The invention relates to optical communication, and, more particularly, to a full range offset correction technique for coherent optical orthogonal frequency-division multiplexing OFDM Systems.
Orthogonal frequency division multiplexing (OFDM) is a prevailing multiplexing technique in digital mobile communication systems and starts to rise in optical communication systems. It provides flexible multiple access and multiplexing/demultiplexing ability in a low-cost and spectral efficient way. In the OFDM modulation, the data symbols are divided into multiple orthogonal narrow-bandwidth subcarriers and transmitted simultaneously. Due to its feasibility of digital dispersion compensation and narrow-bandwidth subcarriers, it can be very resistible to multi-path fading and chromatic dispersion. This makes it outstand among other modulation schemes.
However, the detection can encounter dramatic difficulty if the frequency of the local oscillator (LO) has an offset from the incoming signal, which may come from Doppler effects, channel fading, or the instability of the frequency synthesizers in transmitters or receivers. The tolerance of the frequency offset is much smaller than the subcarrier spacing, makes it difficult to implement in the environment of large frequency offset. For example, in the coherent optical OFDM (CO-OFDM) fiber communication system, the local laser, as a LO, can be shifted by a few GHz due to temperature variation. Fortunately, the detection problem caused by frequency offset can be perfectly corrected if the frequency offset amount is precisely known.
Accordingly, there is a need for a quick wide searching range frequency offset estimator coherent optical OFDM systems.
A method includes synchronizing a received signal with at least two orthogonal frequency division multiplexed OFDM training signals having only in-phase values and being real in the time domain and determining a frequency offset correction from the synchronization of the received signal and training symbols responsive to a cross-correlation between said training symbols to enable estimating all possible frequency offsets for correction for enabling OFDM demodulation of said received signal.
These and other advantages of the invention will be apparent to those of ordinary skill in the art by reference to the following detailed description and the accompanying drawings.
The present invention is directed to a method that can estimate all possible frequency offset in an OFDM system by two identical training symbols, without complicated computations such as iterative methods or matrix inversion, and without reducing the training symbol size, which may result in the system being more sensitive to noise. Without reduction of the OFDM training symbol size, the inventive method employs a special designed training symbol whose values are all real in time domain. Then a cross-correlation relationship of the two identical training symbols can estimate all possible frequency offsets for correction.
To estimate the frequency offset two identical training symbols [R1 R2 . . . RL]n are sent sequentially, as shown in
r1n=[R1e−j2πΔfTs R2e−j2πΔf×2Ts . . . RLe−j2πΔf×LTs]n
r2n=[R1e−j2πΔf(N+1)Ts R2e−j2πΔf×(N+2)Ts . . . RLe−j2πΔf×(N+L)Ts]n
Then the frequency offset is estimated by
Then, the range of the frequency estimation is limited to −1/(NTs) to 1/(NTs), which is equivalent to few subcarrier spacings. The range can be improved by decreasing the size of an OFDM training frame N, but a smaller OFDM training size will result in less information to average out the impairments of noise, and worse equalization capability of the training symbol.
Turning now to
Referring to
where D is the search range. The output of the cross correlation is used to make a frequency offset Δ{circumflex over (f)} estimation 303 according to
The frequency offset correction {circumflex over (r)}n 304 is then {circumflex over (r)}n=r(NTs)×exp(j2πΔfNTs). After frequency offset correction is applied to the received signal r(t), the OFDM demodulation 305 noted with respect to
The two training symbols r1n and r2n are designed such that they originally only have in-phase values. Then after the calculation of cross-correlation, the phase shift of the received signal r(t) is removed by the complex conjugate noted in
The advantage of the invention is that it can estimate all possible frequency offset in OFDM system by two identical training symbols, without complicated computation such as iterative methods or matrix inversion, and without reducing the training symbol size, which may result in the system more sensitive to noise. The searching range can be adjusted by changing the variable D in
While an exemplary drawing and specific embodiment of the present invention has been described and illustrated, it is to be understood that that the scope of the present invention is not to be limited to the particular embodiments discussed. Thus, the embodiments shall be regarded as illustrative rather than restrictive, and it should be understood that variations may be made in those embodiments by workers skilled in the arts without departing from the scope of the present invention as set forth in the claims that follow and their structural and functional equivalents.
This application claims the benefit of U.S. Provisional Application No. 61/227,459, filed on Jul. 22, 2009, entitled, “Full Range Offset Correction Technique for Coherent Optical OFDM Systems”, the contents of which is hereby incorporated by reference.
Number | Name | Date | Kind |
---|---|---|---|
20040141457 | Seo et al. | Jul 2004 | A1 |
Number | Date | Country | |
---|---|---|---|
20110075764 A1 | Mar 2011 | US |
Number | Date | Country | |
---|---|---|---|
61227459 | Jul 2009 | US |