EP4699267A1

EP4699267A1 - Online scanning in multi-sources pimc

Info

Publication number: EP4699267A1
Application number: EP23944842.6A
Authority: EP
Inventors: Hao YE; Jin ELLGARDT; Spyridon PAPASIDERIS; Yuecheng YANG; Marcus WESTLUND; Jinsong Yang
Original assignee: Telefonaktiebolaget LM Ericsson AB
Current assignee: Telefonaktiebolaget LM Ericsson AB
Priority date: 2023-07-11
Filing date: 2023-09-26
Publication date: 2026-02-25
Also published as: WO2025010836A1; CN121488436A

Abstract

The embodiments herein relate to Passive Intermodulation (PIM) delay estimation for multi-sources. In some embodiments, there proposes a method (1200) for estimating a time delay for a PIM source of a radio equipment with multi-PIM sources. In an embodiment, the method may comprise the step of determining (S1201) a time delay window with one or more possible PIM sources of the radio equipment. The method may further comprise the step of determining (S1202) a plurality of time delay values in the time delay window. The method may comprise the step of for each of one or more basis functions, scanning (S1205) each of the plurality of time delay values in the time delay window by performing adaptive algorithm based on the basis function to obtain a set of adaptive parameters, and estimating (S1208) one or more time delays for each PIM source according to the set of adaptive parameters. With the embodiments herein, an adaptive scanning delay estimation method is proposed to replace correlation estimation method, which may improve delay estimation accuracy for multi PIM sources.

Description

ONLINE SCANNING IN MULTI-SOURCES PIMC

CROSS-REFERENCE TO RELATED APPLICATION (S)
This application claims priority to the Greek National Application No. GR 20230100567, entitled “Online Scanning in Multi-Sources PIMC” , filed on July 11, 2023, which is incorporated herein by reference in its entirety.

Technical Field

The embodiments herein relate generally to the field of mobile communication, and more particularly, the embodiments herein relate to Passive Intermodulation (PIM) delay estimation for multi-sources.

Background

PIM is a type of distortion generated by the nonlinearity of passive components, such as filters, duplexers, connectors, antennas and so forth at radio equipment. Traditionally, PIM is a critical problem when the radio equipment such as wireless transceiver is provided with multi-band capabilities.
Figure 1 shows multiple PIM sources in the radio equipment 100. The radio equipment 100 shown in Figure 1 may include one transmitter (TX) branch and one receiver (RX) branch. The PIM signal generated from one TX branch to one respective RX branch may be referred as inline PIM. As shown in Figure 1, the PIM signal considered by the PIM Cancellation (PIMC) may be generated by multi-sources. For example, Power Amplifier (PA) Intermodulation (IM) signal may be located into RX band, as shown by line 101; PIM may be generated from the PA output to the Low Noise Amplifier (LNA) input due to isolator non-linear character, as shown by line 102; PIM may be generated by a metal cavity filter which has a non-linear character, as shown by line 103; PIM may be generated by antenna or cable connected to the radio equipment due to connecter metal rust, as shown by line 104; external PIM may be generated by a metal fence or other non-linear materials, as shown by lines 105 and 106.
The PIM distortions may degrade the RX sensitivity and the uplink signal-to-interference-plus-noise ratio (SINR) . The PIMC algorithm is introduced to improve the RX SINR. The PIMC algorithm may build up a PIM model to generate a simulated PIM signal and subtract the simulated PIM signal from the RX signal. The adaptive filter method, such as Least Squares (LS) , Least Mean Squares (LMS) , Recursive Least Squares (RLS) methods, may be used in the PIM modeling.
The PIMC needs to synchronize TX and RX signals to determine the cancellation start time point. This synchronization function may be implemented by Time Delay Estimation (TDE) . The TDE may capture TX and RX data at the same time and store the captured data into the memory. The TDE algorithm may use the power correlation method to calculate PIM loop delay. Because PIM is caused by nonlinear system, the PIM signal cannot be correlated with TX signal directly. But the power swap of the PIM signal can be correlated with TX signal. The TX and RX power correlation equation is:
wherein the operator is a correlation operator, and the function abs () is an absolution value function.
Figure 2 shows an example correlation curve of TDE at 300^th iteration. As shown in Figure 2, multi sources may cause multi-peak values in the correlation curve, wherein the cable PIM is the biggest (i.e., peak) value, and other PIM sources are just local peaks in small windows. The left peaks illustrate some leakage signals from TX to RX transceiver chips. The PA IM and isolator PIM are mixed because the delay path from the PA to the isolator is too short. The peaks for the PA IM and isolator PIM cannot be distinguished by the TDE.
Because the PIM source in cable is much stronger than other PIM sources, the cable PIM is visualized in this figure. The other peaks are not visualized and may be overwhelmed in the noise floor. As a result, it is difficult to get accurate delay information for all these PIM sources. Currently, the TDE can just get one peak value and locate one PIM which can be the most powerful PIM source. Other PIM sources can be dismissed by the TDE. Then the PIMC just cancel the strongest PIM source such as the cable PIM, other PIM sources are not cancelled.

