CN102355305A - Linear chirp z transform based frequency offset estimation algorithm in M-QAM (M-ary Quadrature Amplitude Modulation) coherent optical communication system - Google Patents
Linear chirp z transform based frequency offset estimation algorithm in M-QAM (M-ary Quadrature Amplitude Modulation) coherent optical communication system Download PDFInfo
- Publication number
- CN102355305A CN102355305A CN2011103034472A CN201110303447A CN102355305A CN 102355305 A CN102355305 A CN 102355305A CN 2011103034472 A CN2011103034472 A CN 2011103034472A CN 201110303447 A CN201110303447 A CN 201110303447A CN 102355305 A CN102355305 A CN 102355305A
- Authority
- CN
- China
- Prior art keywords
- frequency
- qam
- algorithm
- signal
- czt
- 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.)
- Pending
Links
Images
Landscapes
- Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
Abstract
The invention provides a linear chirp z transform based frequency offset estimation algorithm in an M-QAM (M-ary Quadrature Amplitude Modulation) coherent optical communication system, which comprises the following steps that: a smaller amount of M-QAM signal receiving sample values are selected for four-power; the FFT (Fast Fourier Transform) is conducted to obtain a frequency corresponding to the maximum spectral value to be recorded as a roughly estimated value; the linear CET (Chirp z Transform) of four-power signals is conducted near the roughly estimated value to obtain a frequency corresponding to the maximum spectral value; and the frequency is divided by four to obtain a frequency offset estimated value of M-QAM signals. The linear chirp z transform based frequency offset estimation algorithm does not adopt a gradient descent method, a Newton method and other methods depending on a step size and the number of symbols, but adopts a method that the frequency is finely estimated by CZT with a fixed computation load. Since the CZT can be realized by the FFT with fast algorithm, the complexity of the algorithm can be obviously reduced in the algorithm, and meanwhile, the algorithm is applied to all-order QAM signals.
Description
Technical field
The present invention relates to the communications field, more particularly, relate to the Frequency offset estimation scheme in the optical fiber telecommunications system that adopts the M-QAM modulation format in the coherent optical communication.
Background technology
Optical fiber telecommunications system starts from early 1970s, adopts the semiconductor laser of intensity modulated, and the light intensity signal that receives becomes the signal of telecommunication through photodiode converts.This mode is called as intensity modulated/directly detect, and (intensity modulation/direct detection IM/DD), and is widely used in current commercial optical fiber telecommunications system.But directly detections/intensity modulated has its intrinsic deficiency, receives the restriction of noise such as reception and sensitivity, and (spectral efficiency) is not high for spectrum efficiency, can not make full use of bandwidth etc.In the research in early days, adopt the constant differential QPSK modulation format of intensity, and adopt Differential Detection, to improve spectrum efficiency at receiving terminal.This method can pass 2 bits (bit) in each symbol (symbol), theoretic spectrum efficiency can reach 2bit/s/Hz.(signal to noise ratio SNR), can adopt the coherent detection that obtains extensive concern for further raising frequency efficiency improves signal to noise ratio simultaneously.
Coherent detection recovers the amplitude and the phase place of modulation intelligence, can improve spectrum efficiency, improves the signal to noise ratio of receiving end signal simultaneously through local oscillator (local oscillator) light.The principle of coherent reception system as shown in Figure 1, the mixing in 90 ° of frequency mixers of receiving optical signals and local oscillator light signal, the balanced detector on the road of going up detects the in-phase component that receives signal, the quadrature component that receives signal is detected on road down.In order to guarantee the correctness of symbol judgement, the local oscillator light frequency must be consistent with the frequency of transmitter.Generally speaking, local oscillator light frequency f
LOWith transmitter frequency f
TIncomplete same, make the signal that receives on planisphere, rotate the angle of rotation and frequency-splitting Δ f=f with respect to original modulated signal
LO-f
TRelevant.If the signal to this rotation does not carry out frequency offset compensation, directly carry out symbol judgement, can not obtain correct symbol.
