M.Sc Thesis

M.Sc StudentBar Amitay
SubjectChannel Estimation for Sub-Banded Coherent and Self-Coherent
Optical Communication
DepartmentDepartment of Electrical and Computer Engineering
Supervisor PROFESSOR EMERITUS Moshe Nazarathy


Although OFDM-based optical transmission has not yet been commercially deployed, intense research continues in this promising direction. As the OFDM approach continually evolves, improved multiple variants are introduced, such as DFT-spread (DFT-S) OFDM, filter-bank based sub-banded OFDM and combinations thereof. In all OFDM variants, training sequences are used for coarse timing offset (CTO) recovery as well as carrier frequency offset (CFO) recovery, i.e. estimation and correction of the timing window to perform the receiver FFT for OFDM detection, as well as the estimation of the CFO and its subsequent cancellation.

We first introduce an improved CTO and CFO recovery algorithm for OFDM and its variants, based on the so-called Golay Complementary Codes, which are two sequences yielding a discrete delta-sequence upon summing up their aperiodic auto correlations. When used in conjunction with filter-bank based DFT-S OFDM, our CTO estimator is highly tolerant of strong channel impairments such as chromatic dispersion (CD) and polarization mode dispersion. The novel proposed method will be shown to outperform the state-of-the-art scheme for CTO estimation purposes in terms of displaying a distinct single sample peak, as well as provide lower complexity than current delay & correlate techniques. 

CTO estimation methods may also provide an estimation of the CFO, for which there is a tradeoff between estimation accuracy and dynamic range. Our Novel Golay-base method demonstrates a variety of flexible CFO estimation schemes depending on the required accuracy / CFO dynamic range.

Next, we propose to use the recent filter-bank based receivers for robust, accurate, and low-complexity in-service CD estimation impervious to channel impairments. Our novel CD estimation method is not blind but is data-aided by the same training sequences already used for in-service transmission in the sub-banded OFDM receiver. A particular advantage of filter-bank based CD estimation is that the multitude of sub-bands endows the proposed method with diversity resilience, enabling switched combining of the sub-bands. We introduce a simple and robust diversity voting algorithm (DVA) for detecting the outlier sub-bands to be excluded from the CD estimation.

Finally, we conduct a self-coherent transmission experiment for a transmission system, developed in our EE group, which mitigates the pilot fading phenomenon occurs in self-coherent transceivers when transmitting independent data on the two polarizations.