Summary

In view of the above, the embodiments herein propose methods, apparatus, computer readable medium and computer program product for PIM delay estimation for multi-sources.
In some embodiments, there proposes a method for estimating a time delay for a PIM source of a radio equipment with multi-PIM sources. The method may comprise the step of determining a time delay window with one or more possible PIM sources of the radio equipment. The method may further comprise the step of determining a plurality of time delay values in the time delay window. The method may further comprise the steps of for each of one or more basis functions, scanning each of the plurality of time delay values in the time delay window by performing adaptive algorithm based on the basis function to obtain a set of adaptive parameters and estimating one or more time delays for each PIM source according to the set of adaptive parameters.
The proposed method is to introduce a multi-sources PIM estimation of PIMC method. Adaptive scanning delay estimation replaces correlation estimation method (TDE) to improve accuracy and enhance delay dimensions. It can estimate delay for each TX signal in each basis function. The proposed adaptive scanning method can get accurate coefficients and small performance error.
In an embodiment, the set of adaptive parameters includes at least one of: Q adaptive errors of the adaptive algorithm; and Q*M adaptive filter coefficients of the adaptive algorithm. Q is the number of cycles at sampling rate within the time delay window, and M is the number of taps of an adaptive filter used by the adaptive algorithm.
In an embodiment, the step of estimating the one or more time delays may further comprising the step of estimating the one or more time delays according to one or more nulls of the Q adaptive errors for determining one or more respective PIM sources.
The nulls (i.e., minimum values) of adaptive error may show the maximum values of the performance improvement of the PIMC, which is associated with the accuracy of the time delay estimation. Therefore, the estimation of the time delays according to the nulls of the Q adaptive errors may improve the accuracy of the estimation.
In an embodiment, the method may further comprise the step of discarding a PIM source with a null above a first threshold. The first threshold may be set based on a tradeoff between resource usage and performance of a PIMC algorithm.
When the null is higher than the threshold, it illustrates that the PIM is weak and the PIMC cannot get more performance improvement even introduce this PIM source. Whereas when the null is lower than the threshold, it illustrates that this PIM source is strong enough. PIMC performance can be improved after cancelling this PIM source.
In an embodiment, the step of estimating the one or more time delays may further comprising the step of for each of the Q time delay value in the time delay window, adding M adaptive filter coefficients of the adaptive algorithm, to obtain Q added adaptive filter coefficients. The step of estimating the one or more time delays may further comprising the step of calculating a power of coefficients by squaring the absolute value of the added adaptive filter coefficients, for each of the Q time delay values within the time delay window, to obtain Q power of coefficients. The step of estimating the one or more time delays may further comprising the step of estimating the one or more time delays according to one or more peaks of Q power of coefficients, for determining one or more respective PIM sources.
In addition to the nulls of adaptive error, delay estimation may be based on the peak of the coefficients power.
In an embodiment, the method may further comprise the step of discarding a PIM source with a peak below a second threshold. The second threshold may be set based on a tradeoff between resource usage and performance of a PIMC algorithm.
Adaptive error null may be used to estimate delay for each basis function. And this error can also be used to distinguish multi-PIM sources. Coefficients can be used to calculate PIM power in the center delay point of each PIM source. This delay estimation method is gotten from performance calculation. So, it is very accurate and a perfect matching with PIMC performance.
TDE algorithm often cannot get convergence result, but this proposed method can get accurate delay for multi-PIM sources. No TDE means that the power for PIMC may be saved. This is a power efficient design of PIMC.
In an embodiment, the method may further comprise the step of combining two PIM sources, if the difference of the time delays of the two PIM sources is less than M multiples of an interval between adjacent ones among the plurality of time delay values.
For two PIM sources represented by the delays d1 and d2, when abs (d2 -d1) is smaller than M, two polynomials for the two PIM sources may be combined into one PIM polynomial. The tap number is M + (d2 -d1) . If the PIM power is focused on concentrated in the center M taps, M taps may be used to replace M + (d2 -d1) taps, and then the two PIM polynomials may be simplified to a M tap filter.
Therefore, if two PIM sources are close and the delay between them is less than M sampling cycles, they may be combined into one PIM source. One polynomial may illustrate both PIM sources. The adaptive algorithm may get convergence coefficient (s) for this polynomial for the two PIM sources.
In an embodiment, an integer scanning may be performed, in which an interval between adjacent ones among the plurality of time delay values are integer multiples of a sampling cycle.
In an embodiment, a fraction scanning may be performed, in which an interval between adjacent ones among the plurality of time delay values are fractional multiples of the sampling cycle.
By the fraction delay estimation, an accurate delay estimation of for example 1/N cycle of sampling may be obtained. The fraction delay estimation may be used to improve the results of the integer delay estimation. For example, the time delay window of the fraction delay estimation may be set around the previously estimated PIM time delay by the integer delay estimation.
In an embodiment, the method may further comprise the step of determining the one or more basis functions, wherein the one or more basis functions are non-linear basis functions and based on a polynomial or a look-up table (LUT) .
In an embodiment, a TX training sequence or a traffic data sequence may be used as a stimulus signal for the radio equipment. If the traffic data sequence is used as a stimulus signal for the radio equipment, the time delay window may be scanned for a plurality of times during a period of time, to obtain the averaged adaptive parameters, which may be used for the subsequent estimation.
By using the traffic data sequence as a stimulus signal, the PIMC may be performed without affecting normal traffic of the radio equipment. The averaged adaptive parameters over a period of time may improve unstable adaptive errors caused by the unstable power of the traffic data.
In an embodiment, the step of determining a time delay window may further comprise the step of determining a minimum time delay D_MIN and a maximum time delay D_MAX based on the structure of the radio equipment and the surrounding environment the radio equipment located. The step of determining a time delay window may further comprise the step of digitizing the minimum time delay D_MIN and the maximum time delay D_MAX by a sampling cycle. The step of determining a time delay window may further comprise the step of determining the time delay window based on the digitized minimum time delay D_MIN and the digitized maximum time delay D_MAX.
In an embodiment, the time delay window may be based on a previous PIM time delay estimation for the radio equipment.
For example, the time delay window for the fraction delay estimation may be based on the integer delay estimation result. Therefore, the fraction delay estimation may be performed within a smaller window, for example, a delay window with one or two sampling cycles; which may improve the efficiency of the fraction delay estimation.
In an embodiment, the method may further comprise the step of for each of the one or more basis functions, modeling the PIM according to the estimated time delay; and canceling the PIM from a received signal of the radio equipment.
In an embodiment, the method may further comprise the step of monitoring the Q adaptive errors of the adaptive algorithm. The method may further comprise the step of redoing the adaptive scanning to re-estimate the PIM time delay for the radio equipment, if one or more adaptive errors for a basis function are larger than a threshold by: determining a second time delay window around the current time delay; determining a plurality of second time delay values in the second time delay window; scanning each of the plurality of second time delay values in the second time delay window by performing adaptive algorithm based on the basis function to obtain a set of second adaptive parameters; and estimating one or more second time delays for each PIM source according to the set of second adaptive parameters.
Dynamic correction on the delay information may be performed by supervising the adaptive error to match with the fast-changing PIM sources.
In an embodiment, the radio equipment may be included in a network node of a wireless communication network.
In an embodiment, the PIM may be 3-order PIM, 5-order PIM, or higher-order PIM.
In an embodiment, the multi-PIM sources may include one or more internal PIM sources and/or one or more external PIM sources.
The proposed solutions are suitable for both internal and external PIM sources. It is suitable for multi-antenna with multi-band product. It is also suitable for polynomial and look up table PIM model, because both of them use the adaptive algorithm method.
In some embodiments, there proposes a method for estimating a time delay for a PIM source of a radio equipment with multi-PIM sources. The method may comprise the step of determining a first basis function based on a first input signal of the radio equipment. The method may further comprise the step of estimating one or more first time delays for the first input signal based on the first basis function. The method may further comprise the step of determining a second basis function based on the first input signal and a second input signal of the radio equipment. The method may further comprise the step of estimating one or more second time delays for the second input signal by referring to the one or more first time delays.
With the above steps, the time delay estimation for the basis function with two elements (input signals, such as TX1, TX2) may be obtained from the time delay estimation for the basis function with one element (input signal, such as TX1) .
In an embodiment, the method may further comprise the step of determining a third basis function based on the first input signal, the second input signal, and a third input signal of the radio equipment. The method may further comprise the step of estimating one or more third time delays for the third input signal by referring to the one or more first time delays and the one or more second time delays.
With the above steps, the time delay estimation for the basis function with three elements (input signals, such as TX1, TX2, TX3) may be obtained from the time delay estimation for the basis function with two elements (input signals, such as TX1, TX2) .
In addition, for higher-order PIM, the time delay estimation for the basis function with more elements may be obtained by performing the above processes repeatedly.
In an embodiment, the step of estimating one or more first time delays for the first input signal may further comprise the step of determining a first time delay window including a first plurality of time delay values. The step of estimating one or more first time delays for the first input signal may further comprise the step of scanning the first time delay window, by performing adaptive algorithm repeatedly based on the first basis function for each of the first plurality of time delay values, to obtain a first set of adaptive parameters. The step of estimating one or more first time delays for the first input signal may further comprise the step of estimating one or more first time delays according to the first set of adaptive parameters.
In an embodiment, the step of estimating one or more second time delays for the second input signal may further comprise the step of determining a second time delay window for the first input signal based on one of the one or more first time delays. The second time delay window may be within the first time delay window, and the second time delay window may have a range around the one of the one or more first time delays. The step of estimating one or more second time delays for the second input signal may further comprise the step of determining a third time delay window for the second input signal. The step of estimating one or more second time delays for the second input signal may further comprise the step of performing a fraction scanning on the second and third time delay window, by performing adaptive algorithm repeatedly based on the second basis function for each of a plurality of second time delay values within the second time delay window and each of a plurality of third time delay values within the third time delay window, to obtain a second set of adaptive parameters. The step of estimating one or more second time delays for the second input signal may further comprise the step of estimating one or more updated first time delays for the first input signal and one or more second time delays for the second input signal according to the second set of adaptive parameters.
In an embodiment, the minimum value of the plurality of second time delay values within the second time delay window may be equal to the one of the one or more first time delays plus a sampling cycle. The maximum value of the plurality of second time delay values within the second time delay window may be equal to the one of the one or more first time delays minus a sampling cycle.
In an embodiment, the step of estimating one or more third time delays for the third input signal may comprise the step of determining a fourth time delay window for the first input signal based on one of the one or more updated first time delays. The fourth time delay window may be within the first time delay window, and the fourth time delay window may have a range around the one of the one or more updated first time delays. The step of estimating one or more third time delays for the third input signal may further comprise the step of determining a fifth time delay window for the second input signal based on one of the one or more second time delays. The fifth time delay window may be within the third time delay window, and the fifth time delay window may have a range around the one of the one or more second time delays. The step of estimating one or more third time delays for the third input signal may further comprise the step of determining a sixth time delay window for the third input signal. The step of estimating one or more third time delays for the third input signal may further comprise the step of performing a fraction scanning on the fourth, fifth and sixth time delay window, by performing adaptive algorithm repeatedly based on the third basis function for each of a plurality of fourth time delay values within the fourth time delay window, each of a plurality of fifth time delay values within the fifth time delay window, and each of a plurality of sixth time delay values within the sixth time delay window, to obtain a third set of adaptive parameters. The step of estimating one or more third time delays for the third input signal may further comprise the step of estimating one or more updated first time delays for the first input signal, one or more updated second time delays for the second input signal, and one or more third time delays according to the second set of adaptive parameters.
In some embodiments, there proposes an apparatus for estimating a time delay for a PIM with multi-PIM sources of a radio equipment, comprising: at least one processor; and a non-transitory computer readable medium coupled to the at least one processor. In an embodiment, the non-transitory computer readable medium may store instructions executable by the at least one processor, whereby the at least one processor may be configured to perform the above methods.
In some embodiments, there proposes a computer readable medium stores computer readable code, which when run on an apparatus, causes the apparatus to perform any of the above methods.
In some embodiments, there proposes a computer program product stores computer readable code, which when run on an apparatus, causes the apparatus to perform any of the above methods.
With the embodiments herein, an adaptive scanning delay estimation method is proposed to replace correlation estimation method, which may improve delay estimation accuracy for multi PIM sources.

