The method of estimation of receiver delay spreading parameter and delay spreading parameter estimation unit
Technical field
The present invention relates to the communications field, particularly a kind of method of estimation of receiver delay spreading parameter and delay spreading parameter estimation unit.
Background technology
In communication system,, just need carry out rational channel estimating for the data of being transmitted in the receiving terminal right demodulation.At present, be used to estimate that the method for wireless channel has two types greatly, one type is based on training sequence, and one type is based on pilot tone.
Be the channel estimating of example explanation with DTMB (Digital Terrestrial/Television Multimedia Broadcasting, digital TV ground multimedia broadcasting) based on training sequence.
At first specifically introduce frame structure and frame head sequence in the DTMB system, the frame head PN sequence as training sequence in the DTMB standard has three kinds: PN420, PN595 and PN945.
Fig. 1 has provided four layers of frame structure based on multi-frame of DTMB system.Signal frame is the most basic transmission unit, comprises frame head and frame.Frame head is filled the PN sequence, and frame can be the data of single carrier mode, also can be the data of multi-carrier mode.Under multi-carrier mode, the size of IFFT (Inverse Fast FourierTransform, fast adverse Fourier transform) is 3780 points.The length of frame head PN sequence has three kinds: 420 data symbols, 595 data symbols and 945 data symbols, thus there are three kinds of signal frame length accordingly.The baseband sampling rate of DTMB system is 7.56MSPS (Mega-Samples-Per-Second, per second sample 1,000,000 times).Therefore, the time span of three kinds of corresponding PN frame head sequences is respectively 420/7.56=55.56 microsecond, 595/7.56=78.703 microsecond and 945/7.56=125 microsecond.And the time span of frame is 500 microseconds.Thereby the time span of each signal frame is respectively 555.56 microseconds, 578.703 microseconds and 625 microseconds accordingly.Be superframe on signal frame, the time span of each superframe is unified to be 125ms.Accordingly, a superframe comprises 225,216 and 200 signal frames respectively.For dividing frame, time span is one minute on the superframe.A branch frame comprises 480 superframes.Up be a day frame again, corresponding one day 24 hours.One day, frame comprised 1440 branch frames.
The building method of three kinds of PN frame head sequences is also inconsistent.Fig. 2 has provided the structure chart of frame head PN420 sequence.The centre is the PN255 sequence, is the m sequence on one 8 rank, and reaching " 1 " through " 0 " to+1 value is the binary character of non-return-to-zero to the mapping transformation of-1 value.The PN420 sequence comprises synchronizing sequence (length is 83 symbols) behind a preamble sequence (length is 82 symbols), PN255 sequence and.Preamble sequence is the Cyclic Prefix sequence of PN255 sequence, and back synchronizing sequence is the cyclic suffix sequence of PN255 sequence.The PN255 sequence adopts LFSR (Linear Feedback Shift Register, linear feedback shift register) to generate, and the initial phase of different LFSR will generate different PN255 sequences.
Fig. 3 has provided the structure of frame head PN945 sequence.The structure of PN945 and the similar of PN420 all have preamble sequence and back synchronizing sequence.The intermediate sequence of PN945 sequence is a PN511 sequence, adopts the m sequence on one 9 rank to generate, and the mapping transformation to-1 value is the binary symbol sequence of non-return-to-zero to+1 value and " 1 " through " 0 " again.The concrete value of PN945 sequence also is variable, and is relevant with the initial phase of LFSR.And in order to reduce adjacent PN420 sequence or the correlation between the PN945 sequence, the DTMB standard has been arranged initial phase once through Computer Simulation specially meticulously.
The building method of frame head PN420 and PN945 is the same, but the building method of the building method of frame head PN595 and PN420 and PN945 differs widely.The PN595 sequence is to adopt 10 rank maximum length pseudo-random binary sequence brachymemmas to form, and is that length is preceding 595 chips of 1023 m sequence.And the initial phase of LFSR that generates the m sequence of this 1023 length is also fixed, and is 0000000001, and promptly the PN595 sequence that adopts of each signal frame is identical.Preceding 595 chips of pseudo random sequence, the mapping transformation to-1 value is the binary symbol sequence of non-return-to-zero to+1 value and " 1 " through " 0 ", is PN595 frame head sequence.
