycle of the average of a upper Soft Inform ation and the relation of described Soft Inform ation threshold value, or, the average of the Soft Inform ation of the computing cycle of the average of more current Soft Inform ation and the relation of described Soft Inform ation threshold value.
Preferably, also comprise before the described step that the average of described Soft Inform ation and default Soft Inform ation threshold value are compared:
Performance simulation is carried out to decoder, draws simulation result, according to described simulation result setting Soft Inform ation threshold value.
By adopting the method in the embodiment of the present invention, Soft Inform ation moves to right by the relation according to Soft Inform ation average and Soft Inform ation threshold value, and the input range of decoder whether is belonged to according to the numerical value of the Soft Inform ation after moving to right, bit truncation is carried out to Soft Inform ation, the interception way of Soft Inform ation can be decided according to actual conditions, can be the maximum in the input range of decoder or minimum value by the Soft Inform ation value of higher magnitude, and then the Soft Inform ation compared with small magnitude can be retained higher precision, thus be conducive to improving decoding performance.
Method in the embodiment of the present invention is described below in detail.Embodiments provide a kind of processing method of the Soft Inform ation for decoding, as shown in Figure 2, the method comprises the following steps:
S201: obtain the Soft Inform ation needed for decoding according to the baseband signal received and pilot signal.
S201a: extract the subcarrier that the decoding data block when pre-treatment needs from the baseband signal received.
For the baseband signal of ofdm system, according to the positional information in the baseband signal of user place, take out the subcarrier X={x with this user data
0, x
1... x
m-1(m is subcarrier number), obtain the subcarrier X when the decoding data block of pre-treatment needs;
S201b: the subcarrier that the decoding data block when pre-treatment needs is carried out to channel estimating and carries out channel compensation.
First, estimated value is got to the pilot signal received, obtains channel estimation value
Then, carry out channel compensation according to channel estimation value, be multiplied by channel estimation value with the subcarrier X that the decoding data block when pre-treatment needs
conjugation, obtain the subcarrier data Y after channel compensation;
S201c: demodulation is carried out to the subcarrier data Y after channel compensation and obtains Soft Inform ation;
In this step, when demodulation is carried out to the subcarrier data Y after channel compensation, corresponding demodulation method can be adopted according to the modulation system that subcarrier data adopts, demodulation mode is not specifically limited herein.
S202: the numerical value of setting Soft Inform ation threshold value T, and set the computing cycle of Soft Inform ation average E.
S202a: setting Soft Inform ation threshold value.
Before the numerical value of setting Soft Inform ation threshold value T, first performance simulation to be carried out to decoder, according to simulation result setting Soft Inform ation threshold value T, multinomial factor to be considered when setting, reach for the purpose of optimum to make the bit error rate performance of system, also can adopt other modes, such as, rule of thumb set Soft Inform ation threshold value.This threshold value T is used for determining Soft Inform ation bit truncation position.Such as, setting threshold T={t
0, t
1... t
n-1(n is threshold number), the Soft Inform ation average calculated within a upper Soft Inform ation mean value computation cycle is E, then when
E≤t
0time, move to right Soft Inform ation 0 bit, to move to right 0bit by Soft Inform ation;
T
0≤ E≤t
1time, move to right Soft Inform ation 1 bit, to move to right 1bit by Soft Inform ation;
T
1≤ E≤t
2time, move to right Soft Inform ation 2 bits, to move to right 2bit by Soft Inform ation;
T
n-2≤ E≤t
n-1time, move to right Soft Inform ation n-1 bit, to move to right n-1bit by Soft Inform ation.
In actual applications, optionally can decide the number of threshold value T, only can set a threshold value, move to right Soft Inform ation when average is greater than this threshold value n1bit, and move to right Soft Inform ation when average is less than this threshold value n2bit, also can set two or more threshold value.
S202b: the computing cycle of setting Soft Inform ation average E.
