CN108566350B - Bluetooth receiving equalization method and device and Bluetooth equipment - Google Patents

Bluetooth receiving equalization method and device and Bluetooth equipment Download PDF

Info

Publication number
CN108566350B
CN108566350B CN201810213703.0A CN201810213703A CN108566350B CN 108566350 B CN108566350 B CN 108566350B CN 201810213703 A CN201810213703 A CN 201810213703A CN 108566350 B CN108566350 B CN 108566350B
Authority
CN
China
Prior art keywords
signal
demodulated
equalization
reference signal
initial
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201810213703.0A
Other languages
Chinese (zh)
Other versions
CN108566350A (en
Inventor
朱旭
符运升
展睿
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Espressif Systems Shanghai Co Ltd
Original Assignee
Espressif Systems Shanghai Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Espressif Systems Shanghai Co Ltd filed Critical Espressif Systems Shanghai Co Ltd
Priority to CN201810213703.0A priority Critical patent/CN108566350B/en
Publication of CN108566350A publication Critical patent/CN108566350A/en
Application granted granted Critical
Publication of CN108566350B publication Critical patent/CN108566350B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L25/03159—Arrangements for removing intersymbol interference operating in the frequency domain
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03878—Line equalisers; line build-out devices
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/0014—Carrier regulation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B5/00—Near-field transmission systems, e.g. inductive or capacitive transmission systems
    • H04B5/70—Near-field transmission systems, e.g. inductive or capacitive transmission systems specially adapted for specific purposes
    • H04B5/72—Near-field transmission systems, e.g. inductive or capacitive transmission systems specially adapted for specific purposes for local intradevice communication
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/0014—Carrier regulation
    • H04L2027/0024—Carrier regulation at the receiver end
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/0014—Carrier regulation
    • H04L2027/0024—Carrier regulation at the receiver end
    • H04L2027/0026—Correction of carrier offset
    • H04L2027/0038—Correction of carrier offset using an equaliser

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Power Engineering (AREA)
  • Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)

Abstract

The embodiment of the application discloses a Bluetooth receiving equalization method, a device and Bluetooth equipment, wherein the method comprises the following steps: sending an initial signal to an equalizer, wherein the equalizer performs equalization processing on the initial signal according to an equalization coefficient to obtain a signal to be demodulated; determining a reference signal according to the initial signal; inputting the reference signal and the signal to be demodulated into a preset training controller, and training the preset training controller through a preset algorithm to generate a parameter training model; the parameter training model is used for calculating loss functions of a reference signal and a signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions; and the preset training controller determines the equalization adjustment coefficient as the equalization coefficient of the equalizer. Through the technical scheme, the received Bluetooth signals can be optimized.

