This paper proposes a robust, energy-efficient multi-user OFDM chaotