Background technology
In Wideband Code Division Multiple Access (WCDMA) (WCDMA) system, because the multipath characteristics that signal is propagated, usually adopt the Rake receiver that each the footpath signal that searches is carried out demodulation, despreading, and merge to come the compensate for channel decline, thereby obtain high as far as possible multipath gain by multipath.
Fig. 1 shows the structured flowchart of Rake receiver in the WCDMA system.As shown in Figure 1, the Rake receiver comprises Multipath searching module 1, channel estimation module 2, demodulation module 3, multipath merge cells 4 and decoding unit 5.Wherein, Multipath searching module 1 is searched for each paths signal that may exist and is offered demodulation module 3 and channel estimation module 2 according to the signal that predetermined judgement principle filters out some paths from multipath signal.Channel estimation module 2 can estimate this locational channel parameter of each data symbol of footpath signal and offer demodulation module 3 according to the pilot symbol information of every footpath signal.The multipath signal that demodulation module 3 utilizes this channel estimation value that the Multipath searching module is screened carries out demodulation and obtains their estimated value.The multipath signal that multipath merge cells 4 comes out demodulation module 4 demodulation offers decoding unit 5 according to certain rule merging and deciphers processing.
Below further describe the process that estimates the locational channel parameter of this each data symbol of footpath signal according to pilot symbol information.In the WCDMA system, up link adopts parallel as shown in Figure 2 pilot channel structure, Dedicated Physical Data Channel (Dedicated Physical Data Channel wherein, abbreviation DPDCH) is used to send data, and Dedicated Physical Control Channel (Dedicated Physical ControlChannel is called for short DPCCH) is used for pilot symbol transmitted.Fig. 2 is the frame structure of up link DPDCH channel and DPCCH channel, as shown in Figure 2, comprises the data symbol of 10 bits, the N that wherein begins at any one time slot of DPCCH channel one frame
PilotIndividual bit is a frequency pilot sign, and table 1 and 2 shows number of pilot symbols N among the figure
PilotBe 3,4,5,6,7 and 8 o'clock sequence of pilot symbols P
iSuppose that the pilot frequency sequence that a certain multi-path location sends is P in j time slot of a frame
i(i=0,1 ... N
Pilot), corresponding channel response is Axe
J θ i, A wherein
iBe i the frequency pilot sign amplitude fading of channel constantly, θ
iBe the channel phase rotation constantly of i frequency pilot sign, then receive pilot frequency sequence
Can pass through thus pilot frequency sequence P
iWith
Execution adds up with equation (1) expression relevant and average calculating operation calculates j time slot interior average channel estimation value or channel parameter C
j:
Table 1
| N
pilot=3
| N
pilot=4
| N
pilot=5
| N
pilot=6
|
Bit# | 0 | 1 | 2 | 0 | 1 | 2 | 3 | 0 | 1 | 2 | 3 | 4 | 0 | 1 | 2 | 3 | 4 | 5 |
Slot#0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 | 1 0 0 0 1 1 1 1 0 1 0 1 1 0 0 | 1 0 1 0 0 1 1 0 1 1 1 0 0 0 0 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 0 0 0 1 1 1 1 0 1 0 1 1 0 0 | 1 0 1 0 0 1 1 0 1 1 1 0 0 0 0 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 0 0 0 1 1 1 1 0 1 0 1 1 0 0 | 1 0 1 0 0 1 1 0 1 1 1 0 0 0 0 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 0 0 0 1 0 0 1 1 0 1 0 1 1 | 0 0 1 0 1 0 0 0 0 1 1 1 0 1 1 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 0 0 0 1 1 1 1 0 1 0 1 1 0 0 | 1 0 1 0 0 1 1 0 1 1 1 0 0 0 0 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 0 0 0 1 0 0 1 1 0 1 0 1 1 | 0 0 1 0 1 0 0 0 0 1 1 1 0 1 1 |
Table 2
| N
pilot=7
| N
pilot=8
|
Bit# | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
Slot#0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 0 0 0 1 1 1 1 0 1 0 1 1 0 0 | 1 0 1 0 0 1 1 0 1 1 1 0 0 0 0 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 0 0 0 1 0 0 1 1 0 1 0 1 1 | 0 0 1 0 1 0 0 0 0 1 1 1 0 1 1 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 0 0 0 1 1 1 1 0 1 0 1 1 0 0 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 0 1 0 0 1 1 0 1 1 1 0 0 0 0 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 0 0 0 1 0 0 1 1 0 1 0 1 1 | 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 0 0 1 0 1 0 0 0 0 1 1 1 0 1 1 |
As shown in Figure 2, comprise N in DPDCH channel and the above-mentioned DPCCH channel time slot corresponding
DataIndividual data symbol (being data bit), N
DataSize by the spreading factor SF_DPDCH (256/2 of DPDCH channel
k) decision.The average channel estimation value that demodulation module can utilize aforesaid way to obtain is carried out demodulation to all data symbols in this time slot, but the precision of channel estimation of this processing mode is not high, and precision of channel estimation is most important in Rake receives, and its accuracy directly has influence on the overall performance of receiver.For this reason, people have proposed following various channel estimation methods with to above-mentioned channel parameter C
jRevise accuracy, mainly comprise with further raising channel estimating:
The maximum-likelihood sequence estimation method, it adopts Viterbi (Viterbi) decoding algorithm to the decoding of coding received signal, revises false channel estimated value wherein, and then reaches the purpose of accurate channel estimating.Feed back, delay by data owing to existing for this technology, thereby need to change the structure of receiver, and increased amount of calculation, is difficult to realization in practice.
