Advanced Orthogonal Frequency Division Multiplexing Differential Chaos Shift Keying Systems in the Presence of Noise and Fading

Abstract

Wireless communication, being inherently susceptible to malicious attacks due to the broadcast nature of its channels, requires enhanced security measures. A rising trend in enhancing the security of wireless communication systems is the adoption of chaotic communication technologies. These technologies use chaotic signals, which have inherent security properties such as non-periodicity, noise-like behavior, and sensitivity to initial values, making them ideal for enhancing the security of information transmission.

The application of chaotic signals proves to be highly beneficial in chaos shift keying modulation technologies, wherein they serve to encrypt the transmitted information symbols. Differential chaos shift keying (DCSK) stands out as an extensively studied modulation scheme within this domain. However, DCSK systems have limitations such as low spectrum efficiency and the need for impractical delay line circuits.

To overcome these limitations, researchers have introduced the multi-carrier differential chaos shift keying (MC-DCSK) scheme. This scheme not only improves the transmission rate of DCSK but also eliminates the requirement for complex delay line circuits. In addition, the MC-DCSK scheme can be further enhanced by integrating the orthogonal frequency division multiplexing (OFDM) technique. The resulting scheme, known as orthogonal frequency division multiplexing differential chaos shift keying (OFDM-DCSK), demonstrates high practical utility.

However, the multi-carrier architecture intrinsic to OFDM-DCSK renders it vulnerable to noise interference. This vulnerability is particularly pronounced in power line communication (PLC) systems, notorious for frequent encounters with impulsive noise. Here, OFDM-DCSK signals face large-amplitude outliers, leading to significant performance declines. Additionally, OFDM-DCSK systems exhibit poor resistance performance to fading, particularly in doubly selective fading channels.

This thesis addresses above-mentioned issues of OFDM-DCSK systems, presenting enhancement strategies and novel system designs to combat various noise environments and fading channels. Firstly, the rank-1 characteristic of the transmitted noise-free symbol matrix is leveraged, leading to a reduction in the dimensionality of time-domain data symbols for the purpose of noise mitigation. This is followed by chaotic demodulation on these symbols to decode information bits. In scenarios with additive white Gaussian noise (AWGN) channel, rank-1 matrix approximation is attained via truncated singular value decomposition (SVD), aligning with the solution of ℓ₂-norm minimization. Whereas, in the context of impulsive noise conditions, an algorithm is formulated employing alternating optimization for ℓ_p-norm based matrix factorization, with 0<p<2.

Next, we devise a sparse-coded OFDM-DCSK system based on rank-1 matrix recovery, adept at managing Gaussian background noise and outlier-contaminated symbols concurrently. By leveraging the rank-1 attribute of the noise-free OFDM-DCSK symbol matrix, the vector outer product is employed for signal modeling, alongside sparse coding to reduce transmission energy. The demodulation of information bits from the sparse-coded signal involves the construction of an objective function, composed of a sum of Frobenius norm for mitigating the Gaussian background noise, and ℓ₀-norm for identifying outlier-contaminated symbols.

Lastly, we propose a singular-vector pre-coding strategy to enhance the reliability of OFDM-DCSK systems over doubly selective fading channels. The channel is partitioned into distinct sub-channels using SVD, leading to a higher signal-to-noise ratio and lower bit error rate (BER). Inspired by the water-filling principle, a subsequent step of symbol permutation at the transmitter, where symbols with higher magnitudes are allocated to the sub-channel with larger gain, contributes to an additional reduction in the BER.

Theoretical analysis and simulations demonstrate that the proposed systems outperform conventional OFDM-DCSK methods in AWGN and impulsive noise environments, including Middleton class A and α-stable noise channels, and achieve higher energy efficiency. Moreover, lower BERs over doubly selective fading channels are achieved.

Date of Award	28 Jun 2024
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Hing Cheung SO (Supervisor)