When setting the computing cycle of Soft Inform ation average E, also can there is multiple set-up mode, generally, the computing cycle of Soft Inform ation average E can be set to the integral multiple in the processing time of 1 frame.Such as, when the processing time of 1 frame is 5ms, the most I of computing cycle of Soft Inform ation average E is set to 5ms, and also the computing cycle of Soft Inform ation average E can be set to 20ms, the setting for the concrete time of the computing cycle of Soft Inform ation average E does not limit in this patent.Within the Soft Inform ation mean value computation cycle of setting, the account form calculating Soft Inform ation average is: by the amplitude sum of Soft Inform ation all in this cycle divided by data amount check, amplitude refers to the absolute value of the numerical value of Soft Inform ation, therefore, no matter Soft Inform ation numerical value is just or bears, the amplitude of Soft Inform ation is all positive number.During each intercepting Soft Inform ation bit, can decide Soft Inform ation bit truncation position in the current Soft Inform ation mean value computation cycle according to the Soft Inform ation average in a upper Soft Inform ation mean value computation cycle, and the Soft Inform ation average calculated within the current Soft Inform ation mean value computation cycle is for the foundation as the Soft Inform ation bit truncation in the next Soft Inform ation mean value computation cycle.
Certainly, also the computing cycle of Soft Inform ation average E can be set to the less time cycle, computing cycle by Soft Inform ation average E is set to the processing time of a frame, at this moment, the Soft Inform ation average in current calculation cycle can be calculated when a frame is disposed, thus utilize the Soft Inform ation average in current calculation cycle to carry out bit truncation to the Soft Inform ation in current calculation cycle.And, other mode can also be had when calculating Soft Inform ation average, such as, it not the computing cycle that setting is fixed, but according to the computing cycle of actual conditions dynamic conditioning Soft Inform ation average, or within the time period not etc., calculate Soft Inform ation average, in a word, as long as the object according to Soft Inform ation average adjustment Soft Inform ation bit truncation mode can be realized.
By calculating Soft Inform ation average, can when Soft Inform ation average be larger, when namely Soft Inform ation amplitude is larger, Soft Inform ation is moved to right more bit; And when Soft Inform ation average is less, when namely Soft Inform ation amplitude is less, Soft Inform ation is moved to right less bit, thus better retain the precision of the less Soft Inform ation of amplitude.The number of bits that Soft Inform ation moves to right is determined according to Soft Inform ation average, can according to the situation dynamic conditioning bit truncation mode of Soft Inform ation, thus realize best decoding performance, and, because the Soft Inform ation mean value computation cycle calculates for the Soft Inform ation in a period of time, so when the data of indivedual Soft Inform ation occur abnormal, because the quantity of Soft Inform ation is very large within this period of time, namely sample size is very large, so the data exception of indivedual Soft Inform ation can't produce too much influence to the Soft Inform ation mean value computation result of entirety.
S203: calculate the Soft Inform ation average in the current Soft Inform ation mean value computation cycle, and the Soft Inform ation threshold value of Soft Inform ation average and setting is compared, determine according to comparative result the mode that current decoding data block Soft Inform ation moves to right, and bit truncation is carried out to the Soft Inform ation after moving to right.
S203a: calculate the Soft Inform ation average in the current Soft Inform ation mean value computation cycle.
In this step, the amplitude of Soft Inform ation and data amount check in the cumulative current Soft Inform ation mean value computation cycle, to calculate the Soft Inform ation average in the current Soft Inform ation mean value computation cycle when current Soft Inform ation mean value computation end cycle, this Soft Inform ation average may be used for as determining in the next Soft Inform ation mean value computation cycle that Soft Inform ation moves to right the foundation of mode.Certainly, also can when current Soft Inform ation mean value computation end cycle, directly utilize the Soft Inform ation average in the current Soft Inform ation mean value computation cycle as determining in the current Soft Inform ation mean value computation cycle that Soft Inform ation moves to right the foundation of mode.
S203b: Soft Inform ation average and Soft Inform ation threshold value are compared, according to comparative result, Soft Inform ation is moved to right.