Description

Bluetooth receiving equalization method and device and Bluetooth equipment
Technical Field
The embodiment of the application relates to a bluetooth communication technology, and in particular relates to a bluetooth receiving equalization method and device and a bluetooth device.
Background
With the development of bluetooth technology, the new generation of bluetooth technology has the advantages of higher speed, larger broadcast capacity, longer transmission distance, etc. For example, bluetooth 5 has twice the transmission rate, four times the transmission distance, and eight times the broadcast capacity of bluetooth 4.x, compared to bluetooth 4.x, the previous version. Although the new generation of bluetooth technology has higher advantages, as the transmission distance of bluetooth communication increases, the interference during transmission increases, and although the transmission distance increases, the effect of the received bluetooth signal is not ideal, so that the receiving of the bluetooth signal needs to be optimized.
Disclosure of Invention
The embodiment of the application provides a Bluetooth receiving equalization method, a Bluetooth receiving equalization device and Bluetooth equipment, which can optimize received Bluetooth signals.
In a first aspect, an embodiment of the present application provides a bluetooth reception equalization method, including:
sending an initial signal to an equalizer, wherein the equalizer performs equalization processing on the initial signal according to an equalization coefficient to obtain a signal to be demodulated;
determining a reference signal according to the initial signal;
inputting the reference signal and the signal to be demodulated into a preset training controller, and training the preset training controller through a preset algorithm to generate a parameter training model; the parameter training model is used for calculating loss functions of a reference signal and a signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions;
and the preset training controller determines the equalization adjustment coefficient as the equalization coefficient of the equalizer.
Optionally, determining a reference signal according to the initial signal comprises:
determining a reference signal to be processed according to the initial signal;
modulating the reference signal to be processed through a reference modulator according to the frequency offset parameter to determine a reference signal;
correspondingly, the parameter training model is also used for calculating phase loss functions of the reference signal and the signal to be demodulated and outputting frequency offset adjustment parameters according to the phase loss functions;
correspondingly, the preset training controller is further configured to determine the frequency offset adjustment parameter as the frequency offset parameter of the reference modulator.
Optionally, before sending the initial signal to the equalizer, the method further includes:
sending the initial signal to an initial filter for filtering processing to obtain a filtered initial signal Rn：
Figure BDA0001597955730000021
Wherein the content of the first and second substances,nfor the initial signal before filtering, vector HLPIs the impulse response of the initial filter, and M is the length of the initial filter;
accordingly, the signal to be demodulated
Figure BDA0001597955730000022
Comprises the following steps:
Figure BDA0001597955730000023
wherein the vector C is an equalization coefficient of the equalizer.
Optionally, the training the preset training controller through a preset algorithm includes:
according to a reference signal InAnd a signal to be demodulated
Figure BDA0001597955730000024
Obtaining a loss function
Figure BDA0001597955730000025
Comprises the following steps:
Figure BDA0001597955730000026
determining a gradient from the loss function
Figure BDA0001597955730000027
Comprises the following steps:
Figure BDA0001597955730000031
determining an equalization adjustment coefficient C' according to the gradient as follows:
Figure BDA0001597955730000032
wherein, the vector C is the equalizing coefficient of the equalizer, the vector η is a first preset learning rate, and J is a set gradient accumulation number.
Optionally, the training the preset training controller through a preset algorithm further includes:
according to a reference signal InAnd a signal to be demodulated
Figure BDA0001597955730000033
Obtaining a phase difference Delta phinComprises the following steps:
Figure BDA0001597955730000034
determining a phase loss function L (Δ Φ) from the phase differencen+1,Φn) Comprises the following steps:
Figure BDA0001597955730000035
determining a phase gradient from the phase loss function
Figure BDA0001597955730000036
Comprises the following steps:
Figure BDA0001597955730000037
determining a frequency offset adjustment parameter omega' according to the phase gradient as follows:
Figure BDA0001597955730000038
wherein the content of the first and second substances,
Figure BDA0001597955730000039
to be the phase of the reference signal,
Figure BDA00015979557300000310
for the phase of the signal to be modulated, omega is the frequency offset parameter, vector etaωK is a set phase gradient accumulation number for the second predetermined learning rate.
Optionally, before the reference modulator performs modulation processing on the reference signal to be processed according to the frequency offset parameter, the method further includes:
filtering the reference signal to be processed by a reference filter to obtain a filterThe later reference signal x to be processedn：
xn＝GtDn；
Wherein D isnFor the reference signal to be processed, vector G is the coefficient of the reference filter;
accordingly, the reference signal InComprises the following steps:
Figure BDA0001597955730000041
where ω is a frequency offset parameter.
Optionally, the determining a reference signal to be processed according to the initial signal includes:
determining a reference signal to be processed according to the packet header data of the initial signal;
correspondingly, after the equalizer performs equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated, the method further includes:
signal to be demodulated
Figure BDA0001597955730000042
Frequency offset correction is carried out to obtain a corrected signal to be demodulated
Figure BDA0001597955730000043
Figure BDA0001597955730000044
Optionally, the equalizing, by the equalizer, the initial signal according to the equalization coefficient to obtain a signal to be demodulated includes:
the equalizer performs equalization processing on the transmission data of the initial signal according to the equalization coefficient to obtain a signal to be demodulated;
correspondingly, after the equalizer performs equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated, the method further includes:
demodulating the signal to be demodulated to obtain a demodulated output signal;
correspondingly, the determining the to-be-processed reference signal according to the initial signal comprises:
determining a reference signal to be processed according to the packet header data of the initial signal;
or
And determining a reference signal to be processed according to the demodulation output signal.
In a second aspect, an embodiment of the present application further provides an equalizing apparatus for bluetooth reception, including:
the equalizer is used for carrying out equalization processing on the initial signal according to the equalization coefficient so as to obtain a signal to be demodulated;
a reference generator for determining a reference signal from the initial signal;
the preset training controller is used for training the preset training controller through a preset algorithm according to the input reference signal and the signal to be demodulated so as to generate a parameter training model; the parameter training model is used for calculating loss functions of a reference signal and a signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions;
