CN104734768A - Method for improving convergence rate of repeater self-excitation suppression system - Google Patents

Abstract

The invention provides a method for improving the convergence rate of a repeater self-excitation suppression system. The method comprises the steps of estimating the gain isolation difference delta r of a repeater and the relative phase delta omicron of a self-excitation signal through a self-transmitting and self-receiving sequence, and selecting the initial tapping coefficient wk (0) raw and the learning stepping coefficient eta of a corresponding self-adapting filter from a pre-stored coefficient lookup table according to the estimated gain isolation difference delta r. By the adoption of the method, a reasonable initialization value of the self-excitation suppression system can be rapidly obtained, the convergence rate of the self-adapting filter can be remarkably improved, and accordingly the purpose of improving the convergence rate of the self-excitation suppression system can be achieved.

Description

Improve the method that straight-forward station self-excitation suppresses the convergence rate of system

Technical field

The present invention relates to wireless communication field, particularly digital radio repeater field.

Background technology

Because the isolation between wireless discharging-directly station dual-mode antenna is inadequate, the part signal of transmitting antenna can leak into reception antenna and form echo interference.If now the gain of repeater is greater than the isolation between dual-mode antenna, the self-excitation of system will be caused.Therefore, self-excitation suppresses system (Interference Cancellation System, ICS) to be widely used in wireless discharging-directly station, for improve wireless discharging-directly station gain, reduce the insulated degree requirement of dual-mode antenna and reduce difficulty of construction.

As shown in Figure 1, be the model schematic of digital radio repeater in prior art.The principle of self-excitation suppression system is the characteristic of channel (such as estimating the characteristic of channel with LMS algorithm) by estimating self-excitation multipath signal, then self-excitation interference signal y (n) estimated is deducted from Received signal strength R (n), to obtain signal e (n) eliminated after self-excitation interference signal for launching, i.e. e (n)=R (n)-y (n).

In self-excitation suppression system, time delay module and sef-adapting filter are used to recover self-excitation interference signal according to the characteristic of channel, and wherein the settings of time delay module are T, i.e. x (n)=e (n-T).

Wherein the tap coefficient of sef-adapting filter is w _k(n), wherein k=0,1 ..., L-1, L are the tap number of sef-adapting filter.The tap coefficient of sef-adapting filter is updated to:

W _k(n+1)=w _kn ()+η e (n) x (n), wherein η is study stepping coefficient.

The self-excitation interference signal recovered after sef-adapting filter is:

$y\left(n\right)=\underset{k=0}{\overset{L-1}{\mathrm{\Σ}}}{w}_k\left(n\right)x(n-k).$

As can be seen here, the gain of convergence rate and study stepping coefficient, signal amplitude, the repeater of sef-adapting filter, the isolation of repeater dual-mode antenna and the time delay of self-excitation signal etc. are relevant.Therefore, when self-excitation suppresses the initialization of system, if the tap coefficient initial value w of sef-adapting filter _k(0), study stepping coefficient η or the settings T of time delay module arranges unreasonable, the convergence rate of sef-adapting filter will be caused slow, and easily disperse.

In order to select the tap coefficient initial value w of best sef-adapting filter _k(0), study stepping coefficient η and the settings T of time delay module, attempt various settings by traversal in prior art and the w that one group of convergence rate is very fast, convergent probability is higher is selected in combination _k(0), η and T.The workload done like this is very big, efficiency is poor and the probability of success is lower.

Summary of the invention

The object of the invention is to, in order to solve the aforementioned problems in the prior, thus provide a kind of method improving straight-forward station self-excitation and suppress the convergence rate of system.Use the method promptly can obtain the initialization value of rational self-excitation suppression system, improve the convergent probability of sef-adapting filter significantly, thus reach the object improving self-excitation and suppress the convergence rate of system.

The invention provides a kind of method improving straight-forward station self-excitation and suppress the convergence rate of system, described repeater comprises receiver, transmitting chain and self-excitation suppress system, described self-excitation suppresses Operation system setting between described receiver and described transmitting chain, described self-excitation suppresses system to comprise sef-adapting filter, time delay module and subtracter, described method comprises the steps: that S10. suppresses to arrange training sequence maker and MUX between system and described transmitting chain in described self-excitation, described MUX is configured to the output signal of described training sequence maker and described self-excitation to suppress the output signal of system alternatively to output to described transmitting chain, S30. described self-excitation of stopping using suppresses system, enable described receiver, described training sequence maker, described MUX and described transmitting chain, S31. the output configuring described MUX is the training sequence T that described training sequence maker exports _s(n), S32. the second Received signal strength R (n) from receiver is received, S40. according to described training sequence T _sn () and described second Received signal strength R (n) estimate the gain isolation difference Δ r of described repeater and the relative phase Δ ω of self-excitation signal, S50. according to the gain isolation difference Δ r estimated, from the Coefficient Look-up Table prestored, the initial tap coefficient values w of corresponding described sef-adapting filter is selected _k(0) _rawwith initial study stepping coefficient η.