The Soft Inform ation threshold value of Soft Inform ation average and setting compared, here, the Soft Inform ation average can used in a computing cycle compares, and also can directly compare by the Soft Inform ation average in current calculation cycle.Soft Inform ation bit truncation is carried out according to comparative result.Suppose that Soft Inform ation average is E, Soft Inform ation threshold value T is n-1, i.e. T={t
0, t
1... t
n-1, then when
E≤t
0time, move to right Soft Inform ation 0 bit, to move to right 0bit by Soft Inform ation;
T
0≤ E≤t
1time, move to right Soft Inform ation 1 bit, to move to right 1bit by Soft Inform ation;
T
1≤ E≤t
2time, move to right Soft Inform ation 2 bits, to move to right 2bit by Soft Inform ation;
T
n-2≤ E≤t
n-1time, move to right Soft Inform ation n-1 bit, to move to right n-1bit by Soft Inform ation.
Such as, the Soft Inform ation mean value computation result in a upper Soft Inform ation mean value computation cycle is 4237, and the threshold value of setting is t
5=4096, t
6=4980, because t
5≤ E≤t
6, so 6bit that Soft Inform ation is moved to right.
S203c: whether the numerical value of the Soft Inform ation after judging to move to right belongs to the input range of decoder, and processes Soft Inform ation according to judged result.
Suppose that the input range of decoder is [-128,127], if the numerical value of the Soft Inform ation after then moving to right does not belong to the input range of decoder, namely this numerical value be still greater than decoder allow input range in maximum or be less than decoder allow input range in minimum value, then the numerical value of this Soft Inform ation is set to the maximum in the input range of decoder or minimum value.Such as, the Soft Inform ation intercepted after 6bit if move to right is 129, then the numerical value of this Soft Inform ation is set to the maximum 127 in the input range of decoder, the Soft Inform ation intercepted after 6bit if move to right is-129, then the numerical value of this Soft Inform ation is set to the minimum value-128 in the input range of decoder; The Soft Inform ation obtained after 6bit if moved to right belongs to the input range that decoder allows, then this Soft Inform ation value is the numerical value moved to right after 6bit, and such as, if the Soft Inform ation intercepted after the 6bit that moves to right is 126, then the numerical value of this Soft Inform ation is 126.The binary system Soft Inform ation of two 16bit mentioned for background technology part of the present invention more below carries out bit truncation, such as 0000000011000001 and 0000000000000011 these two Soft Inform ation are all 16bit, and decoder allows the number of bits of input to be 8bit, namely the number range allowing the Soft Inform ation inputted is [-128, 127], supposed according to the comparative result between the Soft Inform ation average in the upper Soft Inform ation mean value computation cycle and the Soft Inform ation threshold value of setting, need to move to right 1 bit to these two Soft Inform ation, symbol positive number is had for a 16bit, the bit that moves to right refers to: the number of bits of this positive number is constant, be still 16bit, move to right the numerical value on each bit a bit successively, like this, numerical value originally on lowest order is given up, and in original highest order, fill one zero, then 0000000011000001 move to right after a bit and become 0000000001100000, and 0000000000000011 moves to right after a bit and becomes 0000000000000001, because the numerical value of these two Soft Inform ation after moving to right all does not exceed the input range that decoder allows, so directly intercept to the Soft Inform ation after moving to right the number of bits that decoder allows input from low level, after being about to move to right 0000000001100000 directly from low level, intercept 8 bits after be 01100000, and be 00000001 after 0000000000000001 after moving to right directly is intercepted 8 bits from low level, this shows, adopt the mode in the embodiment of the present invention, the precision remaining the less Soft Inform ation of amplitude as much as possible, be compared with in the of 00000000 with directly these two Soft Inform ation all being intercepted in prior art, more be conducive to improving decoding performance.Be described for the Soft Inform ation that amplitude is larger again, such as 16bit Soft Inform ation 0110111100110001, suppose it also to move to right 1bit, 0011011110011000 is become after then moving to right, because the data after moving to right still exceed the input range [-128 that decoder allows, 127], then the direct value of the Soft Inform ation after this being moved to right is the maximum 127 in the input range of decoder permission, and being converted to binary representation is 01111111.Because the numerical value of this Soft Inform ation is inherently very large, namely the bit representing representated by this Soft Inform ation be 0 probability be almost 100%, so we and the concrete numerical value being indifferent to this Soft Inform ation is how many, only needing is directly the maximum 127 in the input range of decoder permission by its value, namely represent that the probability of the bit 0 representated by this Soft Inform ation is 100%, do not lose the precision of this Soft Inform ation.