the preset training controller is further configured to determine the equalization adjustment coefficient as an equalization coefficient of the equalizer.
In a third aspect, an embodiment of the present application further provides a bluetooth device, including the above bluetooth reception equalizing apparatus.
The embodiment of the application discloses a bluetooth reception equalization scheme, including: sending an initial signal to an equalizer, wherein the equalizer performs equalization processing on the initial signal according to an equalization coefficient to obtain a signal to be demodulated; determining a reference signal according to the initial signal; inputting the reference signal and the signal to be demodulated into a preset training controller, and training the preset training controller through a preset algorithm to generate a parameter training model; the parameter training model is used for calculating loss functions of a reference signal and a signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions; and the preset training controller determines the equalization adjustment coefficient as the equalization coefficient of the equalizer. Through the technical scheme, the received Bluetooth signals can be optimized.
Drawings
Fig. 1 is a flowchart of a bluetooth reception equalization method according to an embodiment of the present application;
fig. 2 is a flowchart of another bluetooth reception equalization method according to an embodiment of the present application;
fig. 3 is a block diagram of a bluetooth reception equalizing apparatus according to an embodiment of the present application;
fig. 4 is a block diagram of another bluetooth reception equalizing apparatus according to an embodiment of the present application.
Detailed Description
The present application will be described in further detail with reference to the following drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the application and are not limiting of the application. It should be further noted that, for the convenience of description, only some of the structures related to the present application are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a flowchart of a bluetooth reception equalization method according to an embodiment of the present application, where the method includes:
s110, sending the initial signal to an equalizer, and equalizing the initial signal by the equalizer according to the equalization coefficient to obtain a signal to be demodulated.
Wherein the initial signal may be a received bluetooth signal; the Bluetooth receiving and balancing device establishes connection with an external Bluetooth device, and the external Bluetooth device sends an initial signal to the Bluetooth receiving and balancing device. And the equalizer is used for carrying out equalization processing on the initial signal according to the equalization coefficient so as to obtain a signal to be demodulated.
Optionally, before sending the initial signal to the equalizer, the method further includes: sending the initial signal to an initial filter for filtering processing to obtain a filtered initial signal Rn：
Figure BDA0001597955730000061
Wherein, the initial filter may be a low-pass filter, and specifically may be an FIR (Finite Impulse Response) filter;nis the initial signal before filtering, specifically the nth initial signal received, rn＝[rn rn+1 rn+2 … rn+M-1]tVector HLPIs the impulse response of the initial filter and M is the length of the initial filter.
Correspondingly, the filtered initial signal is sent to an equalizer, and the equalizer performs equalization processing on the filtered initial signal according to an equalization coefficient to obtain a signal to be demodulated
Figure BDA0001597955730000071
Comprises the following steps:
Figure BDA0001597955730000072
wherein, the vector C is the equalizing coefficient of the equalizer; the equalization coefficient C may be a preset equalization coefficient of the equalizer.
And S111, determining a reference signal according to the initial signal.
The initial signal may include header data and transmission data. The first time when the external bluetooth device and the bluetooth reception equalizing device establish connection, the packet header data may be obtained, where the packet header data is a starting part of the initial signal and may be determined by a broadcast issued by the external bluetooth device. Illustratively, if the external bluetooth device is a legacy bluetooth device, the packet header data includes an access code (access code). If the external bluetooth device is a bluetooth low energy device, the packet header data includes a random access preamble (preamble) and an access address (access address).
The transmission data comprises data which are actually required to be transmitted by the Bluetooth connection, correspondingly, the equalizer performs equalization processing on the transmission data of the initial signal according to an equalization coefficient to obtain a signal to be demodulated, and performs demodulation processing on the signal to be demodulated through the demodulator to obtain a demodulation output signal. The demodulator may include a GFSK (gaussian frequency Shift Keying) demodulator, and accordingly, the transmission data of the received initial signal is a GFSK signal. The demodulator may also include a DPSK (Differential Phase Shift Keying) demodulator.
Alternatively, the reference signal may be determined from header data of the initial signal, or from a demodulated output signal.
Determining a reference signal according to the header data of the initial signal, namely using the header data of the initial signal as the reference signal; and determining a reference signal according to the demodulation output signal, namely using the demodulation output signal obtained by the demodulator as the reference signal.
It should be noted that the execution sequence of operation S111 is not limited to that shown in fig. 1, and operation S111 may be executed before operation S110.
S112, inputting the reference signal and the signal to be demodulated into a preset training controller, and training the preset training controller through a preset algorithm to generate a parameter training model; the parameter training model is used for calculating loss functions of the reference signal and the signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions.
Optionally, the preset algorithm may adopt an iterative method, and in the embodiment of the present application, a gradient descent method is adopted. Wherein training the preset training controller through a preset algorithm comprises:
according to a reference signal InAnd a signal to be demodulated
Figure BDA0001597955730000081
Obtaining a loss function
Figure BDA0001597955730000082
Comprises the following steps:
Figure BDA0001597955730000083
determining a gradient from the loss function
Figure BDA0001597955730000084
Comprises the following steps:
Figure BDA0001597955730000085
determining an equalization adjustment coefficient C' according to the gradient as follows:
Figure BDA0001597955730000086
Figure BDA0001597955730000087
wherein, the vector C is the equalizing coefficient of the equalizer, the vector eta is a first preset learning rate, and J is a set gradient accumulation number; wherein the content of the first and second substances,
Figure BDA0001597955730000088
expressed as a vector RnEach element in the (A) is a complex conjugate,
Figure BDA0001597955730000089
expressed as a vector RiEach element in the (A) is a complex conjugate,
Figure BDA00015979557300000810
representing a Hadamard product (Hadamard product),
Figure BDA00015979557300000811
represents: vector η and vector
Figure BDA00015979557300000812
The hadamard product of (a).
