CN106603217B - Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester - Google Patents

Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester Download PDF

Info

Publication number
CN106603217B
CN106603217B CN201710035632.5A CN201710035632A CN106603217B CN 106603217 B CN106603217 B CN 106603217B CN 201710035632 A CN201710035632 A CN 201710035632A CN 106603217 B CN106603217 B CN 106603217B
Authority
CN
China
Prior art keywords
interpolation
sampling
signal
pos
interp
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
CN201710035632.5A
Other languages
Chinese (zh)
Other versions
CN106603217A (en
Inventor
吴帅
周英
吴建兵
刘海溶
蒋芜
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.)
SHENZHEN JIZHI HUIYI TECHNOLOGY Co Ltd
Original Assignee
SHENZHEN JIZHI HUIYI TECHNOLOGY 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 SHENZHEN JIZHI HUIYI TECHNOLOGY Co Ltd filed Critical SHENZHEN JIZHI HUIYI TECHNOLOGY Co Ltd
Priority to CN201710035632.5A priority Critical patent/CN106603217B/en
Publication of CN106603217A publication Critical patent/CN106603217A/en
Application granted granted Critical
Publication of CN106603217B publication Critical patent/CN106603217B/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
    • H04L7/00—Arrangements for synchronising receiver with transmitter
    • H04L7/0054—Detection of the synchronisation error by features other than the received signal transition
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L7/00—Arrangements for synchronising receiver with transmitter
    • H04L7/0016—Arrangements for synchronising receiver with transmitter correction of synchronization errors

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Digital Transmission Methods That Use Modulated Carrier Waves (AREA)

Abstract

The invention provides a method for inhibiting sampling frequency offset of a Bluetooth signal of a wireless comprehensive tester, which comprises the following steps: step S1, down-sampling the input signal; step S2, performing matched filtering on the signal after the down-sampling processing; step S3, setting the initial value of the interpolation control parameter; step S4, traverse all the input symbolic data, and find the interpolation control parameter of the optimal sampling point. The IQ data to be demodulated, which is input by a Bluetooth signal of the wireless comprehensive tester through DPSK modulation, is subjected to down-sampling processing, so that the operation complexity of matched filtering is reduced; and then, carrying out sampling frequency offset suppression on data output by matched filtering through a Gardner algorithm, wherein the method does not depend on additional pilot frequency data to carry out sampling frequency offset suppression, and can be independently used as a module to be processed independently of carrier frequency offset, so that a Bluetooth signal of the wireless comprehensive tester can rapidly obtain good demodulation performance.