S204: the Soft Inform ation input decoder after bit truncation is carried out decoding.
By presetting Soft Inform ation threshold value in the embodiment of the present invention, and carry out dynamic conditioning Soft Inform ation bit truncation mode according to the relation of the Soft Inform ation average calculated and Soft Inform ation threshold value, when Soft Inform ation average is larger, move to right Soft Inform ation more bit, when Soft Inform ation average is less, move to right Soft Inform ation less bit, and, the numerical value still exceeding the Soft Inform ation of decoder input range after moving to right is set to the maximum in the input range of decoder or minimum value, thus can decoding performance be improved.Because, just bit representated by Soft Inform ation is the probability of 0 or 1, do not require that the amplitude inputted is absolutely correct, therefore, when the amplitude of Soft Inform ation is very large, represent this bit be 0 or 1 probability very large, such as, when the input range of decoder is [-128, 127] time, if the Soft Inform ation numerical value of input is 129, the bit then representing representated by this Soft Inform ation be 0 probability be 100% substantially, so the maximum 127 be set in the input range of decoder does not lose any information by the numerical value of this Soft Inform ation, and, for the Soft Inform ation that amplitude is less, moved to right less bit, thus remain the precision of the less Soft Inform ation of amplitude as much as possible.So the Soft Inform ation bit truncation method in the application embodiment of the present invention can improve decoding performance.
Embodiments provide a kind of processing unit of the Soft Inform ation for decoding, as shown in Figure 3, comprising:
Average calculation unit 31, for obtaining the Soft Inform ation needed for decoding according to the signal received, and calculates the average of described Soft Inform ation;
Relatively shift unit 32, with average calculation unit 31 with judge that processing unit 33 is connected, for the average of described Soft Inform ation and default Soft Inform ation threshold value being compared, moves to right described Soft Inform ation according to comparative result;
Judge processing unit 33, whether the numerical value for the Soft Inform ation after judging to move to right belongs to the input range of decoder, if do not belong to the input range of decoder, then the numerical value of the Soft Inform ation after described moving to right is set to the maximum in the input range of decoder or minimum value, as the numerical value after the Soft Inform ation process before moving to right; If belong to the input range of decoder, then the Soft Inform ation after described moving to right is intercepted from low level the number of bits that decoder allows input, as the numerical value after the Soft Inform ation process before moving to right;
Data transmission unit 34, and judges that processing unit 33 is connected with decoder, for the numerical value after described Soft Inform ation process is supplied to decoder for decoding.
Preferably, described compare shift unit 32 for:
Determine the number of bits that described Soft Inform ation moves to right;
According to the number of bits that described Soft Inform ation moves to right, described Soft Inform ation is moved to right.
Preferably, described compare shift unit 32 for:
As E≤t
0time, move to right Soft Inform ation 0 bit;
Work as t
0≤ E≤t
1time, move to right Soft Inform ation 1 bit;
Work as t
1≤ E≤t
2time, move to right Soft Inform ation 2 bits;
Work as t
n-2≤ E≤t
n-1time, move to right Soft Inform ation n-1 bit;
Wherein, t0, t1, t2...tn-2, tn-1 represent Soft Inform ation threshold value, and n represents the number of Soft Inform ation threshold value, and the value of n is positive integer, and E represents the average of Soft Inform ation.
Preferably, described average calculation unit 31 is for the computing cycle of average of setting Soft Inform ation, calculate the average of the Soft Inform ation in described computing cycle, the average of described Soft Inform ation equals the number of amplitude sum divided by the Soft Inform ation in described computing cycle of the Soft Inform ation in described computing cycle.
Preferably, described compare shift unit 32 for:
The average of the Soft Inform ation of the computing cycle of the average of a upper Soft Inform ation and the relation of described Soft Inform ation threshold value, or,
The average of the Soft Inform ation of the computing cycle of the average of more current Soft Inform ation and the relation of described Soft Inform ation threshold value.