C＝[C1C2…n]tEach coefficient C of the vector CnAll using different learning rates etanA first preset learning rate composed of different learning rates: vector η ═ η1η2…N]tThe first predetermined learning rate η may be a window function, specifically, a gaussian window function, a rectangular window function, or other window functions, and N is greater than or equal to 1. The setting of the gradient accumulation number can be preset by a user, and J is larger than or equal to 1. And accumulating the calculated gradients until the gradient reaches the set gradient accumulation number, and updating the equalization coefficient.
Since the bluetooth signal transmitted over a long distance is easily affected by the multipath fading, by setting the first preset learning rate, it can be assumed that the probability distribution of the multipath fading conforms to the set window function, and the influence of the equalization coefficient of the equalizer on the equalization result should also conform to the window function. The first predetermined learning rate of the window function shape therefore represents: the larger the influence of the equalization adjustment coefficient is, the larger the learning rate corresponding to the equalization adjustment coefficient is, so that the convergence of the equalization adjustment coefficient can be accelerated. The equalizer can effectively equalize the Bluetooth signals affected by the multipath attenuation according to the trained and updated equalization coefficients.
And S113, the preset training controller determines the equalization adjustment coefficient as the equalization coefficient of the equalizer.
That is, the equalization coefficient C of the equalizer is updated to C'. According to the embodiment of the application, the reference signal is determined according to the initial signal, and the equalization coefficient is trained and updated by adopting the preset algorithm according to the reference signal and the signal to be demodulated, so that the equalizer can effectively equalize the received Bluetooth signal according to the trained equalization coefficient.
As another implementation manner of this embodiment, on the basis of the foregoing embodiment, a bluetooth reception equalization method includes the following steps:
and S120, sending the initial signal to an equalizer, and equalizing the initial signal by the equalizer according to the equalization coefficient to obtain a signal to be demodulated.
And S121, determining a reference signal to be processed according to the initial signal.
Correspondingly, determining a reference signal according to the packet header data of the initial signal, or determining a reference signal according to the demodulated output signal, is: and determining a reference signal to be processed according to the packet header data of the initial signal, or determining the reference signal to be processed according to the demodulation output signal.
Since the reference signal determined according to the initial signal may not necessarily be recognized by the preset training controller, the reference signal to be processed needs to be determined according to the initial signal and modulated so that the preset training controller may recognize the reference signal.
And S122, modulating the reference signal to be processed through the reference modulator according to the frequency offset parameter to determine the reference signal.
Optionally, before performing modulation processing on the reference signal to be processed, the method further includes: filtering the reference signal to be processed by a reference filter to obtain a filtered reference signal x to be processedn：
xn＝GtDn；
Wherein the reference filter may be a gaussian filter, in particular a gaussian windowed FIR filter, DnFor reference signals to be processed, Dn＝[Dn Dn+1 Dn+2 … Dn+O-1]tVector G is the coefficients of the reference filter and O is the length of the reference filter.
Correspondingly, the reference signal to be processed is modulated by the reference modulator according to the frequency offset parameter to determine the reference signal InComprises the following steps:
Figure BDA0001597955730000101
wherein, the reference modulator may be an FM modulator, ω is a frequency offset parameter, and the frequency offset parameter may be preset. The reference signal determined by the reference modulator may be implemented by means of a table look-up or by means of CORDIC (Coordinate Rotation Digital Computer) operations.
S123, inputting the reference signal and the signal to be demodulated into a preset training controller, and training the preset training controller through a preset algorithm to generate a parameter training model; the parameter training model is further used for calculating phase loss functions of the reference signal and the signal to be demodulated and outputting frequency offset adjustment parameters according to the phase loss functions.
Optionally, the training the preset training controller through a preset algorithm further includes:
according to a reference signal InAnd a signal to be demodulated
Figure BDA0001597955730000111
Obtaining a phase difference Delta phinComprises the following steps:
Figure BDA0001597955730000112
determining a phase loss function L (Δ Φ) from the phase differencen-1,Φn) Comprises the following steps:
Figure BDA0001597955730000113
determining a phase gradient from the phase loss function
Figure BDA0001597955730000114
Comprises the following steps:
Figure BDA0001597955730000115
determining a frequency offset adjustment parameter omega' according to the phase gradient as follows:
Figure BDA0001597955730000116
Figure BDA0001597955730000117
wherein the content of the first and second substances,
Figure BDA0001597955730000118
to be the phase of the reference signal,
Figure BDA0001597955730000119
the phase of the signal to be modulated, omega being the frequency offset parameter, vector etaωK is a set phase gradient accumulation number for the second predetermined learning rate.
And S124, the preset training controller is used for determining the frequency offset adjustment parameter as the frequency offset parameter of the reference modulator.
That is, the frequency offset parameter ω is updated to the frequency offset adjustment parameter ω'.
According to the embodiment of the application, the phase loss function of the reference signal and the signal to be demodulated is calculated, so that the fluctuation of the phase difference between the reference signal and the signal to be demodulated in time can be minimized, the frequency deviation adjusting parameter obtained through training is determined as the frequency deviation parameter, and the phase difference between the reference signal and the signal to be demodulated obtained through processing by the reference modulator tends to be constant. By training and updating the frequency deviation parameter and the equalization parameter at the same time, the equalizer can further reach a normal and stable working state, and the equalization of the received Bluetooth signal can be further improved.
Optionally, as another implementation manner of the embodiment of the present application, on the basis of the foregoing embodiment, the determining a reference signal to be processed according to the initial signal may be further implemented by:
and determining a reference signal to be processed according to the packet header data of the initial signal.
That is, the packet header data may be obtained at the beginning of the connection between the bluetooth receiving equalization apparatus and the external bluetooth apparatus, and then a to-be-processed reference signal may be determined according to the packet header data of the initial signal, and the to-be-processed reference signal may be input into the memory. The pre-set training controller may retrieve the reference signal to be processed directly from the memory.
Correspondingly, after the equalizer performs equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated, the method further includes:
signal to be demodulated
Figure BDA0001597955730000121
Frequency offset correction is carried out to obtain a corrected signal to be demodulated