Brief Description of the Drawings

The accompanying drawings, which are incorporated herein and form part of the specification, illustrate various embodiments of the present disclosure and, together with the description, further serve to explain the principles of the disclosure and to enable a person skilled in the pertinent art to make and use the embodiments disclosed herein. In the drawings, like reference numbers indicate identical or functionally similar elements, and in which:
Figure 1 shows multiple PIM sources in the radio equipment;
Figure 2 shows an example correlation curve of TDE at 300^th iteration;
Figure 3 shows scanning PIM sources in the window;
Figure 4 shows an example of the adaptive algorithm to estimate the coefficients;
Figure 5 shows the coefficients power of the adaptive algorithm scanning of Figure 4;
Figure 6 shows adaptive error of the adaptive algorithm scanning of the Figure 4;
Figure 7 shows the comparison of adaptive error and coefficient power;
Figure 8A shows the delay estimation based on the nulls of adaptive error;
Figure 8B shows the maximum memory length estimation based on the width of peak of the coefficients;
Figure 9 shows the fraction delay estimation for a PIM source;
Figure 10 shows example architecture for multi-TX;
Figure 11 shows adaptive scanning for 3D delay exhaust searching;
Figure 12 is a schematic flow chart showing an example PIM delay estimation method for multi-PIM sources, according to the embodiments herein;
Figure 13 is a schematic flow chart showing another example PIM delay estimation method for multi-PIM sources, according to the embodiments herein;
Figure 14 is a schematic block diagram showing an example apparatus for PIM delay estimation for multi-PIM sources, according to the embodiments herein; and
Figure 15 is a schematic block diagram showing an example computer-implemented apparatus, according to the embodiments herein.