Along with the development of Digital Signal Processing and hardware handles ability, can estimate that to frequency offseting value Δ f the method through software compensates it then through the mode of Digital Signal Processing.In the middle of the Frequency Estimation technology of up-to-date proposition, different algorithm for estimating has been proposed to different modulation formats.These frequency estimation algorithms or directly handle (time domain approach) to received signal, or utilize and receive signal frequency-domain information and carry out the estimation (frequency domain method) of frequency shift (FS) Δ f.For adopting the time domain estimation approach, in order to improve the precision of Frequency Estimation, need more symbolic number or sampled value (sample) usually, and then need more memory cell and than multiplied unit and bigger processing delay.
In order to improve spectrum efficiency as much as possible, need to adopt more complicated modulation format.(M-ary quadrature amplitude modulation M-QAM) is carried in data message on the amplitude and phase place of symbol the quadrature amplitude phase modulated simultaneously, in identical bandwidth, can transmit more bits.It is higher that but the demodulation of M-QAM modulating system requires frequency accuracy, needs more accurate frequency estimating methods, considers the real-time requirement in the commercial system, and the complexity of algorithm can not be too high.In nearest progress, the researcher proposes the frequency offset compensation scheme to M-QAM in succession, and these methods can be divided into the method for time domain and the method for frequency domain substantially.
In time domain approach, diverse frequency estimating methods is arranged to the M-QAM of different rank.For the signal of 16-QAM, can adopt the method for phase difference to carry out Frequency Estimation.Yet this need classify to the receiving symbol of 16-QAM; And this sorting technique only is applicable to the 16-QAM modulation format; Be not suitable for the QAM modulation of other exponent numbers, this method has only been used the part symbol of 16-QAM simultaneously, and strict to the symbol of participating in Frequency Estimation.
In frequency approach, have to receive to M-QAM that the phase information of signal handles, also there is couple M-QAM to receive that signal itself handles.To the signal of 16-QAM modulation format, its part symbol is rotated, a global maximum appears in the frequency spectrum of its phase place, and this maximum is relevant with the frequency offseting value Δ f that 16-QAM receives signal.But this method also has its limitation, at first is the modulation format for high-order (M＞16), the classification more complicated of symbol, and be not easy to the design of program; Secondly, in order correctly to estimate frequency offseting value, (fast Fourier transform, FFT) computing increase system complexity need to carry out twice identical fast Fourier transform; The 3rd, for obtaining high estimation accuracy, it is bigger to carry out counting of FFT, 2
14Magnitude causes the complexity of algorithm high.
Because the symmetry of M-QAM symbol has the scholar to propose directly the reception signal of M-QAM to be handled.After carrying out 4 powers to received signal, its frequency spectrum also global maximum can occur, and this maximum is f=4 Δ f in the frequency of correspondence.FFT computing through counting less earlier; Obtain the rough estimate evaluation of frequency shift (FS), this is worth near actual value 4 Δ f, then through optimized methods such as gradient descent method, Newton methods; Near this coarse value, seek maximum spectrum value, thereby obtain accurate frequency estimation.Yet the convergence of optimal methods such as gradient descent method or newton's descent method receives the influence of the symbolic number (or sample number) of participating in computing, iteration step length, convergence precision etc., makes the algorithm complex and the estimated accuracy of whole frequency estimation algorithm be difficult to confirm.
Summary of the invention
The technical problem that the present invention will solve is to provide a kind of computational complexity to confirm, operand is little, and the frequency offset estimation algorithm that the system of the QAM modulation format that adopts different rank all is suitable for.The present invention need not to adopt optimized methods such as gradient descent method, Newton method to improve estimated accuracy, and this algorithm adopts chirp Z transform (CZT) to improve estimated accuracy behind the frequency offseting value that obtains rough estimate.It should be noted that CZT can realize through the FFT with fast algorithm, and its operand fixes, and can accurately estimate frequency offseting value Δ f.