Figure BDA0001597955730000122
Figure BDA0001597955730000123
Since the frequency offset parameter of the reference modulator cannot be known in advance, the frequency offset correction needs to be performed on the signal to be demodulated. Signal to be demodulated
Figure BDA0001597955730000124
This may be done by means of a table lookup or by CORDIC operations.
The technical scheme of the embodiment of the application can solve the influence of multipath attenuation on Bluetooth signal receiving, and is based on the assumption that the same signal reaches a receiving end from a transmitting end through different paths, the action of each path on the signal is linear, and the signal reaching the receiving end through different paths is additive. The amplitude of the received signal can thus be equalized in the time domain using an equalizer based on an FIR filter. The embodiment of the application performs equalization integration on signals reaching a receiving end through different paths, trains and updates equalization parameters and frequency offset parameters by using more useful information as reference signals, can reduce the influence of multipath attenuation on signal receiving, improves the signal-to-noise ratio of signal receiving, and improves the sensitivity.
Example two
Fig. 3 is a block connection diagram of a bluetooth reception equalizing apparatus according to a second embodiment of the present application, and as shown in fig. 3, the apparatus includes:
the equalizer 201 is configured to perform equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated.
Wherein the initial signal may be a received bluetooth signal; the Bluetooth receiving and balancing device establishes connection with an external Bluetooth device, and the external Bluetooth device sends an initial signal to the Bluetooth receiving and balancing device. And the equalizer is used for carrying out equalization processing on the initial signal according to the equalization coefficient so as to obtain a signal to be demodulated.
Optionally, an initial filter 202 is further included, for performing a filtering process on the initial signal before the equalizer 201 performs an equalization process on the initial signal to obtain a filtered initial signal Rn：
Figure BDA0001597955730000131
Wherein the initial filter 202 may be a low pass filter,nis the initial signal before filtering, specifically the nth initial signal received, rn＝[rn rn+1 rn+2 … n+M-1]tVector HLPIs the impulse response of the initial filter and M is the length of the initial filter.
Correspondingly, the equalizer 201 performs an equalization process on the filtered initial signal according to the equalization coefficient to obtain a signal to be demodulated
Figure BDA0001597955730000132
Comprises the following steps:
Figure BDA0001597955730000133
wherein, the vector C is the equalizing coefficient of the equalizer; the equalization coefficient C may be a preset equalization coefficient of the equalizer.
A reference generator 203 for determining a reference signal from the initial signal.
Wherein the initial signal may include header data and transmission data. The first time when the external bluetooth device and the bluetooth reception equalizing device establish connection, the packet header data may be obtained, where the packet header data is a starting part of the initial signal and may be determined by a broadcast issued by the external bluetooth device. Illustratively, if the external bluetooth device is a legacy bluetooth device, the packet header data includes an access code (access code). If the external bluetooth device is a bluetooth low energy device, the packet header data includes a random access preamble (preamble) and an access address (access address).
The transmission data includes data actually required to be transmitted by the bluetooth connection, and accordingly, the equalizer performs equalization processing on the transmission data of the initial signal according to the equalization coefficient to obtain a signal to be demodulated. The demodulator may include a GFSK (gaussian frequency Shift Keying) demodulator, and accordingly, the transmission data of the received initial signal is a GFSK signal. The demodulator may also include a DPSK (Differential Phase Shift Keying) demodulator.
Optionally, the reference generator 203 is configured to determine a reference signal according to packet header data of the initial signal, or determine a reference signal according to a demodulation output signal.
The preset training controller 204 is configured to train the preset training controller through a preset algorithm according to the input reference signal and the signal to be demodulated to generate a parameter training model; the parameter training model is used for calculating loss functions of the reference signal and the signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions.
Optionally, the preset algorithm may adopt an iterative method, and in the embodiment of the present application, a gradient descent method is adopted. Wherein training the preset training controller through a preset algorithm comprises:
according to a reference signal InAnd a signal to be demodulated
Figure BDA0001597955730000141
Obtaining a loss function
Figure BDA0001597955730000142
Comprises the following steps:
Figure BDA0001597955730000143
determining a gradient from the loss function
Figure BDA0001597955730000144
Comprises the following steps:
Figure BDA0001597955730000145
determining an equalization adjustment coefficient C' according to the gradient as follows:
Figure BDA0001597955730000151
Figure BDA0001597955730000152
wherein, the vector C is the equalizing coefficient of the equalizer, the vector eta is the first preset learning rate, J is the set gradient accumulation quantity,
Figure BDA0001597955730000153
representing a Hadamard product (Hadamard product),
Figure BDA0001597955730000154
represents: vector η and vector
Figure BDA0001597955730000155
The hadamard product of (a).
C＝[C1C2…Cn]tEach coefficient C of the vector CnAll using different learning rates etanA first preset learning rate composed of different learning rates: vector η ═ η1η2…ηN]tThe first predetermined learning rate η may be a window function, specifically, a gaussian window function, a rectangular window function, or other window functions, and N is greater than or equal to 1. The setting of the gradient accumulation number can be preset by a user, and J is larger than or equal to 1. And accumulating the calculated gradients until the gradient reaches the set gradient accumulation number, and updating the equalization coefficient.
Since the bluetooth signal transmitted over a long distance is easily affected by the multipath fading, by setting the first preset learning rate, it can be assumed that the probability distribution of the multipath fading conforms to the set window function, and the influence of the equalization coefficient of the equalizer on the equalization result should also conform to the window function. The first predetermined learning rate of the window function shape therefore represents: the larger the influence of the equalization adjustment coefficient is, the larger the learning rate corresponding to the equalization adjustment coefficient is, so that the convergence of the equalization adjustment coefficient can be accelerated. The equalizer can effectively equalize the Bluetooth signals affected by the multipath attenuation according to the trained and updated equalization coefficients.
The preset training controller 204 is further configured to determine the equalization adjustment coefficient as an equalization coefficient of the equalizer. That is, the equalization coefficient C of the equalizer is updated to C'.
According to the embodiment of the application, the reference signal is determined according to the initial signal, and the equalization coefficient is trained and updated by adopting the preset algorithm according to the reference signal and the signal to be demodulated, so that the equalizer can effectively equalize the received Bluetooth signal according to the trained equalization coefficient.