Receiver is after receiving signal; The supposition known transmitter is launched which kind of frame head sequence (one of three kinds of frame head sequence PN420, PN595 and PN945) earlier; On receiver, utilize frame head sequence that is stored in local correspondence in advance and the signal that receives to carry out cross-correlation operation then; Seek the output result's of cross-correlation operation amplitude peak value; And then definite original position that receives PN frame head in the signal, know the frame head PN pattern (one of PN420, PN595 or PN945) and the frame head PN sequence of transmitter emission, and estimate carrier frequency offset; After this on these results' basis, also need carry out more meticulous channel estimating.
Time delay expansion is estimated follow-up receiver channel and the training sequence elimination all can be influential; Especially for DTMB etc. because channel quality is relatively poor relatively, the time delay expansion of multipath channel can relatively large system, time delay expansion can exert an influence to estimated result especially.
Summary of the invention
The technical problem that the present invention will solve provides a kind of method of estimation and delay spreading parameter estimation unit of receiver delay spreading parameter; Be applicable to that receiver can confirm the communication system of training sequence; Can carry out the estimation of delay spreading parameter, thereby accomplish more precise channels estimation.
In order to address the above problem, the invention provides a kind of method of estimation of receiver delay spreading parameter, comprising:
(A) reception signal and training sequence through the carrier frequency offset compensation are carried out cross-correlation operation, obtain correlated series;
(B) judge through thresholding, find out in this correlated series the location index value I in last footpath
FirstLocation index value I with last footpath
Last
(C) delay spreading parameter is I
LastDeduct I
FirstAfter poor.
Further, steps A specifically comprises:
Intercepting one piece of data sequence is as pending data sequence r from the reception signal that compensates through carrier frequency offset
Pn(n); With training sequence as secondary processing sequence s
Pn(m); Pending data sequence and secondary processing sequence are carried out cross-correlation operation, obtain correlated series.
Further, in the steps A:
Pending data sequence r
Pn(n) length is 2 * L
Ds, max+ P, n are the index of sample of signal, and the value of n is from L
Sync-L
Ds, maxTo I
Sync+ L
Ds, maxEach integer of+P-1 comprises I
Sync-L
Ds, maxAnd I
Sync+ L
Ds, max+ P-1;
Wherein P is the length of training sequence, and frame synchronization position index is I
Sync, the requirement of the maximum delay expansion of receiver is L
Ds, max
Further, when this method is applied to the DTMB system, in the steps A training sequence is meant as secondary processing sequence: choose local corresponding PN sequence of being preserved as secondary processing sequence s according to frame head mode that estimates and PN sequence phase
Pn(m), this sequence length is P, and wherein the value of index m is each integer from 0 to P-1, comprises 0 and P-1.
Further, when this method is applied to the DTMB system, when frame head PN pattern is PN420, P=420; When frame head PN pattern is PN595, P=595; When frame head PN pattern is PN945, P=945.
Further, to sequence r
Pn(n) and sequence s
Pn(m) do cross-correlation operation according to following formula, obtain correlated series s
Corr(l):
Wherein conj () expression complex conjugate operation, the value of l is to 2 * L from 0
Ds, maxEach integer, comprise 0 and 2 * L
Ds, max
Further, step B specifically comprises:
(B1) the power factor threshold T is set
Power
(B2) ask the power sequence of correlated series;
(B3) ask the average power P of power sequence
Corr, avg
(B4) power sequence P is judged on sample-by-sample ground from front to back
CorrWhether each sample (l) is greater than T
Power* P
Corr, avgWherein, greater than T
Power* P
Corr, avgThe index value of first sample be I
First, greater than T
Power* P
Corr, avgThe index value of last sample be I
Last
Further, when this method is applied to the DTMB system, among the step B1, T
PowerBe 16.