Description

Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester
Technical Field
The invention relates to a method for inhibiting sampling frequency offset, in particular to a method for inhibiting sampling frequency offset of a Bluetooth signal of a wireless comprehensive tester.
Background
Because the clocks of the transmitter and the receiver of the communication system cannot be completely consistent, the receiving end cannot track the change of the crystal oscillator of the transmitting end without error due to the reasons of the offset of the crystal oscillator and the like, the clock of the receiving end is always slower or faster than the clock of the transmitting end, the frequency offset of the sampling clock can be generated, and the receiving end needs to carry out bit synchronization. The bit synchronization function is to track the optimal sampling point of each symbol to make a decision. When the input signal is deviated, the NCO can be adjusted according to the timing error between the local clock and the received signal, so that the sampling clock output by the NCO tracks the optimal sampling point.
Conventional digital receivers readjust the sampling clock by extracting clock information of the received signal, i.e., synchronous clock recovery. In an all-digital receiver, a clock independent of a transmitting end is generally adopted to directly sample a received signal, and then an approximate value of the signal at an optimal decision sampling moment is obtained through interpolation operation, and the method is called asynchronous clock recovery. One typical processing algorithm is the Gardner algorithm, which has the advantage of accommodating a wide range of rates of the baseband signal without changing the local sampling clock. The Gardner timing synchronization loop based on the feedback structure is widely used in practice because it does not require auxiliary data, only requires two samples per symbol, and is independent of the carrier phase, and thus has low implementation complexity.
The Bluetooth protocol 2.0 is newly added with an EDR (enhanced Data rate) type and supports 2M and 3M rates, wherein a frame header uses GFSK modulation, a payload uses pi/4-DQPSK modulation in 2M, and a payload uses 8DPSK modulation in 3M. Both are differential phase modulation signals, i.e. DPSK modulation. Fig. 2 shows the frame format of EDR packets, and a GUARD interval (GUARD) of about 5us is added between two modulated signals, GFSK and DPSK, indicated in fig. 2, including pi/4-DQPSK or 8DPSK signals.
During production testing, the DUT is typically connected to a comprehensive tester by a wired connection. The DUT and the comprehensive tester are two independent systems, so noise influence, especially inconsistency of crystal oscillators, causes inconsistency of sampling frequency, and accordingly sampling frequency offset occurs in time domain sampling of the comprehensive tester, and demodulation performance of the comprehensive tester is affected. The synchronization code may typically be generated by obtaining a 48-bit MAC address of the DUT. And then the estimation and compensation of sampling frequency offset are carried out through the known synchronous code words.
However, most bluetooth DUTs in the market are not directly marked with their MAC addresses during production, and in production test, it is necessary to quickly know the quality of signals sent by the DUTs. One common method in the prior art can perform sampling frequency offset estimation through synchronization code words of 10 symbols before a DPSK symbol, but the method is limited by the length of the synchronization code word, and is not ideal for estimating the sampling frequency offset.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a method for inhibiting the sampling frequency offset of the Bluetooth signal of the wireless comprehensive measuring instrument, which can overcome the timing error jitter and reduce the operation complexity.
Therefore, the invention provides a method for inhibiting sampling frequency offset of a Bluetooth signal of a wireless comprehensive tester, which comprises the following steps:
step S1, down-sampling the input signal;
step S2, performing matched filtering on the signal after the down-sampling processing;
step S3, setting the initial value of the interpolation control parameter Interp _ Pos (k);
step S4, go through all the input symbol data to obtain the interpolation control parameter Interp _ pos (k) of the best sampling point.
In a further development of the invention, in step S3, an initial value of the Gardner algorithm is set, wherein an initial value m of the integer part of the interpolated estimate is set00, the initial value u of the fractional part of the interpolated estimate0＝0。
A further refinement of the invention is that said step S4 comprises the following sub-steps:
step S401, the integer part m of the matched and filtered data is estimated according to interpolation through an interpolation filterkPerforming interpolation operation to obtain an interpolation filter output;
step S402, calculating a timing error according to the output value of the interpolation filter;
step S403, realizing loop filter output through a loop filter;
step S404, the interpolation controller calculates an interpolation control parameter Interp _ Pos (k) for the output of the loop filter to obtain an interpolation parameter;
step S405, repeating steps S401 to S404 until all the input symbol data are traversed.
The invention is further improved in that in step S401By the formula
Figure GDA0002361924060000021
Figure GDA0002361924060000022
