CN106937374A - A kind of back-off measure and system based on distortion component measurement DCM - Google Patents

Abstract

The invention discloses a kind of back-off measure based on distortion component measurement DCM and system, method includes back-off amount calculation procedure, signal switch process and back-off and amplification procedure.The present invention excludes the linear segment in third-order non-linear, and back-off amount is predicted using the distortion component of influence systematic function in third-order non-linear power.The envelope fluctuation characteristic of signal is characterized using the distortion component of third-order output, required back-off amount is more accurately predicted.Simultaneously as the measure can more accurately characterize back-off amount, there is reliability higher as signal dynamics reduction criterion.Compared with traditional PAPR and CM measures, the back-off amount needed for power amplifier PA can be more accurately predicted has more preferable reliability as signal Optimality Criteria.

Description

A kind of back-off measure and system based on distortion component measurement DCM

Technical field

The present invention relates to a kind of back-off measure based on distortion component measurement DCM and system.

Background technology

Wireless communication signals often have that envelope fluctuation is big, such as ofdm signal.Signal envelope fluctuation is big then Needing power amplifier (power amplifier, PA) has big linear dynamic range, and otherwise, signal enters inelastic region Domain, can cause serious out of band spectrum to extend and inband distortion.However, increasing power amplifier simply in side circuit design The range of linearity be infeasible.

It is to avoid signal from entering nonlinear area most simply directly that certain back-off is carried out to power amplifier input signal Method, but, if back-off is too small, do not reach the mission nonlinear distortion requirement of considered critical in communication standard；Instead It, then can make the reduction of PA efficiency.So, the back-off amount required for Accurate Prediction is extremely important.

Predict that the back-off amount of power amplifier generally requires reference signal Dynamic Degree value.Additionally, believing in baseband waveform design Number dynamic optimization method is also required to accurate signal dynamics measurement criterion.Therefore, which kind of measure metric signal bag is carried out using Network dynamic is very crucial.

At present, existing two kinds of measurement means：Peak-to-average power ratio (peak-to-average power ratio, PAPR) With cubic metric (cubic metric, CM).Based on the assumption that：Signal by after power amplifier produce non-linear distortion with The PAPR of signal reduce and it is dull reduce, PAPR using peak value sampling point come the back-off of prediction signal, however, the hypothesis by Confirm not abundant enough.Consider the Nonlinear Cubic model of simplified memoryless power amplifier：

Y (n)=G₁x(n)+G₃x(n)³,

Wherein, G₁And G₃Power amplifier linearity gain and Nonlinear Cubic gain are represented respectively, and x (n) represents power amplifier input Signal.The definition of CM is：

Wherein, RCM=rms [((| x (n) |)/rms [x (n)])³] be input signal primitive cube measurement, rms [] Root mean square, RCM are asked in expression_refIt is the RCM values of the reference signal of selection, H is constant.

From it is defined above understand CM utilize be third-order non-linear whole power come characterize signal by after system produce Non-linear distortion so that predict back-off amount.Compared to PAPR, CM is by experimental results demonstrate can more accurately predict Inband distortion and band external expansion, and used by 3GPP.In fact, orthogonal according to Gram-Schmidt (Gram-Schmidt) Change theory, the Nonlinear Cubic model of PA is decomposed into：

Wherein, d (n) is the distortion component orthogonal with x (n), and factor gamma ensure that between linear component and distortion component Orthogonality.It can be seen that, containing the part for conducing linearly amplification in third-order non-linear, eliminate the d (n) after linear component It is the distortion component for influenceing systematic function.Therefore, with CM come the back-off amount of the non-linear distortion of gauging system and signal still It is so not accurate enough.

For the deficiency of CM, research and propose with outer cubic metric (out-of-band cubic metric, OCM) To predict the back-off amount needed for power amplifier input signal, such as patent of invention of Application No. CN201410100963.9.OCM profits With to be the out-of-band distortion component that exports of third-order non-linear predict back-off, and influence systematic entirety can be not only Out-of-band distortion, the inband distortion for also having a direct impact signal quality.Therefore, OCM is still not accurate enough.Additionally, OCM is only applicable to Nonlinear power amplifier input signal is the situation of OFDM modulation systems.By contrast, the DCM measures in the present invention consider simultaneously Signal band and out-of-band distortion, and do not limited by input signal modulation type.

The content of the invention

It is an object of the invention to overcome the deficiencies in the prior art, there is provided a kind of power that DCM is measured based on distortion component Rollback measure and system, exclude the linear segment in third-order non-linear, systemic using being influenceed in third-order non-linear power Can distortion component predict back-off amount.