Further, in step s 40, use following formula to the relative phase Δ ω of the gain isolation difference Δ r and self-excitation signal that estimate described repeater:

$\mathrm{\Δr}\·e^{\mathrm{j\Δ\ω}}=\frac{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{T}_s{\left(n\right)}^{*}{R}_p\left(n\right)}{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|T_s\left(n\right)\right|}^2},$

Wherein, N is described training sequence T _sthe length of (n), T _s(n) ^*for described training sequence T _sthe conjugate sequence of (n), R _p(n) for extract from described second Received signal strength R (n) and T _sthe sequence of (n) time unifying.

Further, also comprise after step S50: S51. according to the relative phase Δ ω of self-excitation signal estimated to the initial tap coefficient values w of described sef-adapting filter _k(0) _rawcarry out phase compensation, with the initial tap coefficient values w of the described sef-adapting filter after obtaining phase compensation _k(0).

Further, also comprise after step s 32: S41. is according to the second Received signal strength R (n) and training sequence T _sn () estimates the self-excitation time delay T of self-excitation signal _d; Also comprise after step S41: S52. is according to the self-excitation time delay T estimated _d, record self-excitation time delay and the settings T selecting described time delay module the look-up table of corresponding time delay module settings from what prestore.

Further, in step S41, by described second Received signal strength R (n) and described training sequence T _sn () carries out the method for computing cross-correlation to estimate described self-excitation time delay T _d.

Further, also comprise after step S50: the tap coefficient of described sef-adapting filter is set to w by S60. _k(0), study stepping coefficient is set to η, and the settings of described time delay module are set to T; S61. enable described self-excitation suppresses system, and the tap coefficient opening described sef-adapting filter upgrades until the convergence of described sef-adapting filter.

Preferably, described repeater also comprises the first power regulation module for adjusting received signal power between described receiver and subtracter, and the second power regulation module for adjusting transmit signal power between described transmitting chain and described MUX.

Further, also comprised before step S30: S20. stop using described self-excitation suppress system, described training sequence maker, described MUX, described transmitting chain, described first power regulation module and described second power regulation module, enable described receiver; S21. the first Received signal strength R from receiver is received ₀(n); S22. described first Received signal strength R is obtained by statistics ₀the power P of (n) _l; S23. the first power adjusting factor arranging described first power regulation module is P ₀/ P _l, the second power adjusting factor arranging described second power regulation module is P _l/ P ₀, wherein P ₀for predetermined power value.

Further, also comprise after step S61: the output that S70. configures described MUX is the output that described self-excitation suppresses system, and stops the tap coefficient of described sef-adapting filter to upgrade; S71. enable described first power regulation module and described second power regulation module; S72. the tap coefficient opening described sef-adapting filter upgrades, and described repeater starts normal work.

Further, also comprised before step S10: S01. sets up the off-line simulation model of described repeater, the power setting the e (n) that transmits in described off-line simulation model is P ₀, the relative phase of self-excitation signal is 0; S02. by the many experiments of described off-line simulation model, obtain initial tap coefficient values and the study stepping coefficient of described repeater preferred described sef-adapting filter under the poor grade of various gain isolation, and the initial tap coefficient values of each gain isolation difference and each correspondence is entered described Coefficient Look-up Table with study stepping coefficient storage.

Raising straight-forward station self-excitation provided by the invention is used to suppress the method for the convergence rate of system promptly can obtain the initialization value of rational self-excitation suppression system, improve the convergent probability of sef-adapting filter significantly, thus reach the object improving self-excitation and suppress the convergence rate of system.