Obtaining an interpolation filter output y (kT)i) Where x (-) is the input original data, k is the index value after interpolation output, TiFor the interpolation period, mkEstimating the integer part, u, for interpolationkFor estimating the fractional part of the interpolation, TsDown-sampling the input signal at a post-processing sampling rate, hIFor the interpolation function, m is the interpolation filter coefficient index value, I1And I2Taking a positive integer, (I)2+I1) Is the length of the interpolation filter.
The invention further improves the method and the device, and further comprises a step S5, wherein the step S5 is used for outputting the optimal sampling point after the sampling frequency offset is suppressed; in the step S401, the interpolation function hIA cubic interpolation function or a piecewise parabolic interpolation function is used.
The invention is further improved in that in the step S402, the formula is used
Figure GDA0002361924060000031
Calculating a timing error Err (k), wherein I (-) and Q (-) respectively represent the I signal and Q signal of the input data,
Figure GDA0002361924060000032
and
Figure GDA0002361924060000033
the corrected intermediate sampling points of the I path signal and the Q path signal are respectively, I (kT) and I ((k-1) T) are two optimal sampling points of the I path signal,
Figure GDA0002361924060000034
is the midpoint between the two optimal sampling points of the I path signal, Q (kT) and Q ((k-1) T) are the two optimal sampling points of the Q path signal,
Figure GDA0002361924060000035
the midpoint between the two best sampling points of the Q-path signal.
In a further development of the invention, in step S403, a transfer function of the loop filter is passed
Figure GDA0002361924060000036
Implementing loop filter output Err _ Loop (k), where a1And a2Is a constant related to the bandwidth of the loop filter, z-1The output recursion formula of the Loop filter is Err _ Loop (k) ═ Err _ Loop (k-1) + a as a variable of the transfer function1×Err(k)-a1(1-a2)×Err(k-1)。
In a further improvement of the present invention, in the step S404, the interpolation controller uses the formula inter _ Pos (k) ═ inter _ Pos (k-1) + (N)sA/2 + Err _ Loop (k)) calculates an interpolation control parameter Interp _ Pos (k) for the output of the loop filter; n is a radical ofsThe number of points is sampled for each symbol.
The invention is further improved in that in the step S404, the formula is used
Figure GDA0002361924060000037
And uk＝Interp_Pos(k)-mkObtaining an interpolated estimated integer part mkAnd the fractional part u of the interpolated estimatekWherein, in the step (A),
Figure GDA0002361924060000038
indicating a rounding down operation on the interpolation control parameter Interp _ Pos (k).
In a further improvement of the present invention, in the step S1, if the sampling rate of the received signal is fs1The sampling rate after down-sampling is fs2Then down-sampling the coefficient NratioIs Nratio＝fs1/fs2(ii) a In step S2, performing matched filtering on the down-sampled signal according to the nyquist non-intersymbol interference criterion, where the matched filtering uses a square root raised cosine filter that is the same as the transmit-end shaping filtering.
Compared with the prior art, the invention has the beneficial effects that: IQ data to be demodulated, which is input by a Bluetooth signal of the wireless comprehensive tester through DPSK modulation, is subjected to down-sampling processing, so that the operation complexity of matched filtering is reduced; and then, carrying out sampling frequency offset suppression on data output by matched filtering through a Gardner algorithm, wherein the method does not depend on additional pilot frequency data to carry out sampling frequency offset suppression, and can be independently used as a module to be processed independently of carrier frequency offset, so that a Bluetooth signal of the wireless comprehensive tester can rapidly obtain good demodulation performance.
Drawings
FIG. 1 is a schematic workflow diagram of one embodiment of the present invention;
FIG. 2 is a diagram of an EDR frame format in the prior art;
FIG. 3 is a DPSK modulation schematic block diagram of one embodiment of the invention;
FIG. 4 is a schematic diagram of the output waveform after IQ symbol pulse shaping according to one embodiment of the present invention;
FIG. 5 is a schematic block diagram of the Gardner algorithm implementing sample frequency offset suppression according to an embodiment of the present invention;
FIG. 6 is a detailed workflow diagram of one embodiment of the present invention;
fig. 7 is a functional block diagram of a loop filter according to an embodiment of the present invention.
Detailed Description
Preferred embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
In the embodiment, terms are explained first, and the DUT is a to-be-tested piece; edr (enhanced Data rate) is enhanced bluetooth, which is also a bluetooth signal in this example; rrc (root Raised cosine) is a square root Raised cosine pulse filter, DPSK (differential Phase Shift keying) is differential Phase Shift keying, GFSK (gaussian Frequency Shift keying) is gaussian Frequency Shift keying, GUARD is the direct GUARD interval of GFSK and DPSK modulation, nco (numerically controlled oscillator) is a numerically controlled oscillator, and evm (error Vector magnitude) is the error Vector magnitude.
In the embodiment, a Gardner algorithm is used for restraining the sampling frequency offset of the Bluetooth DPSK signal, wherein for the problem of overlarge sampling rate in a wireless comprehensive measuring instrument, the embodiment performs down-sampling processing on the signal before performing matched filtering; and the error detector of the Gardner algorithm is improved to meet the requirement of DPSK demodulation of the Bluetooth signal of the wireless comprehensive measuring instrument and solve the problem of error jitter detection of a timing loop.