Detailed Description of Embodiments

Embodiments herein will be described in detail hereinafter with reference to the accompanying drawings, in which embodiments are shown. These embodiments herein may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. The elements of the drawings are not necessarily to scale relative to each other.
Reference to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, the appearances of the phrase "in an embodiment" appearing in various places throughout the specification are not necessarily all referring to the same embodiment.
The term "A, B, or C" used herein means "A" or "B" or "C" ; the term "A, B, and C" used herein means "A" and "B" and "C" ; the term "A, B, and/or C" used herein means "A" , "B" , "C" , "A and B" , "A and C" , "B and C" or "A, B, and C" .
PIM Delay Window
Based on the locations of the PIM sources, the PIM sources may be categorized as internal or external sources, and the respective PIM may be categorized as internal or external PIM. For example, the PIM generated in the filters inside the radio equipment may be regarded as the internal PIM; the PIM generated by the metal fence outside the radio equipment may be regarded as the external PIM. The external PIM sources are located outside of the air interface of the radio equipment.
The loop delay for each PIM source may be different from each other. Some weak PIM source can be discarded because they are too weak and cause no distortion to RX sensitivity. In practice, the PIMC may cancel the PIM sources within a time delay window.
In an example, as shown in Figure 1, various PIM sources are considered in the PIMC, in these PIM sources, the first PIM source is PA IM. The loop delay of the PA IM may define the minimum value of the time delay window, i.e., d_MIN. The minimum delay may be measured in lab or factory.
The maximum value d_MAX of the time delay window may be defined outside of the air interface of the radio equipment. Due to the air path loss, the external PIM signal will be critically degraded to a very weak signal after a long distance from the victim radio equipment. As a result, the PIM source with a delay bigger than the maximum delay d_MAX may not distort RX sensitivity, since its PIM strength is weaker than the noise floor. The time delay window for all PIM sources in interest may be from d_MIN to d_MAX, which may be represented as:
DelayWindow_pimc=d_MAX-d_MIN (2)
If the sampling rate is considered, the time delay window may be digitized the range of [0, …, Q-1] , wherein
Here, T_sample is the cycle of sampling, i.e., the sampling cycle. Q is the sample number within the time delay window. In the following description, Q is set to be equal to 100. However, Q may be larger or smaller than 100, according to the length of the time delay window and the cycle of sampling.
Multi PIM sources modeling
The PIMC algorithm may build up a PIM model to regenerate the PIM signal and subtract the regenerated PIM signal from the RX signal, as shown in Figure 1. The PIM model may include several basis functions with nonlinear polynomial to simulate the nonlinearity of PIM sources. The adaptive filter algorithm, such as LS, LMS, RLS methods, may be used to estimate the coefficients of the adaptive filter. Each nonlinear polynomial with respective filter coefficients needs to be studied by the adaptive method.
The PIMC algorithm may regenerate the PIM signal by a polynomial PIM modeling. A complex Volterra polynomial model may be used to regenerate the PIM interference. The term number of the polynomial model may be very large, especially for multi-antenna and multi-band radio equipment, because all TX antennas, bands, conjugation branches join the combination.
If the adaptive filter has enough taps, one long tap filter may be used to illustrate one polynomial basis function. However, in practice, the adaptive filter may have a limited number of taps (for example M taps) . Therefore, the one polynomial basis function may be separated into I polynomials with same basis function (with different time delays) . The I polynomials may illustrate I PIM sources located into the delay window respectively.
For the sake of simplicity, for inline PIM of one branch (one TX and one RX) , a 3-order non-linear PIM at baseband of disperse system with memory effect generated by TX signal may be written as:
The polynomial basis function of this 3-order PIM is:
ψ (tx (n) ) =tx (n) tx (n) tx^* (n) (5)
The PIM model shown by equation (4) includes I PIM sources. Each PIM source has one polynomial filter with a memory length (i.e., m in the equation (4) ) . Each PIM source is a nonlinear term with M memory taps in equation (4) . If M memory taps are used to illustrate a PIM with memory length larger than M, the adaptive algorithm will get a worse performance. M may be selected, so that for most cases, the M memory taps include almost all the PIM power and thus M is good enough to model the PIM source with memory length larger than M.
Therefore, the memory taps M may be defined as (but not limited to) a fix number and use it to handle all the PIM sources. In most cases, M = 5 is good enough for PIM memory effect, at 245.76Msps sampling rate. The PIM in equation (4) may be referred as the PIM with M memory length. h_i (m) in equation (4) are the adaptive filter coefficients for PIM source i.
In this model, the delay of the stimulus tx (n) (i.e., the input signal) is not considered for simplicity. The delay is same for a same PIM source.
Each PIM sources can be distinguished based on the loop delay d_i in equation (4) . It is the loop delay from TX signal across the PIM source to the RX signal. If the interval between different delays d1 and d2 for different PIM sources 1 and 2 are larger than M (i.e., abs (d1 -d2) > M) , then the PIM sources 1 and 2 can be distinguished from each other by the loop delays d1 and d2.
If the interval is less than M (i.e., abs (d1 -d2) < M) , these PIM sources 1 and 2 may be combined to one PIM source with a tap number bigger than M (as shown in the following equation (6) ) .
As shown in equation (6) , when abs (d2 -d1) is very small, the two PIM polynomials may be combined into one PIM polynomial. The tap number is M + (d2 -d1) . If the PIM power is concentrated in the center M taps, M taps may be used to replace M + (d2 -d1) taps, and then the two PIM polynomials may be simplified to a M tap filter.
Therefore, if two PIM sources are very close and the delay between them is less than M sampling cycles, they may be combined into one PIM source. One polynomial shown by equation (6) may illustrate both PIM sources. The adaptive algorithm may get convergence coefficient (s) h_i (m) for this polynomial at one time.
Adaptive scanning
Figure 3 shows scanning PIM sources in the window. Considering a delay window shown in equation (3) , in which the delay range is from d_MIN to d_MAX, Q time delay values are included in this window. As an example but not limit, the interval between the adjacent time delay values are a sampling cycle T_sample.
Given the coefficients h (q) for all the cycles of window q = [0, …Q-1] , the PIM model may be written as (assuming the delay for three inputs tx (n) is same, i.e., d_MIN + q, and omitting the small time differences among them) :
It takes so many multipliers to implement this long polynomial filter. It is much bigger than the maximum memory length M. The actual implementation may use multi polynomials with a fix number of taps (e.g., M taps) .
Then, the equation (7) may be rewritten as:
In equation (8) , each PIMC just has a small memory length M.
Scanning on each time delay q of the window q = [0, …, Q-1] may be used to calculate the PIM_IM3 (n) .
For each step of the scanning, PIM_scan (n) may be written as:
If the time delay is represented in the units of T_sample (i.e., the sampling cycle) , the equation (9) may be rewritten as:
As may be seen from equation (10) , each step of the scanning may be understood as an independent adaptive convergence. Therefore, the adaptive algorithm in the PIMC may be used to calculate the coefficient h_i (m) in the equation (10) .
Then, the equation (8) may be rewritten as:
That is, the PIMC algorithm may scan the window to get the coefficients for each time delay value within the window. The scan counter is from 0 to Q-1 and the number of the filter taps is M.
Figure 4 shows an example of the adaptive algorithm to estimate the coefficients h_i (m) . As shown in Figure 4, the input signal tx (n) may be delayed by q sampling cycles at block 401 for each scanning step, in which q may be from 0 to Q-1. Then, the delayed signal may be further delayed by d_MIN, i.e., D_MIN/T_sample sampling cycles at block 402. Then at block 403, the adaptive algorithm in the PIMC may be used to simulate the PIM in the equation (10) , and thus may calculate the coefficient h_i (m) at block 404. Then, at block 405, the simulated PIM may be subtracted from the RX signal rx (n) , to obtain the error of the adaptive algorithm.
In the scanning in Figure 4, the scanning window is from 1 to 100, that is, the delay i may be from D_MIN/T_sample to D_MIN/T_sample + 99 sampling cycles. Here, the delay i is in the units of integer sampling cycles, and thus the delay interval (i.e., the scanning step) is also in the units of the integer sampling cycles; therefore the scanning may be referred as an integer scanning. The equalizer filter coefficients length is 5 taps (i.e., M = 5) .
Since 5 taps are used for the simulation, for the PIM with memory larger than M, the adaptive algorithm will get an adaptive error from block 405, shown in Figure 4. The number of the taps may impact coefficient width and also impact adaptive error wide and shape.
For the adaptive algorithm at block 403 in Figure 4:
if LS (Least Squares) algorithm is used in this simulation, the equation of h (h_i (m) in equation (10) ) may be for example:
if LMS (Least Mean Squares) algorithm is used in this simulation, the equation of h (h_i (m) in equation (10) ) may be for example:
both LS and LMS may be used to calculate the error of adaptive filter. The power of error may be for example:
By this scanning, a coefficient matrix with size of Q*M may be obtained. That is, in each of the Q steps of the scanning, M coefficients may be obtained.
The coefficient matrix may be used to estimate PIM power for each PIM source.
In an example, M is set to 1. The coefficient matrix may become a Q*1 vector and the Q*1 vector may be used to illustrate PIM power and phase in each point of window.
In an example, the coefficients for a single scanning step (in a row of the matrix) may be summed to get a Q*1 vector.
For M tap coefficients, the power of coefficients for each scanning step may be written as for example:
Power_Mtapcoef (i) = (abs (sum ( [h₀ h₁…hM_-1] ) ) ) ²
(16)
The Q*1 vector may be written as for example:
h_PIM3 (q) = [h₀ h₁…h_Q-1]
(17)
Meanwhile, in this scanning, an error vector with size of Q*1 may be obtained. That is, in each of the Q steps of the scanning, an adaptive error may be obtained.
error (n) =rx (n) -PIM_scan (n)
(18)
That is, the error may be understood as the receiving signal RX minus the stimulated PIM signal. It is the signal after performing the PIMC.
Delay estimation from Adaptive scanning
Figure 5 shows the coefficients power of the adaptive algorithm scanning of Figure 4. As shown in Figure 5, the peaks of coefficients power are matched with the delays of multi-PIM sources. This is because that the polynomial with M memory taps is used to simulate the PIM signal (see equations (9) or (10) ) , and most of the PIM power is concentrated in the M taps. Therefore, the peak of coefficients power may reflect the power of the PIM sources.
From Figure 5, three PIM sources may be distinguished, with memory length M. PIM delay value for each PIM source may be obtained from Figure 5, for example d₀, d₁, …, d_i.
Figure 6 shows adaptive error of the adaptive algorithm scanning of the Figure 4. As shown in Figure 6, the nulls (i.e., minimum values) of adaptive error are matched with the delays of multi-PIM sources. This is because that the error may be understood as the receiving signal minus the stimulated PIM signal. Therefore, the error may reflect the performance improvement of the PIMC. That is, the lower the power of the error is, the better the performance improvement of the PIMC is.