The purpose of the present invention is achieved through the following technical solutions：A kind of power that DCM is measured based on distortion component Rollback measure, including back-off amount calculation procedure, signal switch process and back-off and amplification procedure；

Described back-off amount calculation procedure includes following sub-step：

S11：Input baseband signal, and the baseband signal is carried out to calculate acquisition DCM metrics, including following sub-step Suddenly：

S111：Calculate the average power content P of baseband signal_av；

S112：Calculated distortion component power value P_di；

S113：According to P_avValue and P_diValue calculates DCM metrics；

S12：The back-off amount of power amplifier is predicted using the DCM metrics；

Described signal switch process includes following sub-step：

S21：Input baseband signal, and DA conversions are carried out to the baseband signal；

S22：Signal to completing DA conversions carries out frequency up-conversion operation；

Described back-off includes following sub-step with amplification procedure：

S31：The signal that signal switch process is obtained enters according to the back-off amount that back-off amount calculation procedure is obtained Row back-off；

S32：Signal input power amplifier after back-off is amplified.

Average power content P in step S111_avComputing formula it is as follows：

In formula, x (n) represents baseband discrete time-domain signal, and N represents that signal sampling is counted, and E [] is represented and asked expectation, n=0, 1,...,N-1。

Distortion product power value P in step S112_diComputing formula it is as follows：

The computing formula of DCM metrics is as follows in step S112：

Computing formula in step S12 using the back-off amount of DCM metrics prediction power amplifier is as follows：

PBO=k₁×DCM+k₂(dB)；

In formula, k₁And k₂Prediction parameter and constant offset are corresponded to respectively.

A kind of back-off gauging system that DCM is measured based on distortion component, including：

Back-off metric calculation module：Power for the baseband signal being input into calculate prediction power amplifier Rollback amount, including：

Average power content P_avComputing unit：Average power content P for calculating the baseband signal of input_av；

Distortion product power value P_diComputing unit：For according to average power content P_avCalculated distortion component power value P_di；

DCM metric computing units：For according to P_avValue and P_diValue calculates DCM metrics；

Back-off amount computing unit：Back-off amount for predicting power amplifier according to the DCM metrics；

Signal conversion module：For carrying out DA conversions and frequency up-conversion operation to the baseband signal being input into；

Back-off and amplification module：The signal that signal conversion module is obtained is obtained according to back-off metric calculation module To back-off amount carry out back-off after, input power amplifier is amplified.

Average power content P_avThe computing formula of computing unit is as follows：

In formula, x (n) represents baseband discrete time-domain signal, and N represents that signal sampling is counted, and E [] is represented and asked expectation, n=0, 1,...,N-1。

Distortion product power value P_diThe computing formula of computing unit is as follows：

The computing formula of DCM metric computing units is as follows：

The computing formula of back-off amount computing unit is as follows：

PBO=k₁×DCM+k₂(dB)；

In formula, k₁And k₂Prediction parameter and constant offset are corresponded to respectively.

The beneficial effects of the invention are as follows：The linear segment in third-order non-linear is excluded, using shadow in third-order non-linear power The distortion component of acoustic system performance predicts back-off amount.The Envelop waves of signal are characterized using the distortion component of third-order output Dynamic characteristic, is more accurately predicted required back-off amount.Simultaneously as the measure can more accurately characterize work( Rate rollback amount, has reliability higher as signal dynamics reduction criterion.Compared with traditional PAPR and CM measures, can be with Back-off amount needed for power amplifier PA is more accurately predicted, has more preferable reliability as signal Optimality Criteria.

Brief description of the drawings

Fig. 1 is the inventive method flow chart；

Fig. 2 is present system block diagram；

Fig. 3 is the experimental provision block diagram in embodiment；

When Fig. 4 is system ACLR=35dBc, the back-off amount of the straight line that slope is 1 and the pass of PAPR increments have been made It is schematic diagram；

When Fig. 5 is system ACLR=35dBc, the relation of the back-off amount and PAPR increments of having made fit line is illustrated Figure；

When Fig. 6 is system ACLR=35dBc, the back-off amount of fit line and the relation schematic diagram of RCM increments have been made；

When Fig. 7 is system ACLR=35dBc, the back-off amount of fit line and the relation schematic diagram of DCM increments have been made；

When Fig. 8 is system EVM=10%, the back-off amount of the straight line that slope is 1 and the relation of PAPR increments have been made Schematic diagram；

When Fig. 9 is system EVM=10%, the back-off amount of fit line and the relation schematic diagram of PAPR increments have been made；

When Figure 10 is system EVM=10%, the back-off amount of fit line and the relation schematic diagram of RCM increments have been made；

When Figure 11 is system EVM=10%, the back-off amount of fit line and the relation schematic diagram of DCM increments have been made；

Power prediction error distribution histogram when Figure 12 is ACLR=35dBc；

Power prediction error distribution histogram when Figure 13 is system EVM=10%.