For the EDR bluetooth system of the wireless integrated measuring instrument, that is, for the bluetooth signal of the wireless integrated measuring instrument described in this example, assuming that the access code and the packet header have been correctly demodulated, the baseband signal model of the DPSK received signal may be expressed as: r (t) ═ ΣicigT(t-iT- τ) + n (t), wherein { ciTransmitting symbol sequence, DPSK modulating the Bluetooth signal of wireless comprehensive measuring instrument,
Figure GDA0002361924060000051
τ is the time offset. gTFor the sending end shaping filter, the bluetooth protocol provides for the use of square root raised cosine filter, and n (t) is gaussian noise.
After matched filtering, y (t) sigmaicih (t-iT- τ) + n (t), wherein h (t) gT(t)*gR(t) raised cosine matched filtering; the Gardner algorithm timing error detector can be represented by
Figure GDA0002361924060000052
Figure GDA0002361924060000053
Where Re {. is a real part of the complex number, y*(. cndot.) denotes taking complex conjugation. y ((k-1) T) and y (kT) are the optimal sampling points,
Figure GDA0002361924060000054
the middle sampling zero. Err (k) is the clock error. If no sampling frequency offset exists in an ideal situation, if the values of the I path and the Q path of the kth sampling point and the (k-1) T sampling point of the modulation signal are equal, y ((k-1) T) -y (kT) is 0, and if the signs are opposite, the values of the I path and the Q path are equal
Figure GDA0002361924060000055
This sampling point should be zero. That is, if the synchronization is correct, err (k) is 0. Assume timing advance, err (k)<0, assuming timing hysteresis, err (k)>0。
However, the Gardner algorithm is proposed based on BPSK/QPSK modulation, for the I-path of BPSK modulation and the I-path and Q-path of QPSK modulation. If the polarities of two adjacent optimal sampling points are reversed, the average value of the middle point should be zero when there is no timing error. If the polarities of the two adjacent optimal sampling points are not reversed, the two peaks are equal, and the difference value is zero. Thus, for BPSK/QPSK modulation, the timing error can always be made 0 when there is no timing error. For pi/4-DQPSK modulation and 8DPSK modulation of the Bluetooth signal (EDR) of the wireless comprehensive measuring instrument, the modulated symbols of the I path and the Q path have a plurality of values, for example, for Bluetooth DPSK modulation, the possible values of the two paths of modulated IQ are
Figure GDA0002361924060000056
Five values, when the sign changes from-1 to 1, or from
Figure GDA0002361924060000057
Become into
Figure GDA0002361924060000058
Etc., the mean value of the intermediate points is 0 in the absence of timing errors, but in other cases, such as the sign changes from-1 to
Figure GDA0002361924060000059
The average value of the intermediate point is not zero, which may cause jitter of timing error and affect demodulation performance of the receiving end.
The sampling frequency offset suppression method is suitable for a Bluetooth signal (EDR) of a wireless integrated measuring instrument, can overcome timing error jitter of a Gardner algorithm under the condition of modulation symbol amplitude diversity, and reduces the operation complexity of the whole algorithm under the condition of higher sampling rate; the method is used in a Bluetooth demodulation system of a wireless comprehensive measuring instrument, and can meet the requirements of frequency offset estimation and suppression.
As shown in fig. 3, binary information {0,1} is original data to be transmitted, and is first converted into two paths through serial-parallel conversion, then the two paths of signals are subjected to differential phase coding to correspond bit information to position information of DPSK constellation points, and orthogonal IQ two-path signals i (k) and q (k) are output, and the signals after symbol mapping are pulse signals, have a large number of high-frequency components, and are not suitable for transmission on a channel, so that after shaping filtering is required, a signal spectrum changes, a high-frequency part is suppressed, and crosstalk between signals is avoided, so as to reduce an error rate, a roll-off coefficient β of an RRC filter defined in a bluetooth protocol is 0.4, a modulated signal bandwidth is (1+ β)/(2T), where T is 1 μ s, and a frequency function form of a root-raised cosine filter is given in the protocol:
Figure GDA0002361924060000061
the time domain response h (t) of the filter can be easily obtained by inverse fourier transform. For DPSK modulation of bluetooth signals, possible values of the output symbols of the IQ modulation are
Figure GDA0002361924060000062
Fig. 4 is a schematic diagram of an output waveform of the I-channel signal after pulse forming.
As can be seen from fig. 4, due to the diversity and randomness of the DPSK modulation values, even if the signal is sampled without any sampling frequency offset, when the modulation symbol polarity is reversed, I ((k) in fig. 4 appears1-1) T) and I ((k)1T), at a point in the middle between the two optimal sampling points
Figure GDA0002361924060000063
It is also possible that I ((k) in fig. 4) occurs when the modulation symbol polarity is not inverted2-1) T) and I ((k)2T) are not equal, sampling values are obtained
Figure GDA0002361924060000064
I((k2-1) T), and I ((k)2T) substituting into formula
Figure GDA0002361924060000065