Further, step B4 specifically comprises:
(91) S set of structure, sample-by-sample ground judges that whether each sample in the power sequence is greater than T from front to back
Power* P
Corr, avgIf greater than, then the index value of this sample is the element of S set just; Otherwise, be not the element of S set;
(92) if S set for null set, then first element value of S set is the location index value in last footpath, is designated as I
FirstLast element value of S set is the location index value in last footpath, is designated as I
Last
The present invention also provides a kind of receiver delay spreading parameter estimation unit, comprising:
The correlated series generation module is used to receive the reception signal through the carrier frequency offset compensation, and itself and training sequence are carried out cross-correlation operation, obtains correlated series;
The thresholding judge module is used for judging through thresholding, finds out in the said correlated series the location index value I in last footpath
FirstLocation index value I with last footpath
Last
The delay spreading parameter computing module is used to receive I
FirstAnd I
Last, and use I
LastDeduct I
FirstObtain delay spreading parameter L
Ds
Further, said correlated series generation module comprises:
Pending sequence generates submodule, be used to receive reception signal through the carrier frequency offset compensation, and therefrom intercepting one piece of data sequence as pending data sequence r
Pn(n);
The secondary processing sequence generation module is used to provide training sequence as secondary processing sequence s
Pn(m);
The cross-correlation operation submodule is used for pending data sequence r
Pn(n) and secondary processing sequence s
Pn(m) carry out cross-correlation operation, obtain correlated series s
Corr(l).
Further, said pending sequence generates the pending data sequence r that submodule generates
Pn(n) length is 2 * L
Ds, max+ P, wherein n is the index of sample of signal, the value of n is from L
Sync-L
Ds, maxTo I
Sync+ L
Ds, maxEach integer of+P-1 comprises I
Sync-L
Ds, maxAnd I
Sync+ L
Ds, max+ P-1;
Wherein P is the length of frame head PN sequence, and frame synchronization position index is I
Sync, the requirement of the maximum delay expansion of receiver is L
Ds, max
Further; When this estimation unit was applied to the DTMB system, said secondary processing sequence generates submodule provided training sequence to be meant as secondary processing sequence: secondary processing sequence generates submodule and chooses local corresponding PN sequence of being preserved as secondary processing sequence s according to frame head mode that estimates and PN sequence phase
Pn(m), this sequence length is P, and wherein the value of index m is each integer from 0 to P-1, comprises 0 and P-1.
Further, when this estimation unit is applied to the DTMB system, when frame head PN pattern is PN420, P=420; When frame head PN pattern is PN595, P=595; When frame head PN pattern is PN945, P=945.
Further, said cross-correlation operation submodule carries out cross-correlation operation to two sequences according to following formula, obtains correlated series s
Corr(l):
Wherein conj () expression complex conjugate operation, the value of l is to 2 * L from 0
Ds, maxEach integer, comprise 0 and 2 * L
Ds, max
Further, said thresholding judge module comprises:
The threshold value sub module stored is used to store set power factor threshold T
Power
Power sequence generates submodule, is used to ask the power sequence of correlated series;
Average power generates submodule, is used to ask the average power P of power sequence
Corr, avg
Index value is judged submodule, and whether each sample that is used for sample-by-sample ground from front to back judging power sequence is greater than T
Power* P
Corr, avg, and will be greater than T
Ower* P
Corr, avgThe index value of first sample be designated as I
First, will be greater than T
Power* P
Corr, avgThe index value of last sample be designated as I
Last
Further, when this method was applied to the DTMB system, said threshold value was provided with the set power factor threshold T of submodule
PowerBe 16.
Further, said index value is judged submodule sample-by-sample ground judgement from front to back power sequence P
CorrWhether each sample (l) is greater than T
Power* P
Corr, avg, and will be greater than T
Power* P
Corr, avgThe index value of first sample be designated as I
First, will be greater than T
Power* P
Corr, avgThe index value of last sample be designated as I
LastSpecifically be meant:
Index value is judged S set of submodule structure, and power sequence P is judged on sample-by-sample ground from front to back
CorrWhether each sample (l) is greater than T
Power* P
Corr, avgIf greater than, the index value of this sample is designated as the element of S set; Otherwise, be not designated as the element of S set; If S set is not null set, then index value judges that submodule is designated as I with first element value of S set
First, this is the location index value in last footpath, and last element value is designated as I
Last, this is the location index value in last footpath.
The present invention is directed to receiver and can confirm the communication system of training sequence, proposed a kind of receiver and carried out the scheme that delay spreading parameter is estimated, can make receiver obtain comparatively accurate time delay expansion.Further, the present invention has provided the processing procedure of detailed cross-correlation operation process, thresholding judgement, and the processing details of estimation scheme has been carried out the refinement description.Further, the present invention also is directed against DTMB system and other and the similar system of DTMB and has provided concrete parameter value.In addition, the present invention gives preferred power factor threshold value, makes the reliability of estimation scheme stronger.