Wei Na (Wiener) filtering method, this is optimum in theory method, but needs the second-order statistic of known received signal, in practice owing to the randomness of signal of communication, impossible known these information, thereby be difficult to specific implementation.
Kalman (Kalman) filtering method, in a single day this is a kind of tracking prediction channel estimation method, needs the data feedback, thereby requires to change receiver structure, and mistake occurs, the estimation of mistake will last till next frequency pilot sign, thereby effect is unsatisfactory.
The Gauss interpolation method, this is a kind of non-linear interpolation algorithm, simulation result shows when travelling carriage is in high-speed case, is difficult to obtain precise channels and estimates.
The multi-slot weighted average method, this is one of method preferably at present, this algorithm adopts the method for symmetrical weighting to the pilot value of a plurality of time slots, provides the channel estimation value between frequency pilot sign.This algorithm is easier in realization, but when travelling carriage reaches certain speed, is difficult to make correct estimation.
Multi-slot linear/non-linear interpolation algorithm by the pilot tone correlation between time slot being carried out linearity or non-linear interpolation can improve channel estimated accuracy, is one of method of better performances.But when travelling carriage reached fair speed, the channel estimating performance of conventional linear interpolation algorithm had reduction, but not the computational complexity of linear interpolation algorithm is high more a lot of than linear interpolation method.
Summary of the invention
Therefore the purpose of this invention is to provide a kind of up link Rake receiver channel method of estimation, it can accurately determine the channel circumstance of average channel estimation value when adapting to the travelling carriage high-speed mobile.
According to a kind of up link Rake receiver channel method of estimation of the present invention, described Rake receiver is carried out following steps to the received signal that comprises Dedicated Physical Control Channel and Dedicated Physical Data Channel on arbitrary multi-path location:
(1) obtains the average channel estimation value of each time slot according to the sequence of pilot symbols in each time slot in the described Dedicated Physical Control Channel and local known corresponding sequence of pilot symbols;
(2) utilize the average channel estimation value of the last time slot that obtains in above-mentioned steps (1) mode and the average channel estimation value of current time slots, generate in the described Dedicated Physical Control Channel average channel estimation value of all data symbols in the current time slots by linear interpolation;
And when the spreading factor of Dedicated Physical Control Channel during greater than the spreading factor of Dedicated Physical Data Channel, described method further may further comprise the steps:
(3a) data symbol in the corresponding time slot of Dedicated Physical Data Channel is divided into plurality of data symbols group, every group of data symbol number that comprises equals the spreading factor of Dedicated Physical Control Channel and the ratio of the spreading factor of Dedicated Physical Data Channel, described every group corresponding with adjacent two data symbols in the Dedicated Physical Control Channel corresponding time slot, and previous in the back conduct in two data symbols corresponding and two the corresponding data symbols of next group with current group; And
(3b) utilize the average channel estimation value of two adjacent in the Dedicated Physical Control Channel corresponding time slot corresponding data symbols, generate the average channel estimation value of interior all data symbols of this group by linear interpolation with every group.
Therefore method of the present invention adopts the secondary linear interpolation processing to provide accurate average channel estimation value for each data symbol in the time slot in the Dedicated Physical Data Channel, thereby precision of channel estimation is provided.Particularly, when travelling carriage high-speed motion or channel fading rate were higher, method of the present invention can greatly improve the accuracy of channel estimating.
Embodiment
Below by accompanying drawing preferred embodiment of the present invention is described.