Figure GDA0002361924060000066
Whether the timing is advanced or retarded, formula
Figure GDA0002361924060000067
Figure GDA0002361924060000068
A negative number is obtained. The Gardner algorithm therefore does not accurately extract the instantaneous timing errors of DPSK modulation, which, although averaged to zero from a large data sample, can cause timing jitter. Therefore, if demodulation is performed at the receiving end, the formula is directly used
Figure GDA0002361924060000069
Figure GDA00023619240600000610
When error detection is carried out, great system self-noise can be generated, and the EVM of the integrated tester for measuring the Bluetooth signals is poor.
As can be seen from fig. 4, the value of DPSK is asymmetric, so that the midpoint value may not be zero regardless of whether the polarity between adjacent symbols is reversed. It can be seen from fig. 4 that the midpoint value corresponds to the zero point offset of (I (kt)) + I (k-1) T))/2, and therefore, the intermediate sampling point is only zeroed, and therefore, the formula
Figure GDA00023619240600000611
Becomes the timing error detection formula of the Gardner algorithm
Figure GDA00023619240600000612
Wherein, I and Q respectively represent IQ two-path signals,
Figure GDA00023619240600000613
and
Figure GDA00023619240600000614
the corrected intermediate sampling point. I (kT) and I ((k-1) T) are two optimal sampling points,
Figure GDA00023619240600000615
for the midpoint between the two optimal samples, error values are calculated for the I and Q signals, respectively, and then summed.
Therefore, the block diagram of the present example for performing sampling frequency offset suppression on DPSK modulation signal of the wireless integrated instrument bluetooth signal by using Gardner algorithm is shown in fig. 5, wherein it is assumed that both the access code and the packet header of the wireless integrated instrument bluetooth signal (EDR) are correctly synchronized and demodulated, and therefore, fig. 5 omits the synchronization and carrier frequency offset estimation of the demodulation part of the wireless integrated instrument bluetooth signal (EDR), and only shows the block diagram for performing sampling frequency offset suppression on DPSK signal by using Gardner algorithm.
In summary, as shown in fig. 1 and fig. 6, this embodiment provides a method for suppressing sampling frequency offset of a bluetooth signal of a wireless comprehensive tester, including the following steps:
step S1, down-sampling the input signal;
step S2, performing matched filtering on the signal after the down-sampling processing;
step S3, setting the initial value of the interpolation control parameter Interp _ Pos (k);
step S4, traversing all the input symbol data, and obtaining the interpolation control parameter Interp _ Pos (k) of the optimal sampling point;
and step S5, outputting the optimal sampling point after the sampling frequency offset is suppressed.
In this example, step S1 is to perform down-sampling on the input IQ signals i (n) and q (n) to effectively reduce the complexity of the matched filter, aiming at the problem of high sampling rate of the integrated wireless measuring instrument. This is because the general wireless comprehensive tester can support the test of multiple wireless protocols, so the single carrier waveThe same sampling rate is used for multicarrier systems, and therefore the sampling rate is typically above 80 MHz. Assume that the received signal has a sampling rate fs1The sampling rate after down-sampling is fs2Then down-sampling the coefficient NratioIs Nratio＝fs1/fs2. The down-sampling process is to perform every coefficient N on the original signalratioA new signal is formed by taking one data. The step S5 is a preferred step.
In step S2, the signal after down-sampling is matched and filtered according to the nyquist non-intersymbol interference criterion, the matched filter uses the square root raised cosine filter which is the same as the transmit-end shaped filter, and the formula of the matched filter is
Figure GDA0002361924060000071
Where β is the roll-off factor, preferably β ═ 0.4, T is the symbol period, preferably T ═ 1 μ s, f is the frequency variable, and h (f) is the frequency function of the square root raised cosine filter.
In step S3, the initial value of the interpolation control parameter inter _ pos (k) is set as the initial value of the Gardner algorithm, wherein the initial value m of the integer part of the interpolation estimation value is set00, the initial value u of the fractional part of the interpolated estimate0＝0。
Step S4 in this example includes the following substeps:
step S401, the integer part m of the matched and filtered data is estimated according to interpolation through an interpolation filterkPerforming interpolation operation to obtain an interpolation filter output;
step S402, calculating a timing error according to the output value of the interpolation filter;
step S403, realizing loop filter output through a loop filter;
step S404, the interpolation controller calculates an interpolation control parameter Interp _ Pos (k) for the output of the loop filter to obtain an interpolation parameter;
step S405, repeating steps S401 through S404 until all input symbol data are traversed, assuming a total of received dataThe length is N, until the calculated interpolation control parameter Interp _ Pos (k) is more than or equal to N-NsThe sampling frequency deviation suppression method is quitted and the best sampling point after the sampling frequency deviation suppression is output, NsThe number of sample points per symbol, k is the number of cycles. In step S401, the integer part m of the matched and filtered data is estimated according to interpolation0Performing interpolation operation by formula