Specific embodiment

Technical scheme is described in further detail below in conjunction with the accompanying drawings：

As shown in figure 1, a kind of back-off measure that DCM is measured based on distortion component, including back-off gauge Calculate step, signal switch process and back-off and amplification procedure；

Described back-off amount calculation procedure includes following sub-step：

S11：Input baseband signal, and the baseband signal is carried out to calculate acquisition DCM metrics, including following sub-step Suddenly：

S111：Calculate the average power content P of baseband signal_av, average power content P in step S111_avComputing formula such as Under：

In formula, x (n) represents baseband discrete time-domain signal, and N represents that signal sampling is counted, and E [] is represented and asked expectation, n=0, 1,...,N-1。

S112：Calculated distortion component power value P_di, distortion product power value P in step S112_diComputing formula it is as follows：

This step excludes the linear segment in third-order non-linear, using the mistake that systematic function is influenceed in third-order non-linear power True component further calculates back-off amount.

S113：According to P_avValue and P_diValue calculates DCM metrics, and the computing formula of DCM metrics is as follows in step S112：

S12：The back-off amount of power amplifier is predicted using the DCM metrics, the DCM is utilized in step S12 The computing formula of the back-off amount of metric prediction power amplifier is as follows：

PBO=k₁×DCM+k₂(dB)；

In formula, k₁And k₂Prediction parameter and constant offset are corresponded to respectively.

Described signal switch process includes following sub-step：

S21：Input baseband signal, and DA conversions are carried out to the baseband signal；

S22：Signal to completing DA conversions carries out frequency up-conversion operation；

Described back-off includes following sub-step with amplification procedure：

S31：The signal that signal switch process is obtained enters according to the back-off amount that back-off amount calculation procedure is obtained Row back-off；

S32：Signal input power amplifier after back-off is amplified.

Realization based on the above method, as shown in Fig. 2 the present embodiment additionally provides a kind of distortion component that is based on measures DCM Back-off gauging system, including：

Back-off metric calculation module：Power for the baseband signal being input into calculate prediction power amplifier Rollback amount, including：

Average power content P_avComputing unit：Average power content P for calculating the baseband signal of input_av；

Distortion product power value P_diComputing unit：For according to average power content P_avCalculated distortion component power value P_di；

DCM metric computing units：For according to P_avValue and P_diValue calculates DCM metrics；

Back-off amount computing unit：Back-off amount for predicting power amplifier according to the DCM metrics；

Signal conversion module：For carrying out DA conversions and frequency up-conversion operation to the baseband signal being input into；

Back-off and amplification module：The signal that signal conversion module is obtained is obtained according to back-off metric calculation module To back-off amount carry out back-off after, input power amplifier is amplified.

Accordingly, average power content P_avThe computing formula of computing unit is as follows：

In formula, x (n) represents baseband discrete time-domain signal, and N represents that signal sampling is counted, represents and ask expectation, n=0, 1,...,N-1。

Accordingly, distortion product power value P_diThe computing formula of computing unit is as follows：

Accordingly, the computing formula of DCM metrics computing unit is as follows：

Accordingly, the computing formula of back-off amount computing unit is as follows：

PBO=k₁×DCM+k₂(dB)；

In formula, k₁And k₂Prediction parameter and constant offset are corresponded to respectively.

The reliability of extracting method and system in order to verify, the present embodiment employs various different signals (including LTE letters Number, WLAN signal, MC-CDMA signals), modulation system (M-QAM, M=4,16,64) and sub-carrier number (N=64,256) carry out Experiment.Especially, the present embodiment uses ofdm signal, QPSK to modulate, and 64 subcarriers and 4 times of over-samplings are illustrated.

As shown in figure 3, the experimental provision of our uses, including computer, signal generator, signal analyzer, signal decline Subtract device and power amplifier, wherein the model Agilent E4438c of signal generator, the model of signal analyzer Keysight N9020B, power amplifier parameter：Saturation power value is 33dBm, and gain is 26dB.Computer occurs with signal Device is connected, and the two ends of power amplifier are connected with signal generator and signal attenuator respectively, the output end of signal attenuator As the input of signal analyzer.