As another implementation manner of this embodiment, on the basis of the above embodiment, a bluetooth reception equalizing apparatus includes:
the equalizer 201 is configured to perform equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated.
A reference generator 203 for determining a reference signal from the initial signal.
Optionally, the reference generator 203 is configured to determine a reference signal to be processed according to packet header data of the initial signal, or determine a reference signal to be processed according to a demodulation output signal.
Since the reference signal determined according to the initial signal may not necessarily be recognized by the pre-set training controller 204, the reference signal to be processed needs to be determined according to the initial signal and modulated so that the pre-set training controller can recognize the reference signal.
And the reference modulator 210 is configured to perform modulation processing on the reference signal to be processed according to the frequency offset parameter to determine a reference signal.
Optionally, the apparatus further includes a reference filter 211, configured to perform filtering processing on the reference signal to be processed before performing modulation processing on the reference signal to be processed, so as to obtain a filtered reference signal x to be processedn：
xn＝GtDn；
Wherein the reference filter may be a gaussian filter, in particular a gaussian windowed FIR filter, DnFor reference signals to be processed, Dn＝[Dn Dn+1 Dn+2 … Dn+o-1]tVector G is the coefficients of the reference filter and O is the length of the reference filter.
Correspondingly, the reference signal to be processed is modulated by the reference modulator according to the frequency offset parameter to determine the reference signal InComprises the following steps:
Figure BDA0001597955730000161
wherein, the reference modulator may be an FM modulator, ω is a frequency offset parameter, and the frequency offset parameter may be preset. The reference signal determined by the reference modulator may be implemented by means of a table look-up or by means of CORDIC (Coordinate Rotation Digital Computer) operations.
Presetting a training controller 204, and training the preset training controller through a preset algorithm to generate a parameter training model; the parameter training model is further used for calculating phase loss functions of the reference signal and the signal to be demodulated and outputting frequency offset adjustment parameters according to the phase loss functions.
Optionally, the training the preset training controller through a preset algorithm further includes:
according to a reference signal InAnd a signal to be demodulated
Figure BDA0001597955730000171
Obtaining a phase difference Delta phinComprises the following steps:
Figure BDA0001597955730000172
determining a phase loss function L (Δ Φ) from the phase differencen-1,Φn) Comprises the following steps:
Figure BDA0001597955730000173
determining a phase gradient from the phase loss function
Figure BDA0001597955730000174
Comprises the following steps:
Figure BDA0001597955730000175
determining a frequency offset adjustment parameter omega' according to the phase gradient as follows:
Figure BDA0001597955730000176
Figure BDA0001597955730000177
wherein the content of the first and second substances,
Figure BDA0001597955730000178
to be the phase of the reference signal,
Figure BDA0001597955730000179
for the phase of the signal to be modulated, omega is the frequency offset parameter, vector etaωK is a set phase gradient accumulation number for the second predetermined learning rate.
The preset training controller 204 is further configured to determine the frequency offset adjustment parameter as the frequency offset parameter of the reference modulator.
That is, the frequency offset parameter ω is updated to the frequency offset adjustment parameter ω'.
According to the embodiment of the application, the phase loss function of the reference signal and the signal to be demodulated is calculated, so that the fluctuation of the phase difference between the reference signal and the signal to be demodulated in time can be minimized, the frequency deviation adjusting parameter obtained through training is determined as the frequency deviation parameter, and the phase difference between the reference signal and the signal to be demodulated obtained through processing by the reference modulator tends to be constant.
Optionally, as another implementation manner of the embodiment of the present application, on the basis of the above embodiment, as shown in fig. 4, the reference generator 203 is configured to determine a reference signal to be processed according to packet header data of the initial signal.
That is, the packet header data may be obtained at the beginning of the connection between the bluetooth receiving equalizing apparatus and the external bluetooth apparatus, and then a to-be-processed reference signal may be determined according to the packet header data of the initial signal, and may be input into the memory 213. The pre-set training controller 204 may retrieve the reference signal to be processed directly from the memory 213.
Correspondingly, after the equalizer 201 performs equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated, the method further includes:
a frequency offset correction module 212 for correcting the frequency offset of the signal to be demodulated
Figure BDA0001597955730000181
Frequency offset correction is carried out to obtain a corrected signal to be demodulated
Figure BDA0001597955730000182
Figure BDA0001597955730000183
Since the frequency offset parameter of the reference modulator 210 cannot be known in advance, the frequency offset correction needs to be performed on the signal to be demodulated. Signal to be demodulated
Figure BDA0001597955730000184
This may be done by means of a table lookup or by CORDIC operations.
The technical scheme of the embodiment of the application can solve the influence of multipath attenuation on Bluetooth signal receiving, and is based on the assumption that the same signal reaches a receiving end from a transmitting end through different paths, the action of each path on the signal is linear, and the signal reaching the receiving end through different paths is additive. The amplitude of the received signal can thus be equalized in the time domain using an equalizer based on an FIR filter. The embodiment of the application performs equalization integration on signals reaching a receiving end through different paths, trains and updates equalization parameters and frequency offset parameters by using more useful information as reference signals, can reduce the influence of multipath attenuation on signal receiving, improves the signal-to-noise ratio of signal receiving, and improves the sensitivity.
EXAMPLE III
An embodiment of the present application further provides a bluetooth device, where the bluetooth device includes the bluetooth reception equalizing apparatus according to any of the above embodiments. The Bluetooth device can be a Bluetooth earphone, a Bluetooth sound box, a Bluetooth mobile phone or a Bluetooth wearable device and the like.
Reference may be made to the above description for specific embodiments, which are not repeated herein.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present application and the technical principles employed. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, although the present application has been described in more detail with reference to the above embodiments, the present application is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present application, and the scope of the present application is determined by the scope of the appended claims.