The present invention all is suitable for the M-QAM modulation signal of any exponent number.At first receive signal and carry out 4 power computings, search the maximum of its frequency spectrum through two steps then, and then estimate the frequency offseting value Δ f that M-QAM receives signal.Substep estimates to reduce significantly the computational complexity of algorithm, makes frequency estimation algorithm keep higher precision simultaneously.
Other characteristics of the present invention and advantage will be elaborated in specification, claims and accompanying drawing subsequently, also can understand through embodiment of the present invention simultaneously.
Description of drawings
Through the description of carrying out below in conjunction with accompanying drawing to embodiment, above-mentioned and/or other purposes of the present invention and advantage will become apparent, wherein:
Fig. 1 illustrates the schematic diagram of coherent detection
Fig. 2 illustrates the Frequency Estimation skew algorithm for estimating of M-QAM
Fig. 3 illustrates the rough estimate and the corresponding amplitude spectrum of thin estimating part of 256-QAM
Fig. 4 illustrates the Frequency Estimation deviation under the different signal to noise ratios of 64-QAM
Fig. 5 illustrates the Frequency Estimation deviation of the different frequency deviant of 64-QAM
Embodiment
To combine accompanying drawing that execution mode of the present invention is described in detail below.
Fig. 1 has shown the optical receiver principle that adopts coherent detection.In last branch road, local oscillator light signal and receiving optical signals direct conversion, through behind the balanced detector, output be the in-phase component of modulation signal, can be designated as the real part of M-QAM signal; In road down, the local oscillator light signal is earlier through 90 ° of time delay devices, and then with the receiving optical signals mixing, through behind the balanced detector, output be the quadrature component of modulation signal, be designated as the imaginary part of M-QAM signal.What exported on upper and lower road is the signal of telecommunication, in follow-up digital signal processing chip, needs to carry out discrete sampling to it earlier and handles.
Fig. 2 demonstrates the algorithm flow that the method that adopts Digital Signal Processing is carried out M-QAM communication system Frequency offset estimation, and its detailed arthmetic statement is following.
Sampled value after the discretization is y (k), and it is carried out 4 power computings, receives signal for M-QAM, and no matter why order of modulation M is worth, and all carries out 4 power computings.
For the data behind 4 powers, to find the solution its amplitude spectrum through FFT, and find the corresponding frequency of its maximum, this frequency values is the rough estimate value of 4 * Δ f, is designated as
According to the viewpoint of communication, y
4(k) be analytic signal, its amplitude spectrum is asymmetric at frequency domain about frequency origin, will theoretic impact value near 4 * Δ f, occur, on numerical computations, is a global maximum.Therefore obtain y finding the solution through FFT
4(k) behind the frequency spectrum,, can estimate 4 * Δ f roughly through searching for the maximum of its amplitude spectrum.In order to reduce the complexity of computing, the points N of FFT, just y
4Counting (k) should not be excessive and should be 2 integral number power, simultaneously in order to improve precision, needs further to estimate.
Obtaining the frequency values of rough estimate
After,
Neighbouring to y
4(k) carry out M point CZT.CZT can near the frequency spectrum of analytic signal optional frequency, and can be calculated by FFT through the Bluestein structure, so CZT has the fast algorithm consistent with FFT.In theory, when consideration of noise not, y
4(k) frequency spectrum is symmetrical about 4 * Δ f, and has
Wherein δ is the frequency domain resolution of FFT in the rough estimate.In fact, noise can not be eliminated fully, in order between precision and complexity, to try to achieve balance, can
The interval on to y
4(k) carry out CZT.And similar rough estimates, the CZT obtained, the search of its maximum amplitude spectrum can be obtained by high-precision frequency offset value, denoted?
Because the estimated frequency offset value is 4 × Δf, therefore need to divide 4 to obtain the M-QAM received signal frequency offset estimate value, denoted?