Fig. 3 is the flow chart according to channel estimation process method of the present invention.As shown in Figure 3, received signal is imported correlator after sampling, in correlator, received signal is at first made related operation and is carried out summation operation with scrambler and orthogonal variable length spreading code, owing to comprised the user profile in the spread spectrum communication system in scrambler and the orthogonal variable length spreading code, therefore can utilize above-mentioned relevant accumulating operation from the sampling received signal, to extract the received signal relevant with the specific user, comprise the DPCCH data that only are subjected to channel disturbance in this signal, the data accumulation length of sum unit is 256 herein, equals the spreading code length of DPCCH channel.Subsequently, utilize pilot frequency sequence that aforementioned equation (1) comprises pilot field in each time slot in the DPCCH channel on each multi-path location and the local corresponding pilot frequency sequence that produces implement relevant add up with average calculating operation to obtain the average channel estimation value of each time slot.
As shown in Figure 3, be sent to through the received signal one tunnel of related operation and accumulating operation and ask the mould squaring cell to ask computing module-square obtaining the being used for correlation energy value that multipath is chosen, and another road received signal be sent to channel estimation module finish according to of the present invention based on the secondary linear interpolation channel estimating and offer demodulation module 3 among Fig. 1 to carry out demodulation.
Below be that example is described the channel estimation methods based on the secondary linear interpolation of the present invention with the channel estimation process of data bit in j time slot of a certain frame of special-purpose physical control channel and Dedicated Physical Data Channel among Fig. 2.
According to method of the present invention, at first utilize aforesaid equation (1), according to (j-1) and j preceding N of time slot of above-mentioned a certain frame in the Dedicated Physical Control Channel
PilotCorresponding known pilot symbols shown in individual bit and the table 1 and 2 calculates the average channel estimation value C of this time slot in the Dedicated Physical Control Channel
J-1And C
j
Then utilize the average channel estimation value C of (j-1) individual (being last time slot) and j (being current time slots) of above-mentioned steps acquisition
J-1And C
jGenerate the average channel estimation value C of interior all 10 the data symbols of j time slot of this frame of Dedicated Physical Control Channel
J, n(n=1,2 ... 10), adopt the linear interpolation mode for this reason, that is, make the average channel estimation value C of these 10 data symbols
J, nAccording to the sequencing in j time slot, equally spaced be distributed in from average channel estimation value C
J-1To C
jBetween.It is worthy of note that the average channel estimation value is a plural number, so the average channel estimation value C of the data symbol of linear interpolation
J, nIn complex plane, equally spaced be distributed in C
J-1With C
jBetween line segment on, the closer to the estimated value of the data symbol of time slot starting position just the closer to C
J-1In one embodiment of the invention, (2) expression that establishes an equation under available between the average channel estimation value of each data symbol and last time slot and the current time slots average channel estimation value:
C
j,n=C
j-1+(n-1)×(C
j-C
j-1)/N (2)
Here n=1,2 ..., N, N are the number of data bit in the time slot, value is 10 in the DPCCH channel.By equation (2) as seen, the estimated value C of j interior the 1st the data symbol of time slot
J, 1The average channel estimation value of the individual time slot of value (j-1) promptly is positioned on the end points of line segment.But the linear interpolation mode shown in the equation (2) is not unique, for example can utilize down the mode shown in (3) that establishes an equation to carry out linear interpolation yet, at this moment the estimated value C of last data symbol
J, 10Value is the average channel estimation value of j time slot:
C
j,n=C
j-1+n×(C
j-C
j-1)/N (3)
Here n=1,2 ..., N, N are the number of data bit in the time slot, value is 10 in the DPCCH channel.In addition, the average channel estimation value C of data symbol
J, nEven must necessarily not leave no choice but be positioned on the end points of line segment, promptly can be not with C
J-1Or C
jBe value, as long as the estimated value C of these data symbols
J, nIn complex plane, be distributed in from C according to the sequencing of these data symbols in time slot
J-1To C
jBetween line segment on and be equally spaced getting final product each other, so alleged linear interpolation or the linear interpolation of the present invention should be understood with this.For other time slot in the Dedicated Physical Control Channel, the average channel estimation value of its each symbol also obtains in above-mentioned this mode, and the rest may be inferred.
According to the indication of the transformat combination in the DPCCH channel time slot (Transport FormatCombination Indicator, be called for short TFCI) can obtain the spreading factor SF_DPDCH of DPDCH channel, but its value is 256,128,64,32,16,8 and 4 not wait.For the DPCCH channel, its spreading factor SF_DPCCH is always 256.As shown in Figure 2, if the spreading factor of DPDCH channel is 256 (being k=0), interior identical of the bit number in each time slot of DPDCH channel and DPCCH channel then, all be 10, this moment, DPDCH channel time slot was corresponding one by one with each data symbol of DPDDH channel time slot, so the average channel estimation value that obtains with above-mentioned interpolation method that can utilize each data symbol of DPCCH channel time slot is carried out demodulation to the corresponding data symbol of DPDDH channel time slot.