Preferably, the processing unit of the described Soft Inform ation for decoding, also comprises:
Threshold setting unit, for carrying out performance simulation to decoder, draws simulation result, according to described simulation result setting Soft Inform ation threshold value.
By adopting the device in the embodiment of the present invention, Soft Inform ation moves to right by the relation according to Soft Inform ation average and Soft Inform ation threshold value, and the input range of decoder whether is belonged to according to the numerical value of the Soft Inform ation after moving to right, bit truncation is carried out to Soft Inform ation, the interception way of Soft Inform ation can be decided according to actual conditions, can be the maximum in the input range of decoder or minimum value by the Soft Inform ation value of higher magnitude, and then the Soft Inform ation compared with small magnitude can be retained higher precision, thus be conducive to improving decoding performance.
Fig. 4 adopts in prior art directly to intercept in Soft Inform ation higher bit position and the present invention dynamically to intercept the simulation performance comparison diagram that these two kinds of methods of Soft Inform ation bit carry out decoding respectively.What a curve below in Fig. 4 represented is the performance curve adopting the method directly intercepting Soft Inform ation higher bit position in prior art to carry out decoding, a curve is above the performance curve adopting the method for the dynamic intercepting Soft Inform ation bit in the present invention to carry out decoding, as can be seen from the figure when signal to noise ratio is identical, the error rate of carrying out decoding by the method in the present invention is more much lower than the method for prior art, and namely decoding performance is significantly improved.
Obviously, those skilled in the art can carry out various change and modification to the present invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.
Claims (9)
1. for a processing method for the Soft Inform ation of decoding, it is characterized in that, the method comprises the following steps:
Obtain the Soft Inform ation needed for decoding according to the signal received, and calculate the average of described Soft Inform ation;
The average of described Soft Inform ation and default Soft Inform ation threshold value are compared, according to comparative result, described Soft Inform ation is moved to right;
Whether the numerical value of the Soft Inform ation after judging to move to right belongs to the input range of decoder, if do not belong to the input range of decoder, then the numerical value of the Soft Inform ation after described moving to right is set to the maximum in the input range of decoder or minimum value, as the numerical value after the Soft Inform ation process before moving to right; If belong to the input range of decoder, then the Soft Inform ation after described moving to right is intercepted from low level the number of bits that decoder allows input, as the numerical value after the Soft Inform ation process before moving to right;
Numerical value after described Soft Inform ation process is supplied to decoder for decoding;
Wherein, describedly according to comparative result, described Soft Inform ation to be moved to right, comprising:
Determine the number of bits that described Soft Inform ation moves to right;
According to the number of bits that described Soft Inform ation moves to right, described Soft Inform ation is moved to right;
Wherein, the number of bits that moves to right of described Soft Inform ation is determined according to such as under type:
As E≤t
0time, move to right Soft Inform ation 0 bit;
Work as t
0<E≤t
1time, move to right Soft Inform ation 1 bit;
Work as t
1<E≤t
2time, move to right Soft Inform ation 2 bits;
......
Work as t
n-2<E≤t
n-1time, move to right Soft Inform ation n-1 bit,
Wherein, t
0, t
1, t
2t
n-2, t
n-1represent Soft Inform ation threshold value, n represents the number of Soft Inform ation threshold value, and the value of n is positive integer, and E represents the average of Soft Inform ation.
2. the method for claim 1, it is characterized in that, the implementation of the average of described calculating Soft Inform ation comprises: the computing cycle of the average of setting Soft Inform ation, calculate the average of the Soft Inform ation in described computing cycle, then the account form of the average of described Soft Inform ation comprises: by the amplitude sum of the Soft Inform ation in the described computing cycle number divided by the Soft Inform ation in described computing cycle.
3. method as claimed in claim 2, it is characterized in that, the computing cycle of the average of described Soft Inform ation is the integral multiple in the processing time of a frame.