Description of drawings
Fig. 1 is based on the sketch map of four layers of frame structure of multi-frame in the existing DTMB system;
Fig. 2 is the structural representation of the frame head PN420 sequence in the existing DTMB system;
Fig. 3 is the structural representation of the frame head PN945 sequence in the existing DTMB system;
Fig. 4 is the practical implementation flow chart of the method for estimation of receiver delay spreading parameter of the present invention;
Fig. 5 is in the method for estimation of receiver delay spreading parameter of the present invention, carries out the practical implementation flow chart that thresholding is judged;
Fig. 6 is the practical implementation block diagram of the estimation unit of receiver delay spreading parameter of the present invention;
Fig. 7 is in the estimation unit of receiver delay spreading parameter of the present invention, the practical implementation block diagram of correlated series generation module;
Fig. 8 is in the estimation unit of receiver delay spreading parameter of the present invention, the practical implementation block diagram of thresholding judge module.
Embodiment
To combine accompanying drawing and embodiment that technical scheme of the present invention is explained in more detail below.
When receiver obtains the frame synchronization position, knows training sequence, and estimate carrier frequency offset, carried out carrying out the estimation of delay spreading parameter after the compensation of carrier frequency offset to received signal; The length of note training sequence is P baseband sample point; Frame synchronization position (this frame synchronization position has identified the starting point of the frame head PN sequence) index that the Synchronous Processing step obtains before the note is L
SyncThe requirement that the note system docking is received the maximum delay expansion of machine is L
Ds, maxRemember under the current multipath channel that the delay spreading parameter that will estimate is L
Ds
The method of estimation of the receiver delay spreading parameter that the present invention proposes, as shown in Figure 4, comprising:
A, reception signal and training sequence that the process carrier frequency offset is compensated carry out cross-correlation operation, obtain correlated series;
B, judge, find out in this correlated series the location index value I in last footpath through thresholding
FirstLocation index value L with last footpath
Last
C, delay spreading parameter are I
LastDeduct I
FirstAfter poor, i.e. L
Ds=I
Last-I
First
Wherein, steps A specifically comprises:
Intercepting one piece of data sequence is as pending data sequence r from the reception signal that compensates through carrier frequency offset
Pn(n); With training sequence as secondary processing sequence s
Pn(m); Pending data sequence and secondary processing sequence are carried out cross-correlation operation, obtain correlated series.
Wherein, in the steps A, training sequence can but be not limited to frame head PN sequence, also can be other sequence, as complete 0 or complete 1 sequence etc.
Wherein, in the steps A, the pending data sequence r of institute's intercepting
Pn(n) length is 2 * L
Ds, max+ P, wherein n is the index of sample of signal, the value of n is from I
Sync-L
Ds, maxTo I
Sync+ L
Ds, maxEach integer of+P-1 comprises I
Sync-L
Ds, maxAnd I
Sync+ L
Ds, max+ P-1; That is:
n=I
sync-L
ds,max,I
sync-L
ds,max+1,……,I
sync+L
ds,max+P-1。
Can know that secondary processing sequence length is P, wherein the value of index m is each integer from 0 to P-1, comprises 0 and P-1, that is: m=0, and 1 ..., P-1.
When this method is applied to similar system of DTMB system or other and DTMB, when frame head PN pattern is PN420, P=420; When frame head PN pattern is PN595, P=595; When frame head PN pattern is PN945, P=945.
When this method is applied to similar system of DTMB system or other and DTMB, in the steps A training sequence is meant as secondary processing sequence: choose local corresponding PN sequence of being preserved as secondary processing sequence s according to frame head mode that estimates and PN sequence phase
Pn(m).
