CN1332544A - Adaptive predicting smoothing RAKE reception method and receiver - Google Patents

Abstract

The present invention can eliminate the multipath effect, channel interference and Doppler effect in mobile communication. The reception method includes the direct treatment of the received RF baseband signal with filter, multiplier and preset data and the direct feeding of the treated signal to the software adjudication decoder. The RAKE receiver of the present invention consists of matching filter, delay search tracker, predicting filter, smoothing filter, delay regulator, multiplier, multipath adder, filter parameter data base, S/N measuring comparator and one-out-of-three selector.

Description

Adaptive predicting smoothing RAKE reception method and receiver

(1) technical field:

The invention belongs to code division multiple access (CDMA) moving communicating field.Relate in particular to adaptive predicting smoothing RAKE reception method and receiver

(2) background technology:

Code division multiple access (CDMA) spread spectrum communication mode is to use many a kind of multi-access modes in the present communication system.Excellent specific properties such as it has anti-interference, and anti-multipath decline and power system capacity are big.In CDMA mobile communication system, the space electromagnetic wave exists direct wave and reflected wave in the air simultaneously, thereby received signal is the multipath stack that transmits after propagating, the multipath effect that Here it is receives.Simultaneously, the random noise of space channel all has random disturbances to the electromagnetic amplitude and the phase place of every paths, and this interference is called channel disturbance.Relatively moving of portable terminal and base station can produce Doppler frequency shift, and the random noise of its meeting and channel is disturbed the electromagnetic phase place of transmitting, the Doppler effect in the mobile communication that Here it is together in channel.For offseting three kinds of top interference, i.e. the multipath effect of Jie Shouing, channel disturbance, the Doppler effect in the mobile communication needs to use RAKE receiver.

Traditional RAKE receiver is that received signal is carried out despreading respectively by P matched filter, and it is leading then different time delays to be done in every footpath, more directly addition.This RAKE receiver, the phase perturbation that requires channel disturbance to bring are limited in the very little scope, and be little to the tolerance of channel disturbance and Doppler frequency shift, in case do not satisfy above-mentioned restriction, the performance rapid deterioration.

The object of the present invention is to provide a kind of multipath effect that can offset above-mentioned reception, channel disturbance, a kind of novel RAKE receiver model of the Doppler effect in the mobile communication, i.e. adaptive predicting smoothing RAKE receiver.

(3) summary of the invention:

Method of the present invention is to use filter, and multiplier etc. in conjunction with the database that can preset, are directly handled the baseband signal behind the receiving terminal radio frequency down-conversion, handles the back output signal and can directly deliver to the soft decoder of declaring.

Concrete grammar of the present invention is to use the bank of filters of three kinds of different lengths, separates through the multipath matched filtering, and multipath merges, and exports three road signals, and this three road signal selects a road of maximum signal to noise ratio to export to the soft decoder of declaring through snr measurement.Simultaneously, the result who selects is outputed to the memory-type database simultaneously, give three groups of new filter parameters of bank of filters configuration, reach the purpose of adaptive adjustment filter tracks channel variation by database.

The present invention has been suitable for using the CDMA mobile communication system of continuous or discontinuous pilot tone, the time-division pilot tone CDMA mobile communication system that is suitable for the shared code channel of data and pilot tone, or data and pilot tone are used the parallel pilot tone CDMA mobile communication system of different code channels.

The adaptive predicting smoothing RAKE receiver that designs based on the inventive method, comprise matched filter, delay time search tracker, predictive filter, smoothing filter, the time delay adjuster, multiplier, the multipath adder, filter parameter database, snr measurement and comparator, three select one gate etc.The input of the matched filter of pilot tone and data is the baseband signals behind the receiving terminal radio frequency down-conversion, the matched filter output of pilot tone links to each other with the input of predictive filter and smoothing filter, the input of linking multiplier after the conjugation is got in the output of predictive filter and smoothing filter, another of this multiplier is input as the output of the matched filter of data, the multipath adder is received in the output of multiplier, the output of multipath adder is received three and is selected one gate input, also input links to each other with comparator with snr measurement in the output of multipath adder, the output of snr measurement and comparator and three selects the selected control end of one gate to link to each other, and the output of snr measurement and comparator also links to each other with the input of filter parameter database.Wherein, snr measurement and comparator comprise measuring appliance and comparator two parts, and the filter parameter database comprises prefabricated parameter controller, related list item and parameter refresh controller three parts.