The nulls (i.e., minimum values) of adaptive error may show the maximum values of the performance improvement of the PIMC, which is associated with the accuracy of the time delay estimation, because more accurate time delay estimation may cause better performance improvement of the PIMC.
Figure 7 shows the comparison of adaptive error and coefficient power. From Figure 7, it can be seen that both the peaks of coefficients power and the nulls of adaptive error are matching with the delays of the multi-PIM sources. In addition, the nulls of adaptive error are more intuitive than the peaks of the coefficients power, since the curve of the adaptive error is sharp.
Figure 8A shows the delay estimation based on the nulls of adaptive error.
Firstly, a threshold is set to distinguish the PIM sources from other signals. The threshold may be adjusted to separate strong PIM signal in interest and other noise or weak PIM not in interest. If the threshold is set lower, more possible PIM sources may be found. As a result, the PIMC performance may be improved, but more computing resources are needed to deal with the PIM sources. Therefore, the threshold may be set based on a tradeoff between a resource usage and the PIMC performance of a PIMC algorithm.
When the null is higher than the threshold, it illustrates that the PIM is weak and the PIMC cannot get more performance improvement even introducing this PIM source. Whereas when the null is lower than the threshold, it illustrates that this PIM source is strong enough. PIMC performance can be improved after cancelling this PIM source.
All of the nulls of the adaptive error lower than the threshold are chosen as PIM sources. When there are several continuous points lower than the threshold (not shown in Figure 8A) , the lowest of them marks the delay of this PIM source; because when this delay is set, the adaptive algorithm can get the best performance improvement and the error is the smallest.
In addition, as shown in Figure 8A, the adaptive filter tap number may be used to separate multi-PIM sources. That is, the null may be assigned with a width of M (M = 5 for Figure 8A) . If the distance between two nulls is less than the filter taps M, the respective two PIM sources may be combined to one PIM source. In addition, similar delay estimation may be also applicable for Figure 5 based on the peak of the coefficients power.
Different memory length
In some embodiments, different memory lengths may be defined for different PIM sources. For the maximum memory length estimation, coefficients peak width is a judgement rule to be used.
Figure 8B shows the maximum memory length estimation based on the width of peak of the coefficients. From Figure 8B, a threshold may be used to distinguish multi-PIM sources. Three PIM sources are found. The delay start point and memory length are for example:
pim1: delay starts at 23, coefficients number is 5 taps, delay is from 23 to 27, maximum of coefficient is 15.07;
pim2: delay starts at 42, coefficients number is 7 taps, delay is from 42 to 48, maximum of coefficient is 22.73;
pim3: delay starts at 62, coefficients number is 7 taps, delay is from 62 to 68, maximum of coefficient is 31.39.
By the above information, the multi-PIM modeling equation may be written as:
By this method, multi PIM sources may be distinguished.
After estimating the memory length M and the delay for the PIM source d_i, the power estimation may be performed for each PIM sources.
Then, the PIMC may perform adaptive estimation in real-time, because the PIM model may be changed in real-time by temperature changing or other impact issues.
Note that, PIM cancellation for PIM5 or higher-order PIM is substantially the same as PIM3, with different basis function (s) .
For external PIM source with multi-antenna and multi-band products, the process is same with the above inline PIM cancellation.
Fraction delay or fine delay estimation
The above inline PIM delay estimation is integer delay estimation, because the unit of delay is one sample cycle. The unit may be too big for the PIMC to match the cancellation signal with the PIM distortion signal. In order to make more accurate delay estimation, fraction delay or fine delay estimation may be performed.
The difference between the fraction delay estimation and the integer delay estimation is that the TX signal passes a multi-phase polynomial filter to delay it by a number of units. The number of units may be [0, …, R*N-1] , and a single unit may be 1/N cycle of sampling.
That is, the whole searching range of the scanning is R cycles of sampling, and the scanning step is 1/N cycle of sampling. One example is: N=16 R=16, that is, the whole searching range of the scanning is 16 cycles of sampling, and the scanning step is 1/16 cycle of sampling.
By the fraction delay estimation within a small range window, a delay accuracy of 1/N cycle of sampling may be obtained.
Scan multi-PIM source by fraction delay window
The fraction delay estimation may be based on the integer delay estimation result. For example, after an integer time delay is obtained from the integer delay estimation shown by Figure 4, the fraction delay estimation may be used to make the integer time delay more accurate. The fraction delay estimation may be performed within a smaller window, for example, a delay window with one or two sampling cycles around the obtained integer time delay.
For example, for each PIM source, a window width is marked as M cycle of sampling, and the scanning step of the fraction scanning may be set as 1/Q cycle of sampling. The window length is M*Q scanning steps.
Figure 9 shows the fraction delay estimation for PIM source. As shown in Figure 9, a fraction delay filter 906 is added before the adaptive filter, to perform the fraction delay estimation. The fraction delay estimation is similar to the integer delay estimation, except that the scanning step is fraction cycle of sampling. The details for the fraction delay estimation is omitted for simplicity.
Scanning PIM source with multi-basis functions
For the radio equipment with multiple transmitting antennas and/or multiple transmitting bands (multi-TX) , the basis function (s) may be set for each PIM lobe. Figure 10 shows example architecture for multi-TX. In Figure 10, two TX antennas and two RX antennas, and each of the TX antennas have two TX bands, each of the TX bands includes two TX carriers.
One theoretical equation of Volterra polynomial for 3-order non-linear system caused by 3 carriers f1, f2 and f3 (just considering one TX and RX antenna with 3 bands) may be:
The victim is carrier b. The frequencies of 3 aggressors are f1, f2, f3. The non-linear memory effect may be reflected on t1, t2, t3 delay. One term of this polynomial may be referred as a NL (non-linear) term.
When multi-path hit for one term is not considered, the delay of the 3 carriers may be aligned. The equation (20) may be simplified as:
When implementing PIMC in the radio equipment for dual bands, tripe bands, or more bands, one band may be simply set as one PIM hit signal if digital process has big enough sampling rate to support all the bands. If the sampling rate is not big enough, one band may be separated into two sub-bands. This configuration may reduce the NL term numbers comparing set carrier as hit signal, because one band may have several carriers. Then the logic resource utilization may be reduced. Each band is a hit signal for one antenna. Each antenna TX band is a hit signal for multi-antennas radio equipment.
For each of the basis functions in the equation (21) , i.e., the delay t1 is same for each items (for example f1, f2, and/or f3) of term, the adaptive scanning method shown in Figure 4 above may be used to estimate PIM sources in the delay windows.
If different delays are estimated for three items (for example f1, f2, and f3) , the delay t1, t2, t3 may be scanned separately.
Scanning PIM source with multi-delays for multi-elements
When searching multi-element delay t1, t2, t3 for one basis function such as TX1TX2TX3*, the Greedy method may be used to search this basis function. In details, an exhaust searching method, i.e., 3D searching strategy may be used.
Firstly, the group delay with same delay t1 for the three TX signals may be estimated (i.e., the three TX signals may be aligned, see equation (21) ) with the above mentioned adaptive scanning method in Figure 4. Then the delta delay from this group delay t1 may be found by a 3-dimension exhaust searching method in a small delta delay windows, because the delta delays between the signals TX1, TX2, and TX3 in one basis function are very small. Therefore, the exhaust delay searching will not take so much calculation time.
Figure 11 shows adaptive scanning for 3D delay exhaust searching. Figure 11 shows the 3D delay exhaust searching for the basis function PIM=band1*band1*conj (band3) , i.e., TX1TX1TX3*. In Figure 11, the 3D figure may show a signal pair with two input signals, for example the first TX1 and TX3, or the second TX1 and TX3. For the 3D delay exhaust searching in Figure 11, the delta delays [0 -2] and [-2 0] have the lowest adaptive error (i.e., nulls of the adaptive error) . Therefore, the estimated delta delay for the first TX1 is 0, the estimated delta delay for the second TX1 is 0, and the estimated delta delay for the TX3 is -2. That is, the estimated time delay for the first TX1 is d1 (same as the group delay) , the estimated time delay for the second TX1 is d1 (same as the group delay) , and the estimated time delay for the TX3 is d1-2 (two sampling cycles in advance relative to the group delay) . Two symmetric delta delay values for the first TX1 and the second TX1 is obtained, since the basis function is PIM= band1*band1*conj (band3) , i.e., the two TX1 have similar time delays, although not exactly the same.
The 3D delay exhaust searching in Figure 11 may find the best delay estimation for three TX signals in one basis function. But it needs many times of adaptive processes to do that. Other searching method may be used to get 3D delta delay values, in addition to the above 3D delay exhaust searching in Figure 11.For example, the following 3D delta delay scanning method may be used to improve multi-TX scanning efficiency.
The aggressors may be TX1, TX2, …, TXn. The victims may be RX1, RX2, …, RXn.
In the first step, the delay for a basis function with one aggressor signal may be estimated.
For example, the basis function may be:
Because the aggressor is TX1, the delays for three terms (the first TX1, the second TX1, and the third TX1) in the polynomial are the same or with a very small delta delay. The delay for the basis function with one aggressor signal may be estimated according to any method, for example the above adaptive scanning method.
After estimating the delay value d1, the estimated delay value d1 may be used as a reference value for other basis functions with same aggressor (s) and same victim (s) .
After all the delays (for one or more PIM sources) of the first basis function with one aggressor signal are found, the delays with two aggressor signals may be estimated.
For example, the second basis function with two aggressor signals may be:
In the equation (23) , the delay from TX1 to RX1 (i.e., d1) is already know in the first step, then the delay for TX2 may be further estimated based on the second basis function. It does not matter whether the TX2 is a conjugation term.
When searching d2, d1 may be scanned in a very small window. For example, d1 may be scanned with a fraction delay estimation within a [-1 1] cycle window, i.e., for example [-15/16, …, 16/16] cycle window, to get an accurate delay for TX1 in the second basis function.
After estimating the delay values d1 and d2, the estimated delay values d1 and d2 may be used as reference values for other basis functions with same aggressor (s) and same victim (s) .
In the third step, the delays with three aggressor signals may be estimated.
For example, the third basis function with three aggressor signals may be:
In the equation (24) , the delays from TX1 to RX1 (i.e., d1) and from TX2 to RX1 (i.e., d2) are already known in the first and second steps, then the delay for TX3 may be further estimated based on the third basis function. It does not matter whether the TX3 is a conjugation term.