Figure GDA0002361924060000084
Figure GDA0002361924060000085
Obtaining an interpolation filter output y (kT)i) Where x (-) is the input original data, k is the index value after interpolation output, TiFor the interpolation period, mkEstimating the integer part, u, for interpolationkFor estimating the fractional part of the interpolation, TsDown-sampling the input signal at a post-processing sampling rate, hIM is an interpolation function, and is an interpolation filter coefficient index value; i is1And I2Taking a positive integer, (I)2+I1) Is the length of the interpolation filter. In actual calculation, the I path signal and the Q path signal output by matched filtering are respectively substituted into a formula
Figure GDA0002361924060000086
Figure GDA0002361924060000087
X (-) to obtain the output value of the interpolation filter. Wherein the interpolation function h of the interpolation filterIPreferably a cubic interpolation function or a piecewise parabolic interpolation function is used.
Since the Gardner algorithm only needs to ensure that the optimal sampling point and the zero point are used for clock error detection, one symbol period only needs to perform interpolation of two points, one is the optimal sampling point and the other is the zero crossing point between the optimal sampling points, so that the T isi2T, wherein TiT is the symbol period for the interpolation period.
In the step S402 described in this example,by the formula
Figure GDA0002361924060000081
Calculating a timing error Err (k), wherein I (-) and Q (-) respectively represent outputs of the I path signal and the Q path signal of the input data after matched filtering,
Figure GDA0002361924060000082
and
Figure GDA0002361924060000083
the corrected intermediate sampling points of the I path signal and the Q path signal are respectively, I (kT) and I ((k-1) T) are two optimal sampling points of the I path signal,
Figure GDA0002361924060000091
is the midpoint between the two optimal sampling points of the I path signal, Q (kT) and Q ((k-1) T) are the two optimal sampling points of the Q path signal,
Figure GDA0002361924060000092
the midpoint between the two best sampling points of the Q-path signal. The optimal sampling point of the I path signal and the optimal sampling point of the Q path signal are the optimal decision points after matched filtering output.
In step S403 in the present example, the transfer function of the loop filter is passed
Figure GDA0002361924060000093
Figure GDA0002361924060000094
Implementing loop filter output Err _ Loop (k), where a1And a2Is a constant related to the bandwidth of the loop filter, z-1Is a variable of a transfer function, z-1To represent
Figure GDA0002361924060000095
The output recursion formula of the Loop filter is Err _ Loop (k) ═ Err _ Loop (k-1) + a1×Err(k)-a1(1-a2) XErr (k-1). Wherein, Err _ Loop (k-1) and Err _ Loop (k) are respectively as aboveThe once calculated loop filter output value and the current calculated loop filter output value, err (k) timing error detector output value; a schematic block diagram of the loop filter is shown in fig. 7.
In this example, step S404 is used to calculate an interpolation control parameter inter _ Pos (k), and the interpolation controller uses the formula inter _ Pos (k) ═ inter _ Pos (k-1) + (N)sAnd/2 + Err _ Loop (k)) further calculating the output of the loop filter to obtain an interpolation control parameter Interp _ Pos (k). By the formula inter _ Pos (k) ═ inter _ Pos (k-1) + (N)sIt can be known that the interpolation control parameter Interp _ Pos (k) is an accumulated value, which mainly ensures that the interpolation filter performs two times of interpolation point calculation in one symbol period. After obtaining the interpolation control parameter Interp _ Pos (k), the interpolation control parameter Interp _ Pos (k) is calculated according to the formula
Figure GDA0002361924060000097
And uk＝Interp_Pos(k)-mkObtaining an interpolated estimated integer part mkAnd the fractional part u of the interpolated estimatekWherein, in the step (A),
Figure GDA0002361924060000098
indicating that the interpolation control parameter Interp _ Pos (k) is rounded down; n is a radical ofsThe number of points is sampled for each symbol.
Therefore, in the embodiment, the Gardner algorithm is used in the wireless comprehensive measuring instrument to effectively suppress the DPSK sampling frequency offset of the Bluetooth signal of the wireless comprehensive measuring instrument; for the multivalue and the randomness of the DPSK modulation signal, formulas are respectively used according to the polarities of adjacent symbols
Figure GDA0002361924060000096
Calculating the timing error of the clock; the method for suppressing the sampling frequency offset of the Bluetooth signal of the wireless comprehensive tester samples the software architecture shown in figure 6.
The IQ data to be demodulated, which is input by a Bluetooth signal of the wireless comprehensive tester through DPSK modulation, is subjected to down-sampling processing, so that the operation complexity of matched filtering is reduced; and then, carrying out sampling frequency offset suppression on data output by matched filtering through a Gardner algorithm, wherein the method does not depend on additional pilot frequency data to carry out sampling frequency offset suppression, and can be independently used as a module to be processed independently of carrier frequency offset, so that a Bluetooth signal of the wireless comprehensive tester can rapidly obtain good demodulation performance.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.