Claims (10)

1. A method for bluetooth reception equalization, comprising:
sending an initial signal to an equalizer, wherein the equalizer performs equalization processing on the initial signal according to an equalization coefficient to obtain a signal to be demodulated;
determining a reference signal according to the initial signal;
inputting the reference signal and the signal to be demodulated into a preset training controller, and training the preset training controller through a preset algorithm to generate a parameter training model; the parameter training model is used for calculating loss functions of a reference signal and a signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions;
and the preset training controller determines the equalization adjustment coefficient as the equalization coefficient of the equalizer.
2. The method of claim 1, wherein determining a reference signal from the initial signal comprises:
determining a reference signal to be processed according to the initial signal;
modulating the reference signal to be processed through a reference modulator according to the frequency offset parameter to determine a reference signal;
correspondingly, the parameter training model is also used for calculating phase loss functions of the reference signal and the signal to be demodulated and outputting frequency offset adjustment parameters according to the phase loss functions;
correspondingly, the preset training controller is further configured to determine the frequency offset adjustment parameter as the frequency offset parameter of the reference modulator.
3. The method of claim 1, wherein prior to sending the initial signal to the equalizer, further comprising:
sending the initial signal to an initial filter for filtering processing to obtain a filtered initial signal Rn：
Figure FDA0002824987270000011
Wherein r isnFor the initial signal before filtering, vector HLPIs the impulse response of the initial filter, M is the length of the initial filter,
Figure FDA0002824987270000021
is HLPTransposing;
accordingly, the signal to be demodulated
Figure FDA0002824987270000022
Comprises the following steps:
Figure FDA0002824987270000023
wherein the vector C is the equalization coefficient of the equalizer, CtIs the transpose of C.
4. The method of claim 1, wherein the training the pre-set training controller through a pre-set algorithm comprises:
according to a reference signal InAnd a signal to be demodulated
Figure FDA0002824987270000024
Obtaining a loss function
Figure FDA0002824987270000025
Comprises the following steps:
Figure FDA0002824987270000026
determining a gradient from the loss function
Figure FDA0002824987270000027
Comprises the following steps:
Figure FDA0002824987270000028
determining an equalization adjustment coefficient C' according to the gradient as follows:
Figure FDA0002824987270000029
wherein, the vector C is the equalizing coefficient of the equalizer, the vector eta is the first preset learning rate, J is the set gradient accumulation quantity,
Figure FDA00028249872700000210
expressed as a vector RnEach element in (a) is complex conjugated.
5. The method of claim 2, wherein said training the pre-set training controller through a pre-set algorithm further comprises:
according to a reference signal InAnd a signal to be demodulated
Figure FDA00028249872700000211
Obtaining a phase difference Delta phinComprises the following steps:
Figure FDA00028249872700000212
determining a phase loss function L (Δ Φ) from the phase differencen-1，ΔΦn) Comprises the following steps:
Figure FDA00028249872700000213
determining a phase gradient from the phase loss function
Figure FDA00028249872700000214
Comprises the following steps:
Figure FDA00028249872700000215
determining a frequency offset adjustment parameter omega' according to the phase gradient as follows:
Figure FDA0002824987270000031
wherein the content of the first and second substances,
Figure FDA0002824987270000032
to be the phase of the reference signal,
Figure FDA0002824987270000033
for the phase of the signal to be modulated, omega is the frequency offset parameter, vector etaωK is a set phase gradient accumulation number for the second predetermined learning rate.
6. The method of claim 2, wherein before the modulation processing of the reference signal to be processed by the reference modulator according to the frequency offset parameter, the method further comprises:
filtering the reference signal to be processed by a reference filter to obtain a filtered reference signal x to be processedn：
xn＝GtDn；
Wherein D isnFor the reference signal to be processed, the vector G is the coefficients of a reference filter, GtIs the transpose of G;
accordingly, the reference signal InComprises the following steps:
Figure FDA0002824987270000034
where ω is a frequency offset parameter.
7. The method according to claim 5 or 6, wherein the determining a reference signal to be processed from the initial signal comprises:
determining a reference signal to be processed according to the packet header data of the initial signal;
correspondingly, after the equalizer performs equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated, the method further includes:
signal to be demodulated
Figure FDA0002824987270000035
Frequency offset correction is carried out to obtain a corrected signal to be demodulated
Figure FDA0002824987270000036
Figure FDA0002824987270000037
8. The method according to any one of claims 2 to 6, wherein the equalizer equalizing the initial signal according to the equalization coefficient to obtain the signal to be demodulated comprises:
the equalizer performs equalization processing on the transmission data of the initial signal according to the equalization coefficient to obtain a signal to be demodulated;
correspondingly, after the equalizer performs equalization processing on the initial signal according to the equalization coefficient to obtain a signal to be demodulated, the method further includes:
demodulating the signal to be demodulated to obtain a demodulated output signal;
correspondingly, the determining the to-be-processed reference signal according to the initial signal comprises:
determining a reference signal to be processed according to the packet header data of the initial signal;
or
And determining a reference signal to be processed according to the demodulation output signal.
9. An equalizing apparatus for bluetooth reception, comprising:
the equalizer is used for carrying out equalization processing on the initial signal according to the equalization coefficient so as to obtain a signal to be demodulated;
a reference generator for determining a reference signal from the initial signal;
the preset training controller is used for training the preset training controller through a preset algorithm according to the input reference signal and the signal to be demodulated so as to generate a parameter training model; the parameter training model is used for calculating loss functions of a reference signal and a signal to be demodulated and outputting an equalization adjustment coefficient according to the loss functions;
the preset training controller is further configured to determine the equalization adjustment coefficient as an equalization coefficient of the equalizer.
10. A bluetooth apparatus, characterized by comprising the bluetooth reception equalizing device according to claim 9.
CN201810213703.0A 2018-03-15 2018-03-15 Bluetooth receiving equalization method and device and Bluetooth equipment Active CN108566350B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810213703.0A CN108566350B (en) 2018-03-15 2018-03-15 Bluetooth receiving equalization method and device and Bluetooth equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810213703.0A CN108566350B (en) 2018-03-15 2018-03-15 Bluetooth receiving equalization method and device and Bluetooth equipment