When searching d3, d1 and d2 may be scanned in a very small window respectively. For example, d1 and d2 may be scanned with fraction delay estimations within a [-1 1] cycle window, i.e., for example [-15/16, …, 16/16] cycle window, to get accurate delays for TX1 and TX2 in the third basis function.
The delays for each term of the basis function for a 3-order PIM may be estimated based on the above three steps. For 5-order PIM or higher-order PIM, similar method with more steps may be used.
By the above scanning steps, the efficiency of adaptive scanning may be improved greatly on the delay estimation for all the basis functions. This is an improved strategy for the 3D exhaust searching method.
Dynamic correcting delay error
External PIM sources may dynamically change at high speed, for example the location and strength of an external PIM source may change quickly and unpredictably. The PIMC algorithm has to do time delay estimation with a small interval time to catch up the delay change of the external PIM sources.
If considering radio sample variance, the delay and power variances of the radio equipment may be located into a small window. If also considering other impact issues, like temperature, the total variance window is bigger than the window for sample variance, but the total variance window still has a small range.
Another drawback of the TDE is that the TDE is not suitable for dynamic PIM sources. Some PIM sources change after a long time. The old delay value is not suitable for new PIM sources, thus TDE need to be redone to get correct delay value. Otherwise, the adaptive error will become bigger and bigger, and the PIMC performance will degrade.
By using the adaptive scanning delay process in the embodiments, after the initial process including adaptive scanning delay information, the PIMC may operate and cancel PIM distortion in real time. But some PIM sources are dynamic and the delay will change after a long time, and then the PIMC performance will degrade by this delay swapping.
A method to correct the delay error is to supervise the adaptive error on every time running the adaptive scanning algorithm. When the adaptive error is becoming bigger and bigger, for example, after the error is bigger than a threshold, it is considered that the PIMC cannot get a good performance and need to redo delay estimation then.
When delay estimation needs to be redone for one basis function, the adaptive scanning may be done in a small window adjacent the old delay value, because generally the delay swaps in a small range. The new scanning doesn’t take much time because the scanning window is small, for example several sampling cycles around the old delay value.
Figure 12 is a schematic flow chart showing an example PIM delay estimation method 1200 for multi-PIM sources, according to the embodiments herein. In an embodiment, the flow chart in Figure 12 may be implemented in a network node with radio equipment using multiple bands and multiple antennas.
As used herein, network node refers to equipment capable, configured, arranged and/or operable to communicate directly or indirectly with a UE and/or with other network nodes or equipment, in a telecommunication network. Examples of network nodes include, but are not limited to, access points (APs) (e.g., radio access points) , base stations (BSs) (e.g., radio base stations, Node Bs, evolved Node Bs (eNBs) and NR NodeBs (gNBs) ) .
Base stations may be categorized based on the amount of coverage they provide (or, stated differently, their transmit power level) and so, depending on the provided amount of coverage, may be referred to as femto base stations, pico base stations, micro base stations, or macro base stations. A base station may be a relay node or a relay donor node controlling a relay. A network node may also include one or more (or all) parts of a distributed radio base station such as centralized digital units and/or remote radio units (RRUs) , sometimes referred to as Remote Radio Heads (RRHs) . Such remote radio units may or may not be integrated with an antenna as an antenna integrated radio. Parts of a distributed radio base station may also be referred to as nodes in a distributed antenna system (DAS) .
Other examples of network nodes include multiple transmission point (multi-TRP) 5G access nodes, multi-standard radio (MSR) equipment such as MSR BSs, network controllers such as radio network controllers (RNCs) or base station controllers (BSCs) , base transceiver stations (BTSs) , transmission points, transmission nodes, multi-cell/multicast coordination entities (MCEs) , Operation and Maintenance (O&M) nodes, Operations Support System (OSS) nodes, Self-Organizing Network (SON) nodes, positioning nodes (e.g., Evolved Serving Mobile Location Centers (E-SMLCs) ) , and/or Minimization of Drive Tests (MDTs) .
The method 1200 may begin with step S1201, in which the network node (such as eNB or gNB) may determine a scanning window, for example a time delay window with one or more possible PIM sources of the radio equipment may be determined. For example, the one or more possible PIM sources of the radio equipment may be found based on the configuration of the radio equipment and the environment surrounding the radio equipment. Then, each of the PIM delay values of the PIM sources under laboratory conditions may be determined by a testing on the radio equipment. A scanning window may be determined based on these PIM delay values under laboratory conditions and stored in a database in advance. The network node may determine the scanning window by reading it from the database. There are usually a plural of time delay values in a time delay window.
In an embodiment, determining the time delay window further comprising the steps of: determining a minimum time delay D_MIN and a maximum time delay D_MAX based on the structure of the radio equipment and the surrounding environment the radio equipment located; digitizing the minimum time delay D_MIN and the maximum time delay D_MAX by a sampling cycle; and determining the time delay window based on the digitized minimum time delay D_MIN and the digitized maximum time delay D_MAX.
In an embodiment, the time delay window may be based on a previous PIM time delay estimation for the radio equipment. For example, the time delay window may be set around the previously estimated PIM time delay.
Then, the method 1200 may proceed to step S1202, in which the network node may determine scanning points. For example, a plurality of time delay values in the time delay window may be determined as the scanning points.
In an example, the scanning points may be [0, 100] cycles of sampling. However, the scanning points may be any points within the scanning windows. The scanning points may be denser at the location of the possible PIM source, and may be sparser at other locations. The scanning points may around the location of the possible PIM source (s) .
In an example, the scanning points may be integer points, which are located at the integer sampling cycle. Or, the scanning points may be fraction points, which are located at the fraction sampling cycle.
Then, the method 1200 may proceed to step S1203, in which the network node may determine one or more basis functions. For example, the network node may determine one or more basis functions based on RX hit group. The RX hit group may include the combinations of TX signals hitting the band of the RX signal. For example, if the combinations of TX signals, e.g., TX1TX1TX1*, TX1TX1TX3*, and TX1TX2TX3*hits the band of the RX signal (for example the RX1 in interest) , the RX hit group may be TX1TX1TX1*, TX1TX1TX3*and TX1TX2TX3*; as a result, the one or more basis functions may be determined as TX1TX1TX1*, TX1TX1TX3*and TX1TX2TX3*. One of these basis functions may be chosen to start the adaptive scanning.
In an example, the one or more basis functions may be non-linear basis functions. In an example, the one or more basis functions may be based on a polynomial or a LUT.
To be noted that, the above step S1203 to determine one or more basis functions is just an option to obtain basis functions. Basis functions can also be configured in advance.
Then, the method 1200 may proceed to step S1204, in which the network node may determine whether all scanning points in the window are scanned, that is all the time delay values in the time delay window to be scanned.
In an example, a TX training sequence may be used as an input signal for the scanning.
In an example, a traffic data sequence may be used as an input signal for the scanning. In this case, the time delay window may be scanned for a plurality of times, during a period of time, to obtain the averaged adaptive parameters, which may be used for the subsequent estimation.
To be noted that, the above step S1204 to determine scanning points is just an option to obtain scanning points. A fixed number of scanning points can also be preconfigured evenly in the time delay window, or a number of scanning points can be preconfigured evenly in the time delay window with a fixed interval between adjacent scanning points.
If not all scanning points in the window are scanned, the method 1200 may proceed to step S1205, in which the network node may scan each of the plurality of scanning points (i.e. the time delay values) in the time delay window by performing adaptive algorithm based on each basis function to obtain a set of adaptive parameters for each basis function, for example the network node may run adaptive algorithm to estimate coefficients, adaptive error and power of each PIM source.
If all scanning points in the window are scanned for a basis function (that is, the adaptive algorithm has been performed based on the basis function) , the method 1200 may proceed to step S1206, the network node may store all the information (for example the set of adaptive parameters, including coefficients, adaptive errors and powers of each PIM source) of this basis function.
For example, the obtained set of adaptive parameters may include Q adaptive errors of the adaptive algorithm and/or the obtained set of adaptive parameters may include Q*M adaptive filter coefficients of the adaptive algorithm. Q may be the number of cycles at sampling rate within the time delay window, and M may be the number of taps of an adaptive filter used by the adaptive algorithm.
Then, the method 1200 may proceed to step S1207, in which the network node may determine whether all basis functions are scanned.
If not all basis functions are scanned, the method 1200 may return to step S1204 to scan other basis function (s) .
If all basis functions are scanned, the method 1200 may proceed to step S1208 to estimate one or more time delays for each PIM source according to the set of adaptive parameters, for each of one or more basis functions.
In an embodiment, the step of estimating the one or more time delays may further comprising the step of estimating the one or more time delays according to one or more nulls of the Q adaptive errors for determining one or more respective PIM sources.
In an embodiment, the method may further comprise the step of discarding a PIM source with a null above a first threshold. The first threshold may be set based on a tradeoff between a resource usage and a PIMC performance of a PIMC algorithm.
In an embodiment, the step of estimating the one or more time delays may further comprising the step of for each of the Q time delay value in the time delay window, adding M adaptive filter coefficients of the adaptive algorithm, to obtain Q added adaptive filter coefficients. The step of estimating the one or more time delays may further comprising the step of calculating a power of coefficients by squaring the absolute value of the added adaptive filter coefficients, for each of the Q time delay values within the time delay window, to obtain Q power of coefficients. The step of estimating the one or more time delays may further comprising the step of estimating the one or more time delays according to one or more peaks of Q power of coefficients, for determining one or more respective PIM sources.