Two kinds of system performance index are employed in experiment：Adjacent Channel Leakage Ratio (Adjacent Channel Leakage Ratio, ACLR) and Error Vector Magnitude (Error Vector Magnitude, EVM), respectively as the power of input signal The performance indications that size is met.

Tested as system performance index using ACLR=35dBc and EVM=10%, comprised the following steps：

S41：Computer random produces 50 frame OFDM symbols, calculates and record the DCM values of each frame signal respectively, PAPR with And RCM values.

S42：Input a signal into the RF power amplification PA shown in Fig. 2, observe the performance indications in signal analyzer in Fig. 2 The input power of signal in ACLR or EVM, and sweep generator, as ACLR=35dBc or EVM=10%, records Now the input power of signal is designated as P_in。

S43：In this 50 frame OFDM symbol, selection envelope rises and falls a minimum frame OFDM symbol as a reference to research work( The relation of rate rollback amount and different metric forms.Make P_in,1And P_in,i, i=2,3 ..., 50 represent respectively reference symbol and other The input power of 49 frame symbols.Similarly, M is made₁And M_i, i=2,3 ..., 50 represent respectively these symbols metric (PAPR, RCM and DCM).So, each frame signal is expressed as P with respect to the back-off amount of reference signal_in,1-P_in,i, i=2,3 ..., 50, Measurement incremental representation is M_i-M₁, i=2,3 ..., 50.Ideally, there is P_in,1-P_in,i=μ (M_i-M₁), i=2,3 ..., 50, and μ is constant.This shows that preferable measure prediction back-off amount does not have error.If in fact, measurement increment and work( Linear relationship between rate rollback amount is better, then will be more accurate using measure prediction back-off amount.

Fig. 4 is depicted as system ACLR=35dBc, and the relation of back-off amount and PAPR increments has simultaneously made out slope It is 1 straight line (this line correspondences conventional prediction backing method in engineering：Input signal PAPR values are how much to retract for how many dB DB), Fig. 5, Fig. 6, Fig. 7 depict the relation of back-off amount and PAPR, RCM, DCM increment respectively, and are made according to data Linear fit straight line.Root-mean-square error (root-mean-square errors RMSE) corresponding to Fig. 4, Fig. 5, Fig. 6, Fig. 7 It is respectively 3.0492,0.3354,0.2703 and 0.1907.As can be seen that the RMSE corresponding to DCM is minimum, RCM takes second place, PAPR It is maximum.

Fig. 8 is depicted as system EVM=10%, and the relation of back-off amount and PAPR increments has simultaneously made out slope and is 1 straight line, Fig. 9, Figure 10, Figure 11 depict the relation of back-off amount and PAPR, RCM, DCM increment respectively, and according to data Linear fit straight line is made.Root-mean-square error (root-mean-square corresponding to Fig. 8, Fig. 9, Figure 10, Figure 11 Errors RMSE) it is respectively 2.3028,0.2847,0.2497 and 0.1266.As can be seen that the RMSE corresponding to DCM is minimum, RCM takes second place, and PAPR is maximum.

In order to further illustrate the reliability of methods described, we illustrate distinct methods predict back-off amount when Error distribution histogram.Figure 12 is the back-off predicated error distribution histogram as system ACLR=35dBc.Figure 13 is to work as During system EVM=10%, back-off predicated error distribution histogram.Scheme to can be seen that at the error distribution 0 of DCM more from two Height, and distribution is tighter, RCM takes second place, and PAPR is minimum.

According to experimental result, it can be seen that the measure based on distortion component measurement DCM proposed by the present invention can be more Plus the back-off amount needed for accurate prediction PA.

Claims (10)

1. it is a kind of based on distortion component measure DCM back-off measure, it is characterised in that：Calculated including back-off amount Step, signal switch process and back-off and amplification procedure；

Described back-off amount calculation procedure includes following sub-step：

S11：Input baseband signal, and the baseband signal is carried out to calculate acquisition DCM metrics, including following sub-step：

S111：Calculate the average power content P of baseband signal_av；

S112：Calculated distortion component power value P_di；

S113：According to P_avValue and P_diValue calculates DCM metrics；

S12：The back-off amount of power amplifier is predicted using the DCM metrics；

Described signal switch process includes following sub-step：

S21：Input baseband signal, and DA conversions are carried out to the baseband signal；

S22：Signal to completing DA conversions carries out frequency up-conversion operation；

Described back-off includes following sub-step with amplification procedure：

S31：The signal that signal switch process is obtained carries out work(according to the back-off amount that back-off amount calculation procedure is obtained Rate retracts；

S32：Signal input power amplifier after back-off is amplified.