The adaptive predicting smoothing RAKE receiver that the present invention proposes, data are handled in real time, system delay is reduced greatly, ability of anti-multipath is strong simultaneously, quick variation adaptability to channel magnitude and phase factor is strong, can the adaptive tracing adjustment to the translational speed of portable terminal, the demodulation signal to noise ratio and the processing speed of the signal that the terminal speed change in the variable-parameter channel moves are improved greatly, improved systematic function.

Beneficial effect of the present invention:

Can offset multipath effect, channel disturbance, the Doppler effect in the mobile communication.

(4) description of drawings:

Fig. 1 is an adaptive predicting smoothing RAKE receiver schematic diagram.

Fig. 2 is that the RAKE of median filter group merges the concrete structure schematic diagram.

Fig. 3 is the schematic diagram of the K footpath pilot filter concrete structure in the median filter group.

Fig. 4 is the specific implementation figure of predictive filter.

Fig. 5 is the specific implementation figure of smoothing filter.

Fig. 6 is an adaptive predicting smoothing RAKE operation of receiver flow chart.

Fig. 7 is the workflow diagram of median filter group.

(5) specific embodiment:

Realize the optimal way of invention below in conjunction with description of drawings:

Referring to Fig. 1: wherein, 1 is short bank of filters, 2 is the median filter group, 3 are long bank of filters, the difference of short bank of filters 1, median filter group 2, long bank of filters 3 is, the length of two kinds of crucial wave filters is max{M-Δ M in the short bank of filters 1,0 } and max{N-Δ N, 0}, the length of two kinds of crucial wave filters is M and N in the median filter group 2, two kinds of crucial filter lengths are min{M+ Δ M in the long bank of filters 3, M_maxAnd min{N+ Δ N, N_max, 4 is the one-out-three gate, 5 is snr measurement And comparator, 6 is the filter parameter database.

Referring to Fig. 2: wherein, 7 is that (matched filter of footpath data of 1＜=K＜=P), 8 is that K directly leads to K The delay time search tracker in matched filter frequently and K footpath, 9 is K footpath pilot filters, 10 is the K footpath Initial delay regulator.

Referring to Fig. 3: wherein, the 11st, the single output of dual input time-division switching, the 12nd, predictive filter, the 13rd, Smoothing filter. The lower part of Fig. 3 is a kind of concrete reality of predictive filter 12 and smoothing filter 13 Existing structure.

Referring to Fig. 4: wherein D is unit delayer.

Referring to Fig. 5: wherein D is unit delayer.

Be described further below in conjunction with Fig. 1

Wherein baseband receiving signals is done the filtering processing of parallel upper, middle and lower three tunnel, its short-and-medium bank of filters 1, The basic structure of median filter group 2, long bank of filters 3 is identical, but crucial filter length ginseng in them Several M are different with N, therefore lack bank of filters 1, median filter group 2, long bank of filters 3 according to this M and N, The filter tap coefficients that obtains in the related list item from filter parameter database 6 is not identical yet. Three choosings The gating control end of one gate 4 is by the output control of snr measurement and comparator 5. Snr measurement and The snr measurement algorithm that comparator 5 uses will preset before operation of receiver, and the signal to noise ratio comparator is used for three Select a maximum. Filter parameter database 6 is mainly by preset controller, related list item and refresh controller Form, preset control section branch content wherein is predicting strategy and level and smooth strategy, and determines according to these strategies Fixed a series of control parameters, the prediction that provides take Fig. 3 and level and smooth specific implementation structure are as example, i.e. two kinds of filters The ripple device all realizes with FIR, and then at this moment control parameter is:

Predicting strategy---interpolation FIR

Level and smooth strategy---level and smooth FIR

The maximum length of two kinds of wave filters---M_MAXAnd N_MAX

The adjustment step-length of two kinds of wave filters---0＜K≤16 are for getting fixed constant, f_MAXBe the maximum frequency deviation of system tolerant, T_CHIPBe the chip time width of CDMA GSM, N is to f_MAXQuantification count the refresh cycle---T_FILTERValue is not less than the contingency table item parts in the symbol speed filter parameter database 6 of pilot tone, the prediction that still provides take Fig. 4, Fig. 5 and level and smooth specific implementation are as example, then related list item is as index with filter length, be implemented to the association of filter tap parameter, the refresh control part in the filter parameter database 6 is with cycle T_FILTERTo short bank of filters 1, Median filter group 2, the filter parameter in the long bank of filters 3 refreshes.