In an embodiment, the method may further comprise the step of discarding a PIM source with a peak below a second threshold. The second threshold may be set based on a tradeoff between a resource usage and a PIMC performance of a PIMC algorithm.
In an embodiment, the method may further comprise the step of combining two PIM sources, if the difference of the time delays of the two PIM sources is less than M multiples of an interval between adjacent ones among the plurality of time delay values.
Then, the method 1200 may proceed to step S1209, in which the network node may store respective delay (s) and memory length (s) for each of the PIM sources, for subsequent processing.
Then, the method 1200 may proceed to step S1210, in which the network node may configure PIMC according to PIM source information. For example, the network node may model the PIM according to the estimated time delay, for each of the one or more basis functions.
Then, the method 1200 may proceed to step S1211, in which the network node may perform PIMC for all PIM sources in real time. For example, the network node may cancel the PIM from a received signal of the radio equipment, for each of the one or more basis functions.
Then, the method 1200 may proceed to an optional step S1212, in which the network node may perform dynamic supervision for example on the adaptive error. For example, the network node may monitor the Q adaptive errors of the adaptive algorithm estimated in step S1205.
Then, the method 1200 may proceed to an optional step S1213, in which the network node may determine whether the adaptive error is larger than a threshold.
If the adaptive error is not larger than the threshold, the method 1200 may return to step S1210 for PIM modeling and PIMC.
If the adaptive error is larger than the threshold, the method 1200 may proceed to an optional step S1214, in which the network node may redo adaptive scanning in a small window, to re-estimate the PIM time delay for the radio equipment. The steps of S1201-S1209 may be performed for redoing adaptive scanning in the step S1214. For example, the small window may be set as a window around the existing delay value, and the length of the window may be 2 sampling cycles or the like.
Figure 13 is a schematic flow chart showing another example PIM delay estimation method 1300 for multi-PIM sources, according to the embodiments herein. In an embodiment, the flow chart in Figure 13 may be implemented in a network node with radio equipment using multiple bands and multiple antennas. The PIM delay estimation method 1300 may be adapted to the basis function with multi-TXs, for example TX1TX1TX2*or TX1TX2TX3*.
The method 1300 may begin with step S1301, in which the network node (such as eNB or gNB) may determine a first basis function based on a first input signal of the radio equipment, for example TX1TX1TX1*.
Then, the method 1300 may proceed to step S1302, in which the network node may estimate one or more first time delays for the first input signal TX1 based on the first basis function TX1TX1TX1*. For example, the estimation may be similar to the steps S1201-S1209 in Figure 12.
The method 1300 may begin with step S1303, in which the network node (such as eNB or gNB) may determine a second basis function based on the first input signal TX1 and a second input signal TX2 of the radio equipment, for example TX1TX1TX2*.
Then, the method 1300 may proceed to step S1304, in which the network node may estimate one or more updated first time delays for the first input signal TX1 and one or more second time delays for the second input signal TX2 based on the second basis function TX1TX1TX2*. For example, the estimation may be similar to the steps S1201-S1209 in Figure 12. The estimated one or more first time delays for the first input signal TX1 in the step S1302 may be used as a reference. Therefore, the scanning of the first time delay may be done in a small window, for example, -1 sampling cycle to 1 sampling cycle.
With the above steps S1301-S1304, the time delay estimation for the basis function TX1TX1TX2*may be obtained.
The method 1300 may begin with an optional step S1305, in which the network node (such as eNB or gNB) may determine a third basis function based on the first input signal TX1, the second input signal TX2, and a third input signal TX3 of the radio equipment, for example TX1TX2TX3*.
Then, the method 1300 may proceed to an optional step S1306, in which the network node may estimate one or more updated first time delays for the first input signal TX1, one or more updated second time delays for the second input signal TX2, and one or more third time delays for the third input signal TX3 based on the third basis function TX1TX2TX3*. For example, the estimation may be similar to the steps S1201-S1209 in Figure 12. The estimated one or more first time delays for the first input signal TX1 in the steps S1302 and/or S1304 and the estimated one or more second time delays for the second input signal TX2 in the step S1304 may be used as a reference. Therefore, the scanning of the first time delay and the second time delay may be done in a small window, for example, -1 sampling cycle to 1 sampling cycle.
With the above steps S1301-S1306, the time delay estimation for the basis function TX1TX2TX3*may be obtained. In addition, for higher-order PIM, the time delay estimation for the basis function with more elements may be obtained by performing the processes in the steps S1305-S1306 repeatedly.
Figure 14 is a schematic block diagram showing an example apparatus 1400 for PIM power estimation, according to the embodiments herein. In an embodiment, the apparatus 1400 in Figure 14 may be implemented as the above network node.
In an embodiment, the apparatus 1400 may include at least one processor 1401; and a non-transitory computer readable medium 1402 coupled to the at least one processor 1401. The non-transitory computer readable medium 1402 may store instructions executable by the at least one processor 1401, whereby the at least one processor 1401 is configured to perform the steps in the example methods 1200 and 1300 as shown in the schematic flow charts of Figures 12 and 13; the details thereof are omitted here.
Note that, the apparatus 1400 may be implemented as hardware, software, firmware and any combination thereof. For example, the apparatus 1400 may include a plurality of units, circuities, modules or the like, each of which may be used to perform one or more steps of the example methods 1200 and 1300.
Figure 15 is a schematic block diagram showing an example computer-implemented apparatus 1500, according to the embodiments herein. In an embodiment, the apparatus 1500 may be configured as the above mentioned apparatus such as a network node (eNB or gNB) .
In an embodiment, the apparatus 1500 may include but not limited to at least one processor such as Central Processing Unit (CPU) 1501, a computer-readable medium 1502, and a memory 1503. The memory 1503 may comprise a volatile (e.g., Random Access Memory, RAM) and/or non-volatile memory (e.g., a hard disk or flash memory) . In an embodiment, the computer-readable medium 1502 may be configured to store a computer program and/or instructions, which, when executed by the processor 1501, causes the processor 1501 to carry out any of the above mentioned methods.
In an embodiment, the computer-readable medium 1502 (such as non-transitory computer readable medium) may be stored in the memory 1503. In another embodiment, the computer program may be stored in a remote location for example computer program product 1504 (also may be embodied as computer-readable medium) , and accessible by the processor 1501 via for example carrier 1505.
The computer-readable medium 1502 and/or the computer program product 1504 may be distributed and/or stored on a removable computer-readable medium, e.g. diskette, CD (Compact Disk) , DVD (Digital Video Disk) , flash or similar removable memory media (e.g. compact flash, SD (secure digital) , memory stick, mini SD card, MMC multimedia card, smart media) , HD-DVD (High Definition DVD) , or Blu-ray DVD, USB (Universal Serial Bus) based removable memory media, magnetic tape media, optical storage media, magneto-optical media, bubble memory, or distributed as a propagated signal via a network (e.g. Ethernet, ATM, ISDN, PSTN, X. 25, Internet, Local Area Network (LAN) , or similar networks capable of transporting data packets to the infrastructure node) .
Example embodiments are described herein with reference to block diagrams and/or flowchart illustrations of computer-implemented methods, apparatus (systems and/or devices) and/or non-transitory computer program products. It is understood that a block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, may be implemented by computer program instructions that are performed by one or more computer circuits. These computer program instructions may be provided to a processor circuit of a general purpose computer circuit, special purpose computer circuit, and/or other programmable data processing circuit to produce a machine, such that the instructions, which execute via the processor of the computer and/or other programmable data processing apparatus, transform and control transistors, values stored in memory locations, and other hardware components within such circuitry to implement the functions/acts specified in the block diagrams and/or flowchart block or blocks, and thereby create means (functionality) and/or structure for implementing the functions/acts specified in the block diagrams and/or flowchart block (s) .
These computer program instructions may also be stored in a tangible computer-readable medium that may direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable medium produce an article of manufacture including instructions which implement the functions/acts specified in the block diagrams and/or flowchart block or blocks. Accordingly, embodiments of present inventive concepts may be embodied in hardware and/or in software (including firmware, resident software, micro-code, etc. ) that runs on a processor such as a digital signal processor, which may collectively be referred to as “circuitry, ” “a module” or variants thereof.
It should also be noted that in some alternate implementations, the functions/acts noted in the blocks may occur out of the order noted in the flowcharts. For example, two blocks shown in succession may in fact be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Moreover, the functionality of a given block of the flowcharts and/or block diagrams may be separated into multiple blocks and/or the functionality of two or more blocks of the flowcharts and/or block diagrams may be at least partially integrated. Finally, other blocks may be added/inserted between the blocks that are illustrated, and/or blocks/operations may be omitted without departing from the scope of inventive concepts. Moreover, although some of the diagrams include arrows on communication paths to show a primary direction of communication, it is to be understood that communication may occur in the opposite direction to the depicted arrows.
Many variations and modifications can be made to the embodiments without substantially departing from the principles of the present inventive concepts. All such variations and modifications are intended to be included herein within the scope of present inventive concepts. Accordingly, the above disclosed subject matter is to be considered illustrative, and not restrictive, and the appended examples of embodiments are intended to cover all such modifications, enhancements, and other embodiments, which fall within the spirit and scope of present inventive concepts. Thus, to the maximum extent allowed by law, the scope of present inventive concepts are to be determined by the broadest permissible interpretation of the present disclosure including the following examples of embodiments and their equivalents, and shall not be restricted or limited by the foregoing detailed description.
Abbreviations
ANT Antenna
DL/UL Downlink/Uplink
LS Least Squares
LMS Least Mean Squares
NL Non-Linear
PIM Passive Intermodulation Cancellation
PIMC Passive Intermodulation Cancellation
RF Radio Frequency
RLS Recursive Least Squares
RU Radio Unit
TDE Time Delay Estimation
TAB Transceiver Array Boundary
UE User Equipment.

