Simultaneous Determination Of Modulation Types And Signal-To-Noise Ratios In Wireless Communication Systems
Loading...
Date
2019-07-01
Authors
Almohamad, Tarik Adnan
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
Signal parameters determination techniques can offer more reliability, larger bandwidth,
and higher security to the modern wireless systems. However, the performance
comparison of such determination techniques is not straightforward. This can be attributed
to the lack of having benchmarks datasets in the wireless communication research
community as compared to other research fields. Hence, there is a need to propose
potentially-benchmark datasets that can be a future choice for researchers in the wireless
communication domain, motivating the first contribution in this thesis. Most of the up-to-date
solutions of signal parameters determination techniques focus on determining only
one single parameter for example modulation type and assuming the other parameters e.g.
signal-to-noise ratio (SNR) are known rather than on determining multiple signal
parameters jointly. Hence, the desire to enable an autonomous capability of simultaneous
determination of multiple signal parameters has motivated the second and third contributions
that tackle the non-coherent and coherent receivers, respectively. Therefore, the objectives
in this thesis are as follows, firstly, to construct three datasets for the aforesaid receivers to
be utilized for joint determination of modulation types and signal-to-noise ratios, namely
asynchronous amplitudes histograms (AAHs)-based Dataset, two-dimensional
asynchronous sampling in-phase-quadrature histograms (2D-ASIQHs)-based Dataset 1
and 2D-ASIQHs-based Dataset 2. Secondly, to develop a scheme that can simultaneously
determine the modulation type and signal-to-noise ratio in non-coherent receivers by using
one features’ type (AAHs-based features) under a multipath fading scenario. Lastly, to
develop a scheme that can simultaneously determine the modulation type and signal-to-noise
ratio in coherent receivers by employing 2D-ASIQHs features under Rician and
Rayleigh fading. For datasets formation, the three datasets are developed under multipath
fading channels and used to validate the proposed determination schemes in the second
and third contributions. The simultaneous determination AAHs-based scheme enables a
generic intelligent receiver to recognize various modulations that belong to different
categories and reveals that the recognition capability can be extendable further to other
more signal types. The scheme can also simultaneously estimate the SNR values
accurately. The simultaneous determination 2D-ASIQHs-based scheme tackles the
coherent receivers to jointly determine nine modulation types and a wide range of SNRs.
The most significant features are extracted using principal component analysis (PCA) and
then fed to a support vector machine (SVM) tool for the automatic learning process. In the
simulation results, the samples of the datasets reflect 1D-envelope histograms and in-phase-quadrature-based images which comprise various modulation types and SNRs with
different combinations of path gains and delays. In term of modulation recognition
accuracy and mean SNR estimation error, the proposed AAHs-based scheme attains
99.83% and 0.79 dB, respectively. Similarly, the proposed 2D-ASIQHs-based scheme
achieves 99.06% and 1.10 dB, respectively. The obtained results showed that the proposed
schemes outperform the existing state-of-the-art work.