4. method as claimed in claim 2, it is characterized in that, the implementation of the step of the average of described Soft Inform ation and the relation of described Soft Inform ation threshold value comprises:
The average of the Soft Inform ation of the computing cycle of the average of a upper Soft Inform ation and the relation of described Soft Inform ation threshold value, or,
The average of the Soft Inform ation of the computing cycle of the average of more current Soft Inform ation and the relation of described Soft Inform ation threshold value.
5. the method for claim 1, is characterized in that, also comprises before the described step average of described Soft Inform ation and default Soft Inform ation threshold value compared:
Performance simulation is carried out to decoder, draws simulation result, according to described simulation result setting Soft Inform ation threshold value.
6. for a processing unit for the Soft Inform ation of decoding, it is characterized in that, comprising:
Average calculation unit, for obtaining the Soft Inform ation needed for decoding according to the signal received, and calculates the average of described Soft Inform ation;
Relatively shift unit, for the average of described Soft Inform ation and default Soft Inform ation threshold value being compared, moves to right described Soft Inform ation according to comparative result;
Judge processing unit, whether the numerical value for the Soft Inform ation after judging to move to right belongs to the input range of decoder, if do not belong to the input range of decoder, then the numerical value of the Soft Inform ation after described moving to right is set to the maximum in the input range of decoder or minimum value, as the numerical value after the Soft Inform ation process before moving to right; If belong to the input range of decoder, then the Soft Inform ation after described moving to right is intercepted from low level the number of bits that decoder allows input, as the numerical value after the Soft Inform ation process before moving to right;
Data transmission unit, for being supplied to decoder for decoding by the numerical value after described Soft Inform ation process;
Wherein, described compare shift unit for:
Determine the number of bits that described Soft Inform ation moves to right; According to the number of bits that described Soft Inform ation moves to right, described Soft Inform ation is moved to right;
Wherein, described compare shift unit for:
As E≤t
0time, move to right Soft Inform ation 0 bit; Work as t
0<E≤t
1time, move to right Soft Inform ation 1 bit; Work as t
1<E≤t
2time, move to right Soft Inform ation 2 bits; ... work as t
n-2<E≤t
n-1time, move to right Soft Inform ation n-1 bit, wherein, and t0, t1, t2 ... tn-2, tn-1 represent Soft Inform ation threshold value, and n represents the number of Soft Inform ation threshold value, and the value of n is positive integer, and E represents the average of Soft Inform ation.
7. device as claimed in claim 6, it is characterized in that, described average calculation unit is used for: the computing cycle of the average of setting Soft Inform ation, calculate the average of the Soft Inform ation in described computing cycle, the average of described Soft Inform ation equals the number of amplitude sum divided by the Soft Inform ation in described computing cycle of the Soft Inform ation in described computing cycle.
8. device as claimed in claim 7, is characterized in that, described compare shift unit for:
The average of the Soft Inform ation of the computing cycle of the average of a upper Soft Inform ation and the relation of described Soft Inform ation threshold value, or, the average of the Soft Inform ation of the computing cycle of the average of more current Soft Inform ation and the relation of described Soft Inform ation threshold value.