Accompanying drawing explanation

Fig. 1 is the model schematic of digital radio repeater in prior art;

Fig. 2 is the model schematic that raising straight-forward station self-excitation of the present invention suppresses the preferred embodiment middle digital radio repeater of the method for the convergence rate of system;

The working model schematic diagram of digital radio repeater when Fig. 3 is preferred embodiment middle step S20, S21, S22 and S23 of the method for the convergence rate performing raising straight-forward station self-excitation of the present invention suppression system;

The working model schematic diagram of digital radio repeater when Fig. 4 is preferred embodiment middle step S30, S31 and S32 of the method for the convergence rate performing raising straight-forward station self-excitation of the present invention suppression system;

The working model schematic diagram of digital radio repeater when Fig. 5 is the preferred embodiment middle step S60 and S61 of the method for the convergence rate performing raising straight-forward station self-excitation of the present invention suppression system;

The working model schematic diagram of digital radio repeater when Fig. 6 is preferred embodiment middle step S70, S71 and S72 of the method for the convergence rate performing raising straight-forward station self-excitation of the present invention suppression system.

Embodiment

Below in conjunction with accompanying drawing with preferred embodiment suppress the method for the convergence rate of system to be described in further detail to raising straight-forward station self-excitation of the present invention, but not as a limitation of the invention.

With reference to Fig. 2, suppress the model schematic of the preferred embodiment middle digital radio repeater of the method for the convergence rate of system for raising straight-forward station self-excitation of the present invention.This preferred embodiment in, digital radio repeater comprises reception antenna, receiver successively from receiving terminal to transmitting terminal, for adjusting the first power regulation module of received signal power, self-excitation suppresses system, training sequence maker, MUX, for adjusting the second power regulation module of transmit signal power, transmitting chain and transmitting antenna.Wherein, self-excitation suppresses system to comprise time delay module, sef-adapting filter and subtracter.First input signal of MUX is the output signal of training sequence maker, and its second input signal is the output signal that self-excitation suppresses system, and its output signal is the input signal of transmitting chain.MUX is configured to its first input signal and the second input signal alternatively to output to transmitting chain.These modules can control separately enable or inactive, to realize different functions.

This preferred embodiment in, first perform step S01 and S02: the off-line simulation model setting up repeater, the power setting the e (n) that transmits in this model is P ₀, the relative phase of self-excitation signal is 0.By Multi simulation running and the experiment of this model, obtain initial tap coefficient values and the study stepping coefficient of repeater preferred sef-adapting filter under various gain isolation difference grade, and by the initial tap coefficient values of each gain isolation difference and each correspondence with learn stepping coefficient storage and enter Coefficient Look-up Table.Simultaneously by Multi simulation running and the experiment of this model, be obtained from the relation between the self-excitation time delay of energizing signal and the settings of time delay module, be also stored into Coefficient Look-up Table.

The above gain isolation difference refers to the difference between the gain of wireless discharging-directly station and isolation between transmitting and receiving antenna.The scope of the gain isolation difference of usual digital radio repeater, at 25 ~-5dB, by off-line simulation and experiment, can obtain corresponding relation as shown in table 1:

The lower preferred initial tap coefficient values of table 1 various gain isolation difference and study stepping coefficient

Self-excitation suppresses system to be recover self-excitation interference signal by the channel of time delay module and sef-adapting filter simulation self-excitation interference signal, because sef-adapting filter also can bring the time delay of signal, therefore the settings of time delay module are not identical with the self-excitation time delay of self-excitation signal itself.By off-line simulation and experiment, the relation between self-excitation time delay and time delay module settings also can be obtained.

Following execution step S20, S21, S22 and S23: as shown in Figure 3, self-excitation of stopping using suppresses system, training sequence maker, MUX, transmitting chain, the first power regulation module and the second power regulation module, only enable receiver, the signal that now receiver receives is the first Received signal strength R ₀n (), obtains this first Received signal strength R by statistics ₀the power P of (n) _l, the first power adjusting factor arranging the first power regulation module is P ₀/ P _l, the second power adjusting factor arranging the second power regulation module is P _l/ P ₀, wherein P ₀for predetermined power value.

Following execution step S30, S31 and S32: as shown in Figure 4, self-excitation of stopping using suppresses system, the first power regulation module and the second power regulation module, enable receiver, training sequence maker, MUX and transmitting chain, the output of configuration MUX is the first input signal, i.e. training sequence T _sn (), the signal that now receiver receives is the second Received signal strength R (n).

Following execution step S40 and S41: according to training sequence T _sn () and the second Received signal strength R (n) estimate the gain isolation difference Δ r of repeater, the relative phase Δ ω of self-excitation signal and self-excitation time delay T _d.

Particularly, the gain isolation difference Δ r of repeater and the relative phase Δ ω of self-excitation signal is estimated by following formula:

$\mathrm{\Δr}\·e^{\mathrm{j\Δ\ω}}=\frac{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{T}_s{\left(n\right)}^{*}{R}_p\left(n\right)}{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|T_s\left(n\right)\right|}^2},$

Wherein, N is training sequence T _sthe length of (n), T _s(n) ^*for training sequence T _sthe conjugate sequence of (n), R _p(n) for extract from the second Received signal strength R (n) and T _sthe sequence of (n) time unifying.