2. a kind of back-off measure that DCM is measured based on distortion component according to claim 1, its feature exists In：Average power content P in step S111_avComputing formula it is as follows：

$P_{av}=E\[|x\left(n\right){|}^2\]=\frac{1}{N}\underset{n=0}{\overset{N-1}{\Σ}}|x\left(n\right){|}^2;$

In formula, x (n) represents baseband discrete time-domain signal, and N represents that signal sampling is counted, and E [] is represented and asked expectation, n=0, 1,...,N-1。

3. a kind of back-off measure that DCM is measured based on distortion component according to claim 2, its feature exists In：Distortion product power value P in step S112_diComputing formula it is as follows：

$\begin{array}{c}{P}_{di}=E\[\left|x\left(n\right)\right|x\left(n\right){|}^2-\frac{E\[|x\left(n\right){|}^4\]}{E\[|x\left(n\right){|}^2\]}x\left(n\right){|}^2\]\\ =E\[|x\left(n\right){|}^6\]-\frac{\{E\[|x\left(n\right){|}^4\]\}}{P_{av}}\end{array}.$

4. a kind of back-off measure that DCM is measured based on distortion component according to claim 3, its feature exists In：The computing formula of DCM metrics is as follows in step S112：

$DCM=10\log \left(\frac{P_{di}}{{P_{av}}^3}\right)\left(dB\right).$

5. a kind of back-off measure that DCM is measured based on distortion component according to claim 4, its feature exists In：Computing formula in step S12 using the back-off amount of DCM metrics prediction power amplifier is as follows：

PBO=k₁×DCM+k₂(dB)；

In formula, k₁And k₂Prediction parameter and constant offset are corresponded to respectively.

6. it is a kind of based on distortion component measure DCM back-off gauging system, it is characterised in that：Including：

Back-off metric calculation module：Back-off for the baseband signal being input into calculate prediction power amplifier Amount, including：

Average power content P_avComputing unit：Average power content P for calculating the baseband signal of input_av；

Distortion product power value P_diComputing unit：For according to average power content P_avCalculated distortion component power value P_di；

DCM metric computing units：For according to P_avValue and P_diValue calculates DCM metrics；

Back-off amount computing unit：Back-off amount for predicting power amplifier according to the DCM metrics；

Signal conversion module：For carrying out DA conversions and frequency up-conversion operation to the baseband signal being input into；

Back-off and amplification module：The signal that signal conversion module is obtained is obtained according to back-off metric calculation module After back-off amount carries out back-off, input power amplifier is amplified.

7. a kind of back-off gauging system that DCM is measured based on distortion component according to claim 6, its feature exists In：Average power content P_avThe computing formula of computing unit is as follows：

$P_{av}=E\[|x\left(n\right){|}^2\]=\frac{1}{N}\underset{n=0}{\overset{N-1}{\Σ}}|x\left(n\right){|}^2;$

In formula, x (n) represents baseband discrete time-domain signal, and N represents that signal sampling is counted, and E [] is represented and asked expectation, n=0, 1,...,N-1。

8. a kind of back-off gauging system that DCM is measured based on distortion component according to claim 7, its feature exists In：Distortion product power value P_diThe computing formula of computing unit is as follows：

$\begin{array}{c}{P}_{di}=E\[\left|x\left(n\right)\right|x\left(n\right){|}^2-\frac{E\[|x\left(n\right){|}^4\]}{E\[|x\left(n\right){|}^2\]}x\left(n\right){|}^2\]\\ =E\[|x\left(n\right){|}^6\]-\frac{\{E\[|x\left(n\right){|}^4\]\}}{P_{av}}\end{array}.$

9. a kind of back-off gauging system that DCM is measured based on distortion component according to claim 8, its feature exists In：The computing formula of DCM metric computing units is as follows：

$DCM=10\log \left(\frac{P_{di}}{{P_{av}}^3}\right)\left(dB\right).$

10. a kind of back-off gauging system that DCM is measured based on distortion component according to claim 9, its feature exists In：The computing formula of back-off amount computing unit is as follows：

PBO=k₁×DCM+k₂(dB)；

In formula, k₁And k₂Prediction parameter and constant offset are corresponded to respectively.