Referring to Fig. 2, Fig. 3, Fig. 4, Fig. 5 (simultaneously workflow described in conjunction with Figure 7)

In CDMA mobile communication system: (1) establishes T_CHIPThe chip time width, T_DThe time width of data symbol, T_PPilot tone symbol

Time width, common T_PT_DMultiple, get T here_P＝T _D (2) transmit and contain pilot tone, be expressed as S (t)=C_P(t)＋C _D(t)(I _D(t)＋jQ _D(t)), wherein

C _P(t)，C _D(t) be respectively the spreading code of pilot tone and data, I_D(t)＋jQ _DThe data of (t) modulating for QPSK,

Unmodulated data on the pilot tone. Need to prove that pilot tone can be continuous or discontinuous, also can be to lead the time-division

Frequency or parallel pilot tone when pilot tone does not exist, think that the pilot tone item is 0. Pilot tone exists and does not exist of living in

Time slot, should be preset in the system timer, be used for the worker of the matched filter of all pilot tones of control

Do. (3) baseband receiving signals is

In Fig. 2, K (in the processing module group in the footpath of 1＜=K＜=P), K (1＜=K＜=P) the K footpath data of matched filter 7 outputs of footpath data are: $D_K\left(n\right)=\frac{1}{T_D}\underset{t_K+n{T}_D}{\overset{t_K+(n+1)T_D}{\∫}}{\mathrm{\λ}}_K\left(t\right)(I_D(t-t_K)+j{Q}_D(t-t_K))\mathrm{dt}+\frac{1}{T_D}\underset{t_K+n{T}_D}{\overset{t_K+(n+1)T_D}{\∫}}{\mathrm{\λ}}_K\left(t\right)C_P(t-t_K)C_D^{*}(t-t_K)\mathrm{dt}$ $+\frac{1}{T_D}\underset{\underset{I\≠K}{I=1}}{\overset{P}{\mathrm{\Σ}}}\underset{t_K+n{T}_D}{\overset{t_K+(n+1)T_D}{\∫}}{\mathrm{\λ}}_f\left(t\right)(C_P(t-t_I)+C_D(t-t_I)(I_D(t-t_I)+j{Q}_D(t-t_I))){C_D}^{*}(t-t_K)\mathrm{dt}$ $+\frac{1}{T_D}\underset{t_K+n{T}_D}{\overset{t_K+(n+1)T_D}{\∫}}n\left(t\right){C_D}^{*}(t-t_K)\mathrm{dt}$

First is K footpath data item, and second is because C_P(t)，C _D(t) the non-orthogonal K footpath pilot tone that causes pair The interference of K footpath data, the 3rd be other footpath to the directly interference of data of K, the 4th is interchannel noise The interference that produces.

The K footpath pilot tone of delay time search tracker 8 outputs in the matched filter of K footpath pilot tone and K footpath is: $P_K\left(n\right)=\frac{1}{T_P}\underset{t_K+n{T}_P}{\overset{t_K+(n+1)T_P}{\∫}}{\mathrm{\λ}}_K\left(t\right)\mathrm{dt}+\frac{1}{T_P}\underset{t_K+n{T}_P}{\overset{t_K+(n+1)T_P}{\∫}}{\mathrm{\λ}}_K\left(t\right)C_D(t-t_K)(I_D(t-t_K)+j{Q}_D(t-t_K)){C_P}^{*}(t-t_K)\mathrm{dt}$ $+\frac{1}{T_P}\underset{\underset{I\≠K}{I=1}}{\overset{P}{\mathrm{\Σ}}}\underset{t_K+n{t}_P}{\overset{t_K+(n+1)T_P}{\∫}}{\mathrm{\λ}}_I\left(t\right)(C_P(t-t_I)+C_D(t-t_I)(I_D(t-t_I)+j{Q}_D(t-t_I))){C_P}^{*}(t-t_K)\mathrm{dt}$ $+\frac{1}{T_P}\underset{t_K+n{T}_P}{\overset{t_K+(n+1)T_P}{\∫}}n$ $\left(t\right){C_P}^{*}(t-t_K)\mathrm{dt}$ First is K footpath pilot tone, and second is because C_P(t)，C _D(t) the non-orthogonal K footpath data that cause are to the The interference of K footpath pilot tone, the 3rd be other footpath to the interference of K footpath pilot tone, the 4th is the interchannel noise product The interference of giving birth to.