The self-excitation time delay T of self-excitation signal _dby by the second Received signal strength R (n) and training sequence T _sn () is carried out computing cross-correlation, is estimated according to the difference of the position of relevant peaks and the position of sequence head.

Following execution step S50, S51 and S52: the initial tap coefficient values w selecting corresponding sef-adapting filter according to the gain isolation difference Δ r estimated from the Coefficient Look-up Table prestored _k(0) _rawwith study stepping coefficient η, and according to the relative phase Δ ω of the self-excitation signal estimated to the initial tap coefficient values w of selected sef-adapting filter _k(0) _rawcarry out phase compensation, the initial tap coefficient values w of the sef-adapting filter after obtaining phase compensation _k(0).Meanwhile, according to the self-excitation time delay T estimated _dthe settings T of time delay module is selected from the Coefficient Look-up Table prestored.

With reference to table 1, the gain such as estimated isolation difference Δ r is 4dB, then select the gain of 4dB to isolate the preferred initial tap coefficient values of sef-adapting filter corresponding to difference as w _k(0) _raw, the preferred study stepping coefficient η of its correspondence ₇as η.

According to the relative phase Δ ω of the self-excitation signal estimated, to w _k(0) _raweach tap coefficient { w ₀(0) _raw, w ₁(0) _raw..., w _l-1(0) _rawcarry out phase compensation, be namely multiplied by a complex twiddle factor, the initial tap coefficient values w of the sef-adapting filter after obtaining phase compensation _k(0)={ w ₀(0), w ₁(0) ..., w _l-1(0) }={ w ₀(0) _rawe ^{j Δ ω}, w ₁(0) _rawe ^{j Δ ω}..., w _l-1(0) _rawe ^{j Δ ω}.

Following execution step S60 and S61: the tap coefficient of sef-adapting filter is set to w _k(0), study stepping coefficient is set to η, and the settings of time delay module are set to T.As shown in Figure 5, first power regulation module of stopping using and the second power regulation module, enable receiver, self-excitation suppress system, training sequence maker, MUX and transmitting chain, and the output configuring MUX is the first input signal, i.e. training sequence T _sn (), the tap coefficient simultaneously opening sef-adapting filter upgrades, until sef-adapting filter convergence.

Following execution step S70, S71 and S72: as shown in Figure 6, the output of configuration MUX is the second input signal, and namely self-excitation suppresses the output signal e (n) of system, and stops the tap coefficient of sef-adapting filter to upgrade.Enable first power regulation module and the second power regulation module, the tap coefficient then opening sef-adapting filter upgrades.Now sef-adapting filter is in convergence state, and repeater starts normal work.

The effect of above-mentioned power regulation module is the power normalization of the input data in order to make sef-adapting filter is P ₀namely identical with the input signal power of sef-adapting filter when testing with off-line simulation, can ensure that the gain of whole link is identical with simulation model like this, thus can not need again regularized learning algorithm stepping coefficient that sef-adapting filter also can be made to be in convergence state.

Above embodiment is only illustrative embodiments of the present invention, can not be used for limiting the present invention, and protection scope of the present invention is defined by the claims.Those skilled in the art can in essence of the present invention and protection range, and make various amendment or equivalent replacement to the present invention, these are revised or be equal to replacement and also should be considered as dropping in protection scope of the present invention.

Claims (10)

1. the method improving straight-forward station self-excitation and suppress the convergence rate of system, described repeater comprises receiver, transmitting chain and self-excitation and suppresses system, described self-excitation suppresses Operation system setting between described receiver and described transmitting chain, described self-excitation suppresses system to comprise sef-adapting filter, time delay module and subtracter, it is characterized in that, described method comprises the steps:

S10. suppress to arrange training sequence maker and MUX between system and described transmitting chain in described self-excitation, described MUX is configured to the output signal of described training sequence maker and described self-excitation to suppress the output signal of system alternatively to output to described transmitting chain;

S30. described self-excitation of stopping using suppresses system, enable described receiver, described training sequence maker, described MUX and described transmitting chain;

S31. the output configuring described MUX is the training sequence T that described training sequence maker exports _s(n);

S32. the second Received signal strength R (n) from receiver is received;

S40. according to described training sequence T _sn () and described second Received signal strength R (n) estimate the gain isolation difference Δ r of described repeater and the relative phase Δ ω of self-excitation signal;

S50. according to the gain isolation difference Δ r estimated, from the Coefficient Look-up Table prestored, the initial tap coefficient values w of corresponding described sef-adapting filter is selected _k(0) _rawwith initial study stepping coefficient η.