Fig. 3 has shown rough estimate and the thin frequency spectrum precision signal of estimating for two steps.Its modulation format is 256QAM, and frequency offseting value is Δ f=-1.2GHz, and columnar alignment is 256 FFT, and solid line is 65 CZT, and star (*) is the maximum of amplitude spectrum.Correspondent frequency estimated value 4 * Δ f provides in legend.Can know that by figure CZT can further improve the precision of estimation, and can satisfy the requirement of the high accuracy Frequency Estimation precision of high-order QAM modulation signal.
Figure 4 shows the frequency estimation algorithm under different SNR is the mean square error (mean? Square? Error, MSE, which was defined as?
).Wherein modulation format is 64-QAM, and frequency offseting value is Δ f=-1.2GHz, and it is 512 that FFT counts, and it is 65 that CZT counts.Star-like solid line is the mean square error curve that adopts after CZT carefully estimates, square solid line is the mean square error that rough estimate is only arranged.When signal to noise ratio snr during greater than 15dB, the precision of estimation that has adopted CZT is apparently higher than the precision that rough estimate is only arranged, and when SNR=18dB, estimated accuracy improves more than 100 times, is enough to satisfy subsequent phase and estimates the requirement to the Frequency Estimation precision.
This frequency estimation algorithm of mean square error when Fig. 5 illustrates to(for) different frequency shift (FS) (mean square error, MSE).Wherein modulation format is 64-QAM, and it is 512 that signal to noise ratio snr=19dB, FFT count, and it is 65 that CZT counts, and frequency offseting value Δ f is at [1.5GHz, 1.5GHz].Star-like solid line is the mean square error curve that adopts after CZT carefully estimates, square solid line is the mean square error that rough estimate is only arranged.Can know that after carefully estimating, the rough estimate of Frequency Estimation ratio of precision is higher, and for different frequency shift (FS) Δ f, the stable accuracy that CZT estimates is 10
-8Magnitude, and the precision of rough estimate is unfavorable for frequency shift (FS) in fluctuation in a big way.
Through detailed description with reference to top embodiment and accompanying drawing to exemplary indefiniteness, advantage of the present invention and characteristic and realize that method of the present invention can be easier to understanding.Yet the present invention all is suitable for the M-QAM modulation format of different rank, and the embodiment that should not be construed as limited in this explaination.
Claims (4)
1. a kind of algorithm of the M-QAM Frequency offset estimation based on CZT in the coherent optical communication, this algorithm comprise following four each and every one steps:
Carry out 4 power computings to received signal;
Ask the frequency spectrum of the signal behind 4 powers through fast Fourier transform (FFT);
Find the solution the maximum and the corresponding frequency of this spectrum amplitude;
Near said frequencies, ask the frequency spectrum of the signal behind 4 powers through linear chrip z conversion (CZT), and search the corresponding frequency of maximum.
2. the method for claim 1, wherein carry out 4 powers to the received signal, all adopt 4 power computings for the QAM signal of different order of modulation;
Carry out counting of 4 powers and be 2 integral number power.
3. the method for claim 1, the points N of carrying out FFT is 2 integral number power.
