Design and analysis of chaos-based communication systems

Abstract

The introduction of the IoT paradigm for shaping the future 5G network, as billions of new devices would be deployed with different needs in terms of wireless access, is one of the cornerstones in establishing the 5G standard. A large chunk of the IoT network consists of the WSN and LOWPAN devices, known as devices with limited computation and battery life resources. The new standard for IoT devices, i.e., 802.15.4aa specifies several physical layer methods of access for the Low-Rate Wireless Networks using modulations like DSSS, FSK, MPSK, and UWB. The more significant proportion still deploys the DSSS modulation for WSNs, given its excellent performance in terms of security and its simple implementation. The chaos-based modulations offer better performances in comparison with the DSSS and were considered in the literature, given the excellent statistical properties of the chaotic signals. Hence, they were considered an alternative, especially for devices with low capabilities widely seen in WSNs and LOWPANs. The non-coherent chaos-based scheme specifically offers an excellent solution with its simple detection method that enables successful transmission even through very noisy channels and multipath fading. In the literature, the BER performance of non-coherent chaos-based schemes lacks a universal approach in its establishment, given that every scheme is studied apart and its BER performance is computed using Monte Carlo simulations or through the error function for each model. Moreover, the inherent limitation of non-coherent chaos-based systems in terms of data rate and spectral efficiency has been tackled by introducing countless modulation schemes. However, most of the schemes proposed in the literature require advanced signal processing blocks to implement such designs. In light of this, the issue of BER computation is tackled in the second chapter of this thesis, where the BER expressions of several non-coherent chaos-based schemes over the multipath Rayleigh fading channel are studied, and a new metric in the form of a lower bound is introduced. For this purpose, a transformation is applied to the BER expression of non-coherent chaos-based schemes to replace the error function with an equivalent lower bound to the exponential function multiplied by the PDF of the fading channel, and then an anti-derivative is calculated. This yields a true theoretical expression that can be used to accurately compute the BER values over multipath Rayleigh fading channels for different chaos-based systems As a follow-up, the third chapter extends the work done in chapter 2 to a more generalized channel model, namely the multipath Nakagami-m fading channel. The same methodology is applied to the non-coherent chaos-based models in order to derive a closed-form solution of the BER expression. The newly established lower bound to the BER is then computed and compared to the simulation results and proven to deliver a close match with tiny performance gaps. The fourth chapter of the thesis is dedicated to the improvement of the spectral efficiency of the basic DCSK design. Since the DCSK scheme suffers from the transmission of a reference signal that actively occupies half of the transmission time, we opt for a solution that can shorten such time. We combine the DCSK design with the FTN filtering to reduce the spacing between the transmitted symbols and as a result obtain a gain in terms of spectrum usage. The performances of the newly introduced design are evaluated for different channel models and system settings. In addition, the system performances are compared to state-of-the-art designs in terms of BER and spectral efficiency.

Date	20 Dec 2022
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Georges Kaddoum (Supervisor)