2. raising straight-forward station self-excitation according to claim 1 suppresses the method for the convergence rate of system, it is characterized in that,

In step s 40, use following formula to the relative phase Δ ω of the gain isolation difference Δ r and self-excitation signal that estimate described repeater:

$\mathrm{\Δr}\·e^{\mathrm{j\Δ\ω}}=\frac{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{T}_s{\left(n\right)}^{*}{R}_p\left(n\right)}{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|T_s\left(n\right)\right|}^2},$

Wherein, N is described training sequence T _sthe length of (n), T _s(n) ^*for described training sequence T _sthe conjugate sequence of (n), R _p(n) for extract from described second Received signal strength R (n) and T _sthe sequence of (n) time unifying.

3. raising straight-forward station self-excitation according to claim 1 and 2 suppresses the method for the convergence rate of system, it is characterized in that,

Also comprise after step S50: S51. according to the relative phase Δ ω of self-excitation signal estimated to the initial tap coefficient values w of described sef-adapting filter _k(0) _rawcarry out phase compensation, with the initial tap coefficient values w of the described sef-adapting filter after obtaining phase compensation _k(0).

4. raising straight-forward station self-excitation according to claim 3 suppresses the method for the convergence rate of system, it is characterized in that,

Also comprise after step s 32: S41. is according to the second Received signal strength R (n) and training sequence T _sn () estimates the self-excitation time delay T of self-excitation signal _d;

Also comprise after step S41: S52. is according to the self-excitation time delay T estimated _d, record self-excitation time delay and the settings T selecting described time delay module the look-up table of corresponding time delay module settings from what prestore.

5. raising straight-forward station self-excitation according to claim 4 suppresses the method for the convergence rate of system, it is characterized in that, in step S41, by described second Received signal strength R (n) and described training sequence T _sn () carries out the method for computing cross-correlation to estimate described self-excitation time delay T _d.

6. raising straight-forward station self-excitation according to claim 4 suppresses the method for the convergence rate of system, it is characterized in that, also comprises after step S50:

S60. the tap coefficient of described sef-adapting filter is set to w _k(0), study stepping coefficient is set to η, and the settings of described time delay module are set to T;

S61. enable described self-excitation suppresses system, and the tap coefficient opening described sef-adapting filter upgrades until the convergence of described sef-adapting filter.

7. raising straight-forward station self-excitation according to claim 6 suppresses the method for the convergence rate of system, it is characterized in that, described repeater also comprises the first power regulation module for adjusting received signal power between described receiver and subtracter, and the second power regulation module for adjusting transmit signal power between described transmitting chain and described MUX.

8. raising straight-forward station self-excitation according to claim 7 suppresses the method for the convergence rate of system, it is characterized in that, also comprises before step S30:

S20. described self-excitation of stopping using suppresses system, described training sequence maker, described MUX, described transmitting chain, described first power regulation module and described second power regulation module, enable described receiver;

S21. the first Received signal strength R from receiver is received ₀(n);

S22. described first Received signal strength R is obtained by statistics ₀the power P of (n) _l;

S23. the first power adjusting factor arranging described first power regulation module is P ₀/ P _l, the second power adjusting factor arranging described second power regulation module is P _l/ P ₀, wherein P ₀for predetermined power value.

9. raising straight-forward station self-excitation according to claim 8 suppresses the method for the convergence rate of system, it is characterized in that, also comprises after step S61:

S70. the output configuring described MUX is the output that described self-excitation suppresses system, and stops the tap coefficient of described sef-adapting filter to upgrade;

S71. enable described first power regulation module and described second power regulation module;

S72. the tap coefficient opening described sef-adapting filter upgrades, and described repeater starts normal work.

10. raising straight-forward station self-excitation according to claim 9 suppresses the method for the convergence rate of system, it is characterized in that, also comprises before step S10:

S01. set up the off-line simulation model of described repeater, the power setting the e (n) that transmits in described off-line simulation model is P ₀, the relative phase of self-excitation signal is 0;

S02. by the many experiments of described off-line simulation model, obtain initial tap coefficient values and the study stepping coefficient of described repeater preferred described sef-adapting filter under the poor grade of various gain isolation, and the initial tap coefficient values of each gain isolation difference and each correspondence is entered described Coefficient Look-up Table with study stepping coefficient storage.