4. the method for claim 1, the M that counts that carries out CZT satisfies N+M-1=q * 2
n, wherein q and n are integers.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103034472A CN102355305A (en) | 2011-10-10 | 2011-10-10 | Linear chirp z transform based frequency offset estimation algorithm in M-QAM (M-ary Quadrature Amplitude Modulation) coherent optical communication system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103034472A CN102355305A (en) | 2011-10-10 | 2011-10-10 | Linear chirp z transform based frequency offset estimation algorithm in M-QAM (M-ary Quadrature Amplitude Modulation) coherent optical communication system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102355305A true CN102355305A (en) | 2012-02-15 |
Family
ID=45578807
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011103034472A Pending CN102355305A (en) | 2011-10-10 | 2011-10-10 | Linear chirp z transform based frequency offset estimation algorithm in M-QAM (M-ary Quadrature Amplitude Modulation) coherent optical communication system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102355305A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102611520A (en) * | 2012-02-23 | 2012-07-25 | 中兴通讯股份有限公司 | Wavelength tag information analysis device and analysis method |
| CN102833204A (en) * | 2012-08-24 | 2012-12-19 | 西安空间无线电技术研究所 | Frequency offset estimation implementation method |
| CN105406921A (en) * | 2015-10-20 | 2016-03-16 | 中国电子科技集团公司第三研究所 | MFSK water sound communication method and MFSK water sound communication system |
| CN105891600A (en) * | 2015-01-23 | 2016-08-24 | 北京空间飞行器总体设计部 | Quaternary phase shift keying signal spectrum estimation method |
| CN106101044A (en) * | 2016-06-01 | 2016-11-09 | 中国电力科学研究院 | A kind of frequency deviation estimating method based on multistep interpolation for coherent optical communication system |
| CN108983155A (en) * | 2018-07-09 | 2018-12-11 | 重庆大学 | A kind of radar-communication integration waveform design method |
| CN114128154A (en) * | 2019-06-25 | 2022-03-01 | 波尔多大学 | Method for generating a signal comprising a time-sequential chirp, method for estimating a vehicle symbol using this signal, computer program product and corresponding devices |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5500878A (en) * | 1993-06-08 | 1996-03-19 | Nec Corporation | Automatic frequency control apparatus and method therefor |
| CN1520081A (en) * | 2003-01-23 | 2004-08-11 | 上海贝尔阿尔卡特股份有限公司 | Method and equipment for carrier frequency synchronization |
| CN1937607A (en) * | 2006-10-17 | 2007-03-28 | 北京邮电大学 | Quick, high-precision doppler frequency deviation estimating method |
| CN101222473A (en) * | 2008-02-01 | 2008-07-16 | 北京天碁科技有限公司 | Frequency deviation estimation method and device |
-
2011
- 2011-10-10 CN CN2011103034472A patent/CN102355305A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5500878A (en) * | 1993-06-08 | 1996-03-19 | Nec Corporation | Automatic frequency control apparatus and method therefor |
| CN1520081A (en) * | 2003-01-23 | 2004-08-11 | 上海贝尔阿尔卡特股份有限公司 | Method and equipment for carrier frequency synchronization |
| CN1937607A (en) * | 2006-10-17 | 2007-03-28 | 北京邮电大学 | Quick, high-precision doppler frequency deviation estimating method |
| CN101222473A (en) * | 2008-02-01 | 2008-07-16 | 北京天碁科技有限公司 | Frequency deviation estimation method and device |
Non-Patent Citations (3)
| Title |
|---|
| FENG-XIANG GE,ETAL.: "Iterative CZT-Based Frequency Offset Estimation for Frequency-Selective Channels", 《COMMUNICATIONS,2005.ICC 2005.2005 IEEE INTERNATIONAL CONFERENCE ON (VOLUME:4)》 * |
| MEHREZ SELMI,ETAL.: "Accurate Digital Frequency Offset Estimator for Coherent PolMux QAM Transmission Systems", 《OPTICAL COMMUNICATION,2009.ECOC 09.35TH EUROPEAN CONFERENCE ON》 * |
| TADAO NAKAGAWA,ETAL.: "Non-Data-Aided Wide-Range Frequency Offset Estimator for QAM Optical Coherent Receivers", 《OPTICAL FIBER COMMUNICATION CONFERENCE AND EXPOSITION (OFC/NFOEC), 2011 AND THE NATIONAL FIBER OPTIC ENGINEERS CONFERENCE》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102611520A (en) * | 2012-02-23 | 2012-07-25 | 中兴通讯股份有限公司 | Wavelength tag information analysis device and analysis method |
| CN102611520B (en) * | 2012-02-23 | 2017-07-21 | 南京中兴新软件有限责任公司 | A kind of wavelength tag information analysis apparatus and analysis method |