9. device as claimed in claim 6, is characterized in that, also comprise:
Threshold setting unit, for carrying out performance simulation to decoder, draws simulation result, according to described simulation result setting Soft Inform ation threshold value.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010187352.4A CN102263563B (en) | 2010-05-27 | 2010-05-27 | A kind of processing method of the Soft Inform ation for decoding and device |
| PCT/CN2010/077313 WO2011147146A1 (en) | 2010-05-27 | 2010-09-26 | Processing method and device of soft information for decoding |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010187352.4A CN102263563B (en) | 2010-05-27 | 2010-05-27 | A kind of processing method of the Soft Inform ation for decoding and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102263563A CN102263563A (en) | 2011-11-30 |
| CN102263563B true CN102263563B (en) | 2016-03-30 |
Family
ID=45003238
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010187352.4A Active CN102263563B (en) | 2010-05-27 | 2010-05-27 | A kind of processing method of the Soft Inform ation for decoding and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN102263563B (en) |
| WO (1) | WO2011147146A1 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117439849B (en) * | 2023-12-19 | 2024-03-15 | 深圳市国电科技通信有限公司 | Signal demodulation method, device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002023787A3 (en) * | 2000-09-12 | 2002-05-30 | Ericsson Inc | Method and apparatus for soft information generation in joint demodulation of co-channel signals |
| CN101222471A (en) * | 2008-02-04 | 2008-07-16 | 北京北方烽火科技有限公司 | Base band link scaling method suitable for IEEE802.16e base station receiver |
| CN101262231A (en) * | 2008-04-25 | 2008-09-10 | 浙江大学 | A decoding method for block low-density check code and reconstruction of multi-mode decoder |
| CN101534127A (en) * | 2009-04-20 | 2009-09-16 | 西安新邮通信设备有限公司 | Encoding and decoding method for improving decoding efficiency by pilot frequency information and device thereof |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100373840C (en) * | 2005-04-01 | 2008-03-05 | 东南大学 | Method and apparatus for detecting normalized iterative soft interference cancelling signal |
-
2010
- 2010-05-27 CN CN201010187352.4A patent/CN102263563B/en active Active
- 2010-09-26 WO PCT/CN2010/077313 patent/WO2011147146A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002023787A3 (en) * | 2000-09-12 | 2002-05-30 | Ericsson Inc | Method and apparatus for soft information generation in joint demodulation of co-channel signals |
| CN101222471A (en) * | 2008-02-04 | 2008-07-16 | 北京北方烽火科技有限公司 | Base band link scaling method suitable for IEEE802.16e base station receiver |
| CN101262231A (en) * | 2008-04-25 | 2008-09-10 | 浙江大学 | A decoding method for block low-density check code and reconstruction of multi-mode decoder |
| CN101534127A (en) * | 2009-04-20 | 2009-09-16 | 西安新邮通信设备有限公司 | Encoding and decoding method for improving decoding efficiency by pilot frequency information and device thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2011147146A1 (en) | 2011-12-01 |
| CN102263563A (en) | 2011-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1489774A2 (en) | System and method to determinate a bit error probability of received data within a cellular wireless network | |
| MXPA01003389A (en) | Channel estimator with variable number of taps. | |
| JPWO2006098105A1 (en) | Adaptive modulation control system and wireless communication device | |
| CN1402924A (en) | Power efficient equalization | |
| CN101568145A (en) | LTE system CQI reporting implementation method | |
| KR101739153B1 (en) | Process for controling the hsdpa downlink data rate received by a user equipment | |
| CN101567752B (en) | Self-adaptive encoding/decoding method based on low-density parity-check code | |
| CN109348515B (en) | Method and device for adjusting transmission rate in wireless network | |
| CN102263563B (en) | A kind of processing method of the Soft Inform ation for decoding and device | |
| CN109088836A (en) | The data block building method of single carrier frequency domain equalization SOQPSK-TG signal | |
| CN116032702B (en) | Adaptive channel estimation method, apparatus, computer device and storage medium | |
| CN109150783A (en) | A kind of channel estimation methods and device | |
| CN1917488A (en) | Bit error probability estimating system and related method thereof | |
| JP4189389B2 (en) | Equalizer and equalization method | |
| CN107205273B (en) | DCI blind detection data processing method and device | |
| CN111628953B (en) | Method for reducing peak-to-average ratio of OFDM signal | |
| CN1239033C (en) | Discrete Fourier transform based space-time combined inspecting device and method for radio transmission | |
| CN102769600A (en) | Method and device for zoom factor demodulation in orthogonal frequency division multiplexing (OFDM) system receiving machine | |
| CN102577531B (en) | Method, device and base station for allocating power | |
| CN101057468B (en) | Method and device for detecting high-mobility state of mobile terminal and related equipment | |
| CN100583667C (en) | Self-adaptive balancer | |
| CN1599364A (en) | Data equalization method for burst communication | |
| US20090213920A1 (en) | Method and apparatus for detecting high-mobility state of mobile terminal and related device | |
| WO2009077405A1 (en) | Methods and arrangements for channelization code selection | |
| TWI593257B (en) | Device and method of handling sequence estimation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |