الفهرس 
Literature SurveyOnly 14 pages are availabe for public view
 |  | from | 180 |  |  |
| | | from | 180 | | |
Abstract
In
exible and adaptive optical communication systems, e.g.,
elastic optical networks, the modulation format is adapted according
to each user demand based on the channel state information.
Therefore, at the receiver side, it is required to blindly identify the
modulation format of the received signals with high accuracy, which
enables the DSP algorithm embedded at the receiver to be recon-
gured, accordingly. This, in turn, removes the need for sending
end-to-end handshaking information between the transmitter and
the receiver.
In this thesis, two schemes for blind optical modulation format
identi cation (MFI), based on the singular value decomposition
(SVD) and Radon transform (RT) of the constellation diagrams, are
proposed and applied in two di erent domains. In the rst domain,
Jones space domain, constellation diagrams are obtained at optical
signal-to-noise ratios (OSNRs) ranging from 2 to 30 dB for eight
di erent modulation formats as images. The rst scheme depends
on the utilization of feature vectors composed of the singular values
(SVs) of the obtained images, while the second scheme is based on
applying the RT and then getting the SVs.
Di erent classi ers are used and compared for the MFI task. The
e ect of varying the number of samples on the accuracy of the classi
ers is studied for each modulation format. Simulation and experimental
setups are provided to study the eciency of the two schemes
at high bit rates for three dual-polarized modulation formats. A
decimation approach for the constellation diagrams is suggested to
reduce the SVD complexity, while maintaining high classi cation accuracy.
The obtained results reveal that the proposed schemes can
be accurately used to identify the optical modulation format blindly
with classi cation rates up to 100%, even at low OSNR values of 10
dBs.
Moreover, a new application of the proposed schemes in the
Stoke Space (SS), which deals with the dual-polarized modulation
formats, is presented. The SS gives a representation of the received
signal in 3 dimensions. In our case, we take the projection of that
representation on three di erent planes and deal with them as representative
images. Then, we implement SVD of these obtained
images and use their SVs as features. In another implementation,
the RTs are obtained from the three di erent projections, and the
SVD is implemented to get the features. Here, the features are
prepared to train the classi ers. The test scenario is performed on
features extracted in the same manner from projections.
The two schemes proposed are tested for random data transmission.
In addition, the applicability of transmitting images on the
optical channel model is studied and veri ed in di erent cases. Different
types of compression schemes such as SVD, PCA and JPEG
are studied and compared. Moreover, two schemes for image encryption:
The double random phase encoding (DRPE) and optical
scanning holography (OSH) are suggested to secure the image communication
process. DRPE is a full encryption scheme and the OSH
is a partial encryption scheme. Moreover, error correction coding is
considered as a feasible approach for reliable image communication