Wherein, in the steps A to sequence r
Pn(n) and sequence s
Pn(m) do cross-correlation operation, obtain correlated series s
Corr(l), the sample point s of correlated series
Corr(l) obtain according to following formula:
Wherein conj () expression complex conjugate operation, the value of l is to 2 * L from 0
Ds, maxEach integer, comprise 0 and 2 * L
Ds, max, that is: l=0,1 ..., 2 * L
Ds, max
Wherein, step B is as shown in Figure 5, specifically can comprise:
B1, the power factor threshold T is set
Power, this threshold value obtains according to emulation;
B2, ask correlated series s
Corr(l) power sequence P
Corr(l)=| s
Corr(l) |
2, l=0,1 ..., 2L
Ds, max
B3, ask power sequence P
Corr(l) average power P
Corr, avg
B4, the power sequence of sample-by-sample ground judgement from front to back P
CorrWhether each sample (l) is greater than T
Power* P
Corr, avgWherein, greater than T
Power* P
Corr, avgThe index value l of first sample be I
First, greater than T
Power* P
Corr, avgThe index value l of last sample be I
Last
When practical application, the thresholding judgement is not limited to the judgement to power, also can be the judgement to range value, such as the amplitude factor threshold T can be set
Amp, wherein can get
Other processing copies B2 to carry out to B4.
When this method is applied to similar system of DTMB system or other and DTMB, in the step B1, T
PowerCan but be not limited to 16.
Wherein, in the step B3, specifically ask average power P
Corr, avgMethod can but to be not limited to be P
Corr, avg=mean (P
Corr(l)), wherein mean () representes to average operation.
Wherein, step B4 can specifically comprise:
(1) S set of structure, the element of this set is the relative indexing value of every path position in the multipath channel; The method that obtains the element of this set is: power sequence P is judged on sample-by-sample ground from front to back
CorrWhether each sample (l) is greater than T
Power* P
Corr, avgIf greater than, then the index value l of this sample is the element of S set just; Otherwise, be not the element of S set;
(2) if S set for null set, then first element value of S set is the location index value in last footpath, is designated as I
FirstLast element value of S set is the location index value in last footpath, is designated as I
Last
This method can also comprise: if thresholding is judged failure (is null set such as S set), represent that then the delay spreading parameter estimation is unsuccessful, adjustment T
PowerAnd/or L
Ds, max(such as turning T down
Power, transfer big L
Ds, max), if only adjusted T
Power, then return above-mentioned steps B, if adjusted L
Ds, max, return above-mentioned steps A after then generating new pending sequence, promptly carry out the estimation of delay spreading parameter once more, until success.
The receiver delay spreading parameter estimation unit that the present invention proposes, as shown in Figure 6, comprising: correlated series generation module, thresholding judge module and delay spreading parameter computing module; Wherein:
Said correlated series generation module is used to receive the reception signal through the carrier frequency offset compensation, and itself and training sequence are carried out cross-correlation operation, obtains correlated series and sends to the thresholding judge module;
Said thresholding judge module is used for judging through thresholding, finds out in the said correlated series the location index value I in last footpath
FirstLocation index value I with last footpath
Last, and send to the delay spreading parameter computing module;
The delay spreading parameter computing module is used to receive I
FirstAnd I
Last, and use I
LastDeduct I
FirstObtain delay spreading parameter L
Ds
Wherein, said correlated series generation module is as shown in Figure 7, comprises that pending sequence generates submodule, secondary processing sequence generates submodule and cross-correlation operation submodule;
Said pending sequence generates submodule and is used to receive the reception signal through the carrier frequency offset compensation, and therefrom intercepting one piece of data sequence as pending data sequence r
Pn(n), send to the cross-correlation operation submodule;
Said secondary processing sequence generates submodule and is used to provide training sequence as secondary processing sequence s
PnAnd send to the cross-correlation operation submodule (m);
Said cross-correlation operation submodule is used for pending data sequence r
Pn(n) and secondary processing sequence s
Pn(m) carry out cross-correlation operation, obtain correlated series s
Corr(l).
Wherein, said secondary processing sequence generate training sequence that submodule provided can but be not limited to frame head PN sequence, also can be other sequence, as complete 0 or complete 1 sequence etc.
Wherein, said pending sequence generates the pending data sequence r that submodule generates
Pn(n) length is 2 * L
Ds, max+ P, wherein n is the index of sample of signal, the value of n is from I
Sync-L
Ds, maxTo I
Sync+ L
Ds, maxEach integer of+P-1 comprises I
Sync-L
Ds, maxAnd I
Sync+ L
Ds, max+ P-1; That is:
n=I
sync-L
ds,max,I
sync-L
ds,max+1,……,I
sync+L
ds,max+P-1。
Wherein, said secondary processing sequence generates the secondary processing sequence s that submodule sends
Pn(m) length is P, and wherein the value of index m is each integer from 0 to P-1, comprises 0 and P-1, that is: m=0, and 1 ..., P-1.