Claims

A method (1200) for estimating a time delay for a Passive Intermodulation (PIM) source of a radio equipment with multi-PIM sources, comprising:

- determining (S1201) a time delay window with one or more possible PIM sources of the radio equipment;

- determining (S1202) a plurality of time delay values in the time delay window; and

- for each of one or more basis functions,

- scanning (S1205) each of the plurality of time delay values in the time delay window by performing adaptive algorithm based on the basis function to obtain a set of adaptive parameters; and

- estimating (S1208) one or more time delays for each PIM source according to the set of adaptive parameters.
The method (1200) according to claim 1, wherein the set of adaptive parameters includes at least one of:

- Q adaptive errors of the adaptive algorithm; and

- Q*M adaptive filter coefficients of the adaptive algorithm,

wherein Q is the number of cycles at sampling rate within the time delay window, and M is the number of taps of an adaptive filter used by the adaptive algorithm.
The method (1200) according to claim 2, wherein estimating (S1208) the one or more time delays further comprising:

- estimating the one or more time delays according to one or more nulls of the Q adaptive errors for determining one or more respective PIM sources.
The method (1200) according to claim 3, further comprising:

- discarding a PIM source with a null above a first threshold;

wherein the first threshold is set based on a tradeoff between a resource usage and a PIM cancellation (PIMC) performance of a PIMC algorithm.
The method (1200) according to claim 2, wherein estimating (S1208) the one or more time delays further comprising:

- for each of the Q time delay value in the time delay window, adding M adaptive filter coefficients of the adaptive algorithm, to obtain Q added adaptive filter coefficients;

- calculating a power of coefficients by squaring the absolute value of the added adaptive filter coefficients, for each of the Q time delay values within the time delay window, to obtain Q power of coefficients; and

- estimating the one or more time delays according to one or more peaks of Q power of coefficients, for determining one or more respective PIM sources.
The method (1200) according to claim 5, further comprising:

- discarding a PIM source with a peak below a second threshold,

wherein the second threshold is set based on a tradeoff between a resource usage and a PIM cancellation (PIMC) performance of a PIMC algorithm.
The method (1200) according to any of claims 3 to 6, further comprising:

- combining two PIM sources, if the difference of the time delays of the two PIM sources is less than M multiples of an interval between adjacent ones among the plurality of time delay values.
The method (1200) according to any of claims 1 to 7,

wherein an integer scanning is performed, in which an interval between adjacent ones among the plurality of time delay values are integer multiples of a sampling cycle, or

wherein a fraction scanning is performed, in which an interval between adjacent ones among the plurality of time delay values are fractional multiples of the sampling cycle.
The method (1200) according to any of claims 1 to 8, further comprising:

determining (S1203) the one or more basis functions, wherein the one or more basis functions are non-linear basis functions and based on a polynomial or a look-up table (LUT) .
The method (1200) according to any of claims 1 to 9, wherein a TX training sequence or a traffic data sequence is used as a stimulus signal for the radio equipment.
The method (1200) according to any of claims 1 to 10, wherein determining (S1201) a time delay window further comprising:

- determining a minimum time delay D_MIN and a maximum time delay D_MAX based on the structure of the radio equipment and the surrounding environment the radio equipment located;

- digitizing the minimum time delay D_MIN and the maximum time delay D_MAX by a sampling cycle; and

- determining the time delay window based on the digitized minimum time delay D_MIN and the digitized maximum time delay D_MAX.
The method (1200) according to any of claims 1 to 10, wherein the time delay window is based on a previous PIM time delay estimation for the radio equipment.
The method (1200) according to any of claims 1 to 12, further comprising:

- for each of the one or more basis functions,

- modeling (S1210) the PIM according to the estimated time delay; and

- canceling (S1211) the PIM from a received signal of the radio equipment.
The method (1200) according to any of claims 1 to 13, further comprising:

- monitoring (S1212) the Q adaptive errors of the adaptive algorithm; and

- redoing (S1214) the adaptive scanning to re-estimate the PIM time delay for the radio equipment, if one or more adaptive errors for a basis function are larger than a threshold by:

- determining (S1201) a second time delay window around the current time delay;

- determining (S1202) a plurality of second time delay values in the second time delay window;

- scanning (S1205) each of the plurality of second time delay values in the second time delay window by performing adaptive algorithm based on the basis function to obtain a set of second adaptive parameters; and

- estimating (S1208) one or more second time delays for each PIM source according to the set of second adaptive parameters.
The method (1200) according to any of claims 1 to 14,

wherein the radio equipment is included in a network node of a wireless communication network;

wherein the PIM is 3-order PIM, 5 order-PIM, or higher order-PIM; and/or

wherein the multi-PIM sources include one or more internal PIM sources and/or one or more external PIM sources.
A method (1300) for estimating a time delay for a Passive Intermodulation (PIM) with multi-PIM sources of a radio equipment, comprising:

- determining (S1301) a first basis function based on a first input signal of the radio equipment;

- estimating (S1302) one or more first time delays for the first input signal based on the first basis function;

- determining (S1303) a second basis function based on the first input signal and a second input signal of the radio equipment; and

- estimating (S1304) one or more second time delays for the second input signal by referring to the one or more first time delays.
The method (1300) according to claim 16, further comprising:

- determining (S1305) a third basis function based on the first input signal, the second input signal, and a third input signal of the radio equipment; and

- estimating (S1306) one or more third time delays for the third input signal by referring to the one or more first time delays and the one or more second time delays.
The method (1300) according to claim 16, wherein estimating (S1302) one or more first time delays for the first input signal further comprising:

- determining (S1201) a first time delay window including a first plurality of time delay values;

- scanning (S1205) the first time delay window, by performing adaptive algorithm repeatedly based on the first basis function for each of the first plurality of time delay values, to obtain a first set of adaptive parameters; and

- estimating (S1208) one or more first time delays according to the first set of adaptive parameters.
The method (1300) according to claim 18, wherein estimating (S1304) one or more second time delays for the second input signal further comprising:

- determining (S1201) a second time delay window for the first input signal based on one of the one or more first time delays, the second time delay window is within the first time delay window, the second time delay window has a range around the one of the one or more first time delays;

- determining (S1201) a third time delay window for the second input signal;

- performing (S1205) a fraction scanning on the second and third time delay window, by performing adaptive algorithm repeatedly based on the second basis function for each of a plurality of second time delay values within the second time delay window and each of a plurality of third time delay values within the third time delay window, to obtain a second set of adaptive parameters; and

- estimating (S1208) one or more updated first time delays for the first input signal and one or more second time delays for the second input signal according to the second set of adaptive parameters.
The method (1300) according to claim 19, wherein a minimum value of the plurality of second time delay values within the second time delay window is equal to the one of the one or more first time delays plus a sampling cycle, and a maximum value of the plurality of second time delay values within the second time delay window is equal to the one of the one or more first time delays minus a sampling cycle.
The method (1300) according to claim 19 or claim 20, wherein estimating (S1306) one or more third time delays for the third input signal further comprising:

- determining (S1201) a fourth time delay window for the first input signal based on one of the one or more updated first time delays, the fourth time delay window is within the first time delay window, the fourth time delay window has a range around the one of the one or more updated first time delays;

- determining (S1201) a fifth time delay window for the second input signal based on one of the one or more second time delays, the fifth time delay window is within the third time delay window, the fifth time delay window has a range around the one of the one or more second time delays;

- determining (S1201) a sixth time delay window for the third input signal;

- performing (S1205) a fraction scanning on the fourth, fifth and sixth time delay window, by performing adaptive algorithm repeatedly based on the third basis function for each of a plurality of fourth time delay values within the fourth time delay window, each of a plurality of fifth time delay values within the fifth time delay window, and each of a plurality of sixth time delay values within the sixth time delay window, to obtain a third set of adaptive parameters; and

- estimating (S1208) one or more updated first time delays for the first input signal, one or more updated second time delays for the second input signal, and one or more third time delays according to the second set of adaptive parameters.
An apparatus (1400) for estimating a time delay for a Passive Intermodulation (PIM) with multi-PIM sources of a radio equipment, comprising:

- at least one processor (1401) ; and

- a non-transitory computer readable medium (1402) coupled to the at least one processor (1401) , the non-transitory computer readable medium (1402) contains instructions executable by the at least one processor (1401) , whereby the at least one processor (1401) is configured to perform the method (1200, 1300) according to any one of claims 1-21.
A computer readable medium (1402, 1502) comprising computer readable code, which when run on an apparatus (1400, 1500) , causes the apparatus (1400, 1500) to perform the method (1200, 1300) according to any one of claims 1-21.
A computer program product (1504) comprising computer readable code, which when run on an apparatus (1400, 1500) , causes the apparatus (1400, 1500) to perform the method (1200, 1300) according to any one of claims 1-21.