| CN102833204A (en) * | 2012-08-24 | 2012-12-19 | 西安空间无线电技术研究所 | Frequency offset estimation implementation method |
| CN102833204B (en) * | 2012-08-24 | 2015-02-11 | 西安空间无线电技术研究所 | Frequency offset estimation implementation method |
| CN105891600A (en) * | 2015-01-23 | 2016-08-24 | 北京空间飞行器总体设计部 | Quaternary phase shift keying signal spectrum estimation method |
| CN105891600B (en) * | 2015-01-23 | 2019-05-21 | 北京空间飞行器总体设计部 | Four phase shift keying signal spectrum estimation method |
| CN105406921A (en) * | 2015-10-20 | 2016-03-16 | 中国电子科技集团公司第三研究所 | MFSK water sound communication method and MFSK water sound communication system |
| CN106101044A (en) * | 2016-06-01 | 2016-11-09 | 中国电力科学研究院 | A kind of frequency deviation estimating method based on multistep interpolation for coherent optical communication system |
| CN106101044B (en) * | 2016-06-01 | 2019-04-30 | 中国电力科学研究院 | A kind of frequency deviation estimating method based on multistep interpolation for coherent optical communication system |
| CN108983155A (en) * | 2018-07-09 | 2018-12-11 | 重庆大学 | A kind of radar-communication integration waveform design method |
| CN114128154A (en) * | 2019-06-25 | 2022-03-01 | 波尔多大学 | Method for generating a signal comprising a time-sequential chirp, method for estimating a vehicle symbol using this signal, computer program product and corresponding devices |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102355305A (en) | Linear chirp z transform based frequency offset estimation algorithm in M-QAM (M-ary Quadrature Amplitude Modulation) coherent optical communication system | |
| EP2584720B1 (en) | Frequency offset estimation apparatus, receiver apparatus, frequency offset estimation method, and reception method | |
| CN104393917B (en) | Polarization state rapid tracking monitoring method based on Kalman filtering | |
| CN104022981B (en) | A kind of blind carrier frequency bias estimation of quadrature amplitude modulation signal | |
| US8908809B2 (en) | Complexity reduced feed forward carrier recovery methods for M-QAM modulation formats | |
| US8374512B2 (en) | Frequency offset monitoring device and optical coherent receiver | |
| CN101553028B (en) | Frequency offset and phase estimation method based on differential phase in TD-SCDMA communication system receiving synchronization | |
| US8594241B2 (en) | Estimation and compensation method for IQ imbalance | |
| US8189718B2 (en) | Adaptive optimization method and apparatus for average length | |
| EP2656566B1 (en) | Automatic frequency offset compensation in zero-intermediate frequency receivers using minimum-shift keying (msk) signaling | |
| US20150180691A2 (en) | Phase detector | |
| CN108270715A (en) | It is suitble to the carrier recovery system and method for high-order 4096-QAM | |
| US8682182B2 (en) | Blind carrier frequency offset detection for coherent receivers using quadrature amplitude modulation formats | |
| CN103748846B (en) | Frequency deviation estimating method and module in digital communication system | |
| CN106059984A (en) | Digital phase-modulated signal carrier phase offset estimation method | |
| CN108847895A (en) | A kind of blind phase noise compensation method suitable for C-mQAM coherent optical communication system | |
| CN109067680A (en) | A kind of carrier frequency bias estimation and its device of baseband signal | |
| CN104052693B (en) | Frequency error estimation unit and method, frequency error compensation device and optical receiver | |
| CN114531329A (en) | Method, system and application for estimating carrier frequency of multi-path MSK (minimum shift keying) signals | |
| CN105933265B (en) | A kind of pair of QAM signal carries out the phase noise blind estimating method of unbound nucleus | |
| CN116545824A (en) | Frequency offset estimation method, device and receiver | |
| CN109660481A (en) | A kind of quadrature amplitude modulation demodulation method and its planisphere optimization method | |
| CN105072070B (en) | A kind of digital broadband wireless signal synchronized algorithm | |
| US8295402B2 (en) | Optimal blind channel estimation for DQPSK demodulation | |
| US7844011B2 (en) | Apparatus and method for improving symbol error rate performance of M-ary phase shift keying system having quadrature error |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20120215 |