Claims (2)

1. A method for suppressing sampling frequency offset of Bluetooth signals of a wireless comprehensive tester is characterized by comprising the following steps:
step S1, down-sampling the input signal;
step S2, performing matched filtering on the signal after the down-sampling processing;
step S3, setting an initial value of an interpolation control parameter Interp _ pos (k) to 0, where the interpolation control parameter Interp _ pos (k) is a control parameter of an interpolation controller; wherein, an initial value m of an integer part of an interpolation estimation value of an interpolation control parameter Interp _ Pos (k) is set00, the initial value u of the fractional part of the interpolated estimate0＝0；
Step S4, traversing all the input symbol data, and obtaining the interpolation control parameter Interp _ Pos (k) of the optimal sampling point;
the step S4 includes the following sub-steps:
step S401, the integer part m of the matched and filtered data is estimated according to interpolation through an interpolation filterkPerforming an interpolation operation, the interpolation operation being by formula
Figure FDA0002361924050000011
Figure FDA0002361924050000012
To obtain an interpolation filter output y (kT)i) Wherein x (·)) For the input raw data, k is the interpolated output index value, TiFor the interpolation period, mkEstimating the integer part, u, for interpolationkFor estimating the fractional part of the interpolation, TsFor the down-sampled sampling rate of the input signal, the interpolation function hIUsing cubic interpolation function or piecewise parabolic interpolation function, m being the index value of interpolation filter coefficient, I1And I2Taking a positive integer, (I)2+I1) Is the length of the interpolation filter;
step S402, calculating a timing error Err (k) according to the output value of the interpolation filter, wherein the calculation formula of the timing error Err (k) is
Figure FDA0002361924050000013
Wherein, I (-) and Q (-) represent I signal and Q signal of input data respectively,
Figure FDA0002361924050000014
and
Figure FDA0002361924050000015
the corrected intermediate sampling points of the I path signal and the Q path signal are respectively, I (kT) and I ((k-1) T) are two optimal sampling points of the I path signal,
Figure FDA0002361924050000016
is the midpoint between the two optimal sampling points of the I path signal, Q (kT) and Q ((k-1) T) are the two optimal sampling points of the Q path signal,
Figure FDA0002361924050000017
the middle point between two optimal sampling points of the Q path signal is set;
step S403, passing the transfer function of the loop filter
Figure FDA0002361924050000018
Implementing loop filter output Err _ Loop (k), where a1And a2Is a constant related to the bandwidth of the loop filter, z-1The output recursion formula of the Loop filter is Err _ Loop (k) ═ Err _ Loop (k-1) + a as a variable of the transfer function1×Err(k)-a1(1-a2)×Err(k-1)；
Step S404, the interpolation controller calculates the interpolation control parameter Interp _ Pos (k) for the output of the loop filter, and then passes the formula
Figure FDA0002361924050000021
And uk＝Interp_Pos(k)-mkObtaining an interpolated estimated integer part mkAnd the fractional part u of the interpolated estimatekWherein, in the step (A),
Figure FDA0002361924050000022
indicating a rounding down of the interpolation control parameter Interp _ Pos (k), which is given by the formula Interp _ Pos (k) ═ Interp _ Pos (k-1) + (N)sA/2 + Err _ Loop (k)) calculates an interpolation control parameter Interp _ Pos (k) for the output of the loop filter; n is a radical ofsSampling the number of points for each symbol;
step S405, repeating the steps S401 to S404 until all the input symbol data are traversed, wherein the symbol data are IQ two-path signals input by the interpolation filter, and if the total length of the received data is N, the calculated interpolation control parameter Interp _ Pos (k) is not less than N-NsThe sampling frequency deviation suppression method is quitted and the best sampling point after the sampling frequency deviation suppression is output, NsThe number of points is sampled for each symbol.
2. The method for suppressing sampling frequency offset of bluetooth signal in integrated wireless tester as claimed in claim 1, wherein in step S1, if the sampling rate of the received signal is fs1The sampling rate after down-sampling is fs2Then down-sampling the coefficient NratioIs Nratio＝fs1/fs2(ii) a In step S2, performing matched filtering on the down-sampled signal according to the nyquist non-intersymbol interference criterion, where the matched filtering uses a square root raised cosine filter that is the same as the transmit-end shaping filtering.
CN201710035632.5A 2017-01-18 2017-01-18 Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester Active CN106603217B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710035632.5A CN106603217B (en) 2017-01-18 2017-01-18 Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710035632.5A CN106603217B (en) 2017-01-18 2017-01-18 Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester

Publications (2)

Publication Number Publication Date
CN106603217A CN106603217A (en) 2017-04-26
CN106603217B true CN106603217B (en) 2020-03-20

Family

ID=58585952

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710035632.5A Active CN106603217B (en) 2017-01-18 2017-01-18 Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester

Country Status (1)

Country Link
CN (1) CN106603217B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108040026A (en) * 2017-12-05 2018-05-15 上海富芮坤微电子有限公司 A kind of sign synchronization and equalizer for bluetooth receiver
CN108011649B (en) * 2017-12-05 2019-12-24 上海富芮坤微电子有限公司 Method for constructing symbol synchronization circuit in Bluetooth EDR receiver
CN108183841B (en) * 2017-12-29 2020-07-28 深圳市极致汇仪科技有限公司 Base band data processing method and system based on IEEE802.11ah in comprehensive tester
CN110971325B (en) * 2019-12-13 2021-09-10 北京无线电计量测试研究所 Time transfer method based on bit synchronization
CN113630212B (en) * 2021-08-05 2024-03-19 深圳市思码逻辑技术有限公司 Downsampling method and downsampling device for digital signals

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20000060085A (en) * 1999-03-11 2000-10-16 구자홍 apparatus for recovering clock in digital TV
CN103969628A (en) * 2014-04-08 2014-08-06 南京航空航天大学 Compressive-sensing-based synthetic aperture radar PFA imaging algorithm
CN106302296A (en) * 2016-08-30 2017-01-04 广州海格通信集团股份有限公司 High dynamically narrow band signal frequency tracking method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9553714B2 (en) * 2015-06-26 2017-01-24 Broadcom Corporation Frequency multiplier for a phase-locked loop

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20000060085A (en) * 1999-03-11 2000-10-16 구자홍 apparatus for recovering clock in digital TV
CN103969628A (en) * 2014-04-08 2014-08-06 南京航空航天大学 Compressive-sensing-based synthetic aperture radar PFA imaging algorithm
CN106302296A (en) * 2016-08-30 2017-01-04 广州海格通信集团股份有限公司 High dynamically narrow band signal frequency tracking method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于FPGA的π/4QPSK调制解调器的研究与设计;武光辉;《中国优秀硕士论文全文数据库信息科技辑》;20110515;第21-37页 *

Also Published As

Publication number Publication date
CN106603217A (en) 2017-04-26

Similar Documents

Publication Publication Date Title
CN106603217B (en) Sampling frequency offset suppression method for Bluetooth signal of wireless comprehensive tester
US11431534B2 (en) Circuit and method for compensating frequency offset in wireless frequency shift keying communication
JP2526931B2 (en) PSK signal demodulator
CN111343125B (en) 32APSK modulation system receiver synchronization method
CN110912847B (en) GMSK signal demodulation method
US9722845B2 (en) Bluetooth low energy frequency offset and modulation index estimation
CN110300079B (en) MSK signal coherent demodulation method and system
CN111600823B (en) Parallel OQPSK offset quadriphase shift keying demodulator
CN111585933B (en) Receiver burst signal synchronization method and device of single carrier frequency domain equalization system
CN111935046B (en) Low-complexity frequency shift keying signal symbol rate estimation method
KR20170079157A (en) Bluetooth signal receiving method and device using improved symbol timing offset compensation
US20110170583A1 (en) Method for tracking the phase of a signal modulated by continuous phase modulation and synchronization device implementing said method
CN105049174A (en) Carrier and clock combined synchronization method for OQPSK modulation
US6731710B1 (en) Method for rapid carrier frequency estimation in a communication system
CN113115430A (en) High-speed burst digital demodulation system
CN115118564B (en) Carrier frequency deviation estimation method and device
CN117834034A (en) Self-adaptive modulation and demodulation method for 100g industrial grade single-wave optical module
CN109818894A (en) GMSK signal detecting method and detection device under multipath channel
EP1117219B1 (en) Frequency offset correction in the presence of intersymbol interference
KR100760793B1 (en) Correction of quadrature and gain errors in homodyne receives
JP2002232389A (en) Ofdm receiver
CN114465691A (en) Low-complexity constant envelope phase modulation signal sampling deviation estimation and compensation method and system
CN114499646B (en) VDE-SAT channel physical layer signal time domain pilot frequency equalization method and equipment
CN116155668B (en) Anti-frequency offset carrier recovery method, system and storage medium
US6952445B2 (en) Symbol constellations having second-order statistics with cyclostationary phase

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

Address after: 518000 5C, 1, building No. 6, Ting Wei Road, 67 District, Xingdong community, Baoan District Xin'an, Shenzhen, Guangdong.

Applicant after: SHENZHEN JIZHI HUIYI TECHNOLOGY CO., LTD.

Address before: 518000 Building 2, Zone 2, Block 2, Honghualing Industrial South Zone, 1213 Liuxian Avenue, Taoyuan Street, Nanshan District, Shenzhen City, Guangdong Province

Applicant before: SHENZHEN JIZHI HUIYI TECHNOLOGY CO., LTD.

CB02 Change of applicant information
GR01 Patent grant
GR01 Patent grant