When this estimation unit is applied to similar system of DTMB system or other and DTMB, when frame head PN pattern is PN420, P=420; When frame head PN pattern is PN595, P=595; When frame head PN pattern is PN945, P=945.
When this estimation unit was applied to similar system of DTMB system or other and DTMB, said secondary processing sequence generates submodule provided training sequence to be meant as secondary processing sequence: secondary processing sequence generates submodule and chooses the corresponding PN sequence of being preserved this locality as secondary processing sequence s according to frame head mode that estimates and PN sequence phase
Pn(m).
Said cross-correlation operation submodule carries out cross-correlation operation to two sequences according to following formula, obtains correlated series s
Corr(l):
Wherein conj () expression complex conjugate operation, the value of l is to 2 * L from 0
Ds, maxEach integer, comprise 0 and 2 * L
Ds, max, that is: l=0,1 ..., 2 * L
Ds, max
Wherein, said thresholding judge module is as shown in Figure 8, can comprise that threshold value is provided with submodule, power sequence generates submodule, average power generation submodule and index value and judges submodule;
Said threshold value sub module stored is used to store set power factor threshold T
Power, this threshold value obtains according to emulation;
Said power sequence generates submodule and is used to ask correlated series s
Corr(l) power sequence P
Corr(l)=| s
Corr(l) |
2, l=0,1 ..., 2L
Ds, max
Said average power generates submodule and is used to ask power sequence P
Corr(l) average power P
Corr, avg
Said index value judges that submodule is used for the power sequence of sample-by-sample ground judgement from front to back P
Corr(l) each sample in carries out P
Corr(l) whether greater than T
Power* P
Corr, avg, and will be greater than T
Power* P
Corr, avgThe index value l of first sample be designated as I
First, will be greater than T
Power* P
Corr, avgThe index value l of last sample be designated as I
Last
When practical application, the thresholding judge module is not limited to the judgement to power, also can be the judgement to range value, such as threshold value submodule is set and can be used to be provided with the amplitude factor threshold T
Amp, wherein can get
Other each submodule is copied above-mentioned submodule design.
When this method was applied to similar system of DTMB system or other and DTMB, said threshold value was provided with the set power factor threshold T of submodule
PowerCan but be not limited to 16.
Wherein, said average power generate submodule can but be not limited to ask average power P by following formula
Corr, avg:
P
Corr, avg=mean (P
Corr(l)), wherein mean () representes to average operation.
Wherein, said index value is judged submodule sample-by-sample ground judgement from front to back power sequence P
CorrWhether each sample (l) is greater than T
Power* P
Corr, avg, and will be greater than T
Power* P
Corr, avgThe index value l of first sample be designated as I
First, will be greater than T
Power* P
Corr, avgThe index value l of last sample be designated as I
LastSpecifically can be meant:
Index value is judged S set of submodule structure, and the element of this set is the relative indexing value of every path position in the multipath channel; Index value is judged submodule sample-by-sample ground judgement from front to back power sequence P
CorrWhether each sample (l) is greater than T
Power* P
Corr, avgIf greater than, this index value l is designated as the element of S set; Otherwise, be not designated as the element of S set; If S set is not null set, then index value judges that submodule is designated as I with first element value of S set
First, this is the location index value in last footpath, and last element value is designated as I
Last, this is the location index value in last footpath.
Said thresholding judge module can also be used for: if thresholding is judged failure, then T is adjusted in indication
PowerAnd/or L
Ds, maxIf only adjusted T
Power, then the thresholding judge module rejudges; If adjusted L
Ds, max, then indicate pending sequence to generate submodule and generate new pending sequence; Accordingly, generate new pending sequence after, cross-correlation operation submodule and thresholding judge module carry out handled according to new pending sequence successively, if also adjusted simultaneously T
Power, the new T of foundation when then the thresholding judge module is judged
Power
Certainly; The present invention also can have other various embodiments; Under the situation that does not deviate from spirit of the present invention and essence thereof; Those of ordinary skill in the art work as can make various corresponding changes and distortion according to the present invention, but these corresponding changes and distortion all should belong to the protection range of claim of the present invention.