Publications (2)

Publication Number Publication Date
CN108566350A CN108566350A (en) 2018-09-21
CN108566350B true CN108566350B (en) 2021-02-09

Family

ID=63532674

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810213703.0A Active CN108566350B (en) 2018-03-15 2018-03-15 Bluetooth receiving equalization method and device and Bluetooth equipment

Country Status (1)

Country Link
CN (1) CN108566350B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103401821A (en) * 2013-07-22 2013-11-20 苏州英菲泰尔电子科技有限公司 Adaptive equalizer of Bluetooth 4.0 low-power receiver and implementation method of adaptive equalizer
CN104104627A (en) * 2014-08-01 2014-10-15 王红星 Parallel decision feedback balance method and device based on initial parameter passing
CN104143995A (en) * 2013-05-08 2014-11-12 博通集成电路（上海）有限公司 Self-adaptive Bluetooth receiver and method
CN106507274A (en) * 2016-10-13 2017-03-15 珠海煌荣集成电路科技有限公司 A kind of wireless control method of parametric equalizer

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104143995A (en) * 2013-05-08 2014-11-12 博通集成电路（上海）有限公司 Self-adaptive Bluetooth receiver and method
CN103401821A (en) * 2013-07-22 2013-11-20 苏州英菲泰尔电子科技有限公司 Adaptive equalizer of Bluetooth 4.0 low-power receiver and implementation method of adaptive equalizer
CN104104627A (en) * 2014-08-01 2014-10-15 王红星 Parallel decision feedback balance method and device based on initial parameter passing
CN106507274A (en) * 2016-10-13 2017-03-15 珠海煌荣集成电路科技有限公司 A kind of wireless control method of parametric equalizer

Also Published As

Publication number Publication date
CN108566350A (en) 2018-09-21

Similar Documents

Publication Publication Date Title
US20210152402A1 (en) High-speed signaling systems and methods with adaptable, continuous-time equalization
TWI826357B (en) Apparatus and method for iq mismatch calibration and compensation
JP2715662B2 (en) Method and apparatus for diversity reception of time division signals
CN104780037B (en) A kind of clock recovery method, apparatus and system
US6205170B1 (en) Transmission/reception unit with bidirectional equalization
JPH0257373B2 (en)
JP2005538578A (en) Adaptive pre-equalization method and apparatus
CN1394421A (en) Method for receiving signal and receiver
US8688044B2 (en) Method and system for selective equalization enablement based on modulation type
CN103988474B (en) For the method and apparatus of the cable loss equilibrium launch channel in C8PSK HART systems
US20030231725A1 (en) Methods and apparatus for spectral filtering channel estimates
US5295157A (en) Equalization method utilizing a delay circuit
JP3182763B2 (en) Apparatus and method for adjusting signal point, equalizer gain, etc.
IE50470B1 (en) A phase noise correction circuit for a data transmission system
JP5361827B2 (en) Transmitter and transmission method
JP2785858B2 (en) Receiving method using high-speed adaptive filter
CN108566350B (en) Bluetooth receiving equalization method and device and Bluetooth equipment
US7042962B2 (en) LMS-based channel estimation
JPH0228923B2 (en)
WO2001069877A1 (en) Transmission gain regulating method and radio device
JP2002280941A (en) Adaptive equalizer, receiver and tap coefficient calculation method
US7978758B1 (en) Efficient blind equalization for quadrature amplitude modulation
JPH09275426A (en) Demodulator
CN1953441B (en) Device and method for adaptive correction of orthogonal unbalance
JPH05235792A (en) Adaptive equalizer

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
CB02 Change of applicant information
CB02 Change of applicant information

Address after: Room 204, Building 2, 690 Bibo Road, Pudong New Area, Shanghai, 201203

Applicant after: Lexin Information Technology (Shanghai) Co., Ltd.

Address before: Room 204, Building 2, 690 Bibo Road, Pudong New Area, Shanghai, 201203

Applicant before: Yue Xin information technology (Shanghai) Co., Ltd.

GR01 Patent grant
GR01 Patent grant