If the spreading factor SF_DPDCH of DPDCH channel is less than the spreading factor SF_DPCCH of DPCCH channel, then according to method of the present invention, in obtaining special physical control channel time slot, after the average channel estimation value of each data symbol, adopt the linear interpolation mode further to generate all N in j corresponding in the Dedicated Physical Data Channel time slot
DataThe average channel estimation value C of individual data symbol
J, m(m=1,2 ..., N
Data).Particularly, at first with this N
DataIndividual data are divided into 10 groups of data, and every group comprises (SF_DPCCH/SF_DPDCH) individual data symbol, and this every group of data symbol all the adjacent two data symbol with Dedicated Physical Control Channel is corresponding.Particularly, the 1st group of data (i.e. the 1st~(SF_DPCCH/SF_DPDCH) individual data symbol) are corresponding with interior the 10th data of (j-1) individual time slot and j interior the 1st data of time slot, the 2nd group of data (i.e. the individual data symbol in ((SF_DPCCH/SF_DPDCH)+1)~2 * (SF_DPCCH/SF_DPDCH)) are corresponding with interior the 1st data of j time slot and j interior the 2nd data of time slot, the rest may be inferred, and last group data (i.e. the individual data symbol in (9 * (SF_DPCCH/SF_DPDCH)+1)~10 * (SF_DPCCH/SF_DPDCH)) are corresponding with interior the 9th data of j time slot and j interior the 10th data of time slot.Then according to above-mentioned linear interpolation mode, to the estimated value C of the data symbol in every group
J, mCarry out linear interpolation.In the linear interpolation here, the estimated value C of every group of interior (SF_DPCCH/SF_DPDCH) individual data symbol
J, mAccording to the sequencing in j time slot, equally spaced be distributed between the average channel estimation value of the adjacent two data symbol corresponding, for example the average channel estimation value C of the 1st~(SF_DPCCH/SF_DPDCH) individual data symbol with every group
J, mIn complex plane, equally spaced be distributed in estimated value C
J-1,10And C
J, 1Between line segment on, and the closer to the estimated value of the data symbol of time slot starting position just the closer to C
J-1,10, the rest may be inferred, the average channel estimation value C of the individual data symbol in (9 * (SF_DPCCH/SF_DPDCH)+1)~10 * (SF_DPCCH/SF_DPDCH)
J, mIn complex plane, equally spaced be distributed in estimated value C
J, 9And C
J, 10Between line segment on, and the closer to the estimated value of the data symbol of time slot starting position just the closer to C
J, 9
The same with the linear interpolation mode of each data symbol in the aforementioned Dedicated Physical Control Channel, average channel estimation value C
J, mAnd also available equation (2) and (3) expression between the average channel estimation value of two adjacent data symbols, and even need not be value with the average channel estimation value of two adjacent data symbols, as long as the estimated value C of these data symbols
J, mIn complex plane, be distributed on the line segment between the average channel estimation value of two adjacent data symbols of corresponding Dedicated Physical Control Channel and be equally spaced getting final product each other according to sequencing at time slot.
In order to verify the implementation result of channel estimation methods of the present invention, by experiment, with method of the present invention with contrast on systematic function based on four time slot weighted averages with based on the once linear interpolation algorithm.These comparative experimentss are carried out under the 120km/h mobile environment, have measured error rate BER and the signal to noise ratio E of 64kbps, 144kbps and 384kbps business under the whole bag of tricks respectively
b/ N
0Between relation, the results are shown in Fig. 4 (a)~(c), wherein, the corresponding 384kbps business of Fig. 4 (a), the corresponding 144kbps business of Fig. 4 (b), and the corresponding 64kbps business of Fig. 4 (c), and method 1 representative is based on the average weighted channel estimation method of 4 time slots, method 2 representatives are based on the channel estimation method of once linear interpolation, and the channel estimation method based on the secondary linear interpolation that on behalf of the present invention, method 3 propose.
The experimental result of Fig. 4 (a)~(c) shows, does at travelling carriage under the environment of high-speed motion, no matter be business under which kind of speed, quadratic interpolation algorithm of the present invention is at identical signal to noise ratio E
b/ N
0All have minimum error rate BER down, so performance is better than other two kinds of algorithms.Particularly, the present invention will obviously be better than traditional method of only carrying out an interpolation processing by increasing the performance after the interpolation.
Can under the existing hardware condition, realize according to the channel estimation methods based on the secondary linear interpolation of the present invention and can not influence processing speed, therefore improved signal receiving performance, guaranteed that the Rake receiver still has desirable bit error rate performance when the travelling carriage high-speed motion.