The input of the single output of the dual input time-division switching 11 of K footpath pilot filter 9 connects the coupling of K footpath pilot tone The output of the delay time search tracker 8 in wave filter and K footpath starts the K footpath at the time slot that does not have pilot tone The predictive filter 12 of pilot filter 9, the continuity of coming inhibit signal 1 to export. K footpath pilot filtering The input of the smoothing filter 13 of device 9 connects the output of the single output of input time-division switching 11, uses self adaptation Wave filter come level and smooth, to overcome the impact of noise. Predictive filter 12, the parameter of smoothing filter 13 All by filter parameter database 6 periodic refreshings.

Above-mentioned adaptive predicting smoothing RAKE reception method uses other predicting strategy, and level and smooth strategy is namely repaiied Change predictive filter and smoothing filter, do not adopt the FIR structure, and adopt other wave filter (as non-linear Structure, the time become structure etc.).

Above-mentioned adaptive predicting smoothing RAKE reception method adopts the bank of filters more than three tunnel, selects these many One in the signal of road as output.

One group of noise that above-mentioned adaptive predicting smoothing RAKE reception method is measured with the snr measurement comparator Ratio, according to the adjustment step-length of this group signal to noise ratio adaptive determining wave filter, and adaptive control filter parameter The refresh cycle of database.

Above-mentioned adaptive predicting smoothing RAKE reception method is exported with multi-parameter in the snr measurement comparator Comparator replaces the comparator of single output.

Above-mentioned adaptive predicting smoothing RAKE reception method uses multi-parameter output as the filter parameter database Index.

Claims (9)

1. adaptive predicting smoothing RAKE reception method, it is characterized by according to following workflow and carry out: the 1. bank of filters that enters three kinds of different lengths of the baseband signal behind the receiving terminal radio frequency down-conversion, separate through the multipath matched filtering, multipath merges, and exports three road filtering signals.2. with signal to noise ratio as judgment variables, through snr measurement, select a road of maximum signal to noise ratio to export to the soft decoder of declaring.3. according to the result queries database of selecting, give three groups of new filter parameters of bank of filters configuration, reach the purpose of adaptive adjustment filter tracks channel variation by database.

2. the adaptive predicting smoothing RAKE receiver is characterized by by matched filter, delay time search tracker, predictive filter, smoothing filter, the time delay adjuster, multiplier, multipath adder, filter parameter database, snr measurement and comparator, three select one gate composition.

3. adaptive predicting smoothing RAKE receiver as claimed in claim 2, it is characterized in that: the input of the matched filter of pilot tone and data is the baseband signals behind the receiving terminal radio frequency down-conversion, the matched filter output of pilot tone links to each other with the input of predictive filter and smoothing filter, the input of linking multiplier after the conjugation is got in the output of predictive filter and smoothing filter, another of this multiplier is input as the output of the matched filter of data, the multipath adder is received in the output of multiplier, and the output of multipath adder is received three and selected one gate input.Also input links to each other with snr measurement and comparator in the output of multipath adder, and the output and three of snr measurement and comparator selects the selected control end of one gate to link to each other, and the output of snr measurement and comparator also links to each other with the input of filter parameter database.Wherein, snr measurement and comparator comprise measuring appliance and comparator two parts, and the filter parameter database comprises prefabricated parameter controller, related list item and parameter refresh controller three parts.

4. adaptive predicting smoothing RAKE reception method as claimed in claim 1 is characterized in that using other predicting strategy, level and smooth strategy, promptly revise predictive filter and smoothing filter, do not adopt the FIR structure, and adopt other filter (as nonlinear organization, time become structure or the like).

5. adaptive predicting smoothing RAKE reception method as claimed in claim 1 is characterized in that adopting the bank of filters more than three tunnel, selects in these multiple signals one as output.

6. adaptive predicting smoothing RAKE reception method as claimed in claim 1, it is characterized in that one group of signal to noise ratio that the snr measurement comparator is measured, according to the adjustment step-length of this group signal to noise ratio adaptive determining filter, and the refresh cycle of adaptive control filters parameter database.

7. adaptive predicting smoothing RAKE reception method as claimed in claim 1, it is characterized in that in the snr measurement comparator, using one group of other judgment variables, the adjustment step-length of adaptive determining filter, and the refresh cycle of adaptive control filters parameter database.

8. adaptive predicting smoothing RAKE reception method as claimed in claim 1 is characterized in that the comparator with multi-parameter output replaces single comparator of exporting in the snr measurement comparator.

9. adaptive predicting smoothing RAKE reception method as claimed in claim 1 is characterized in that using the index of multi-parameter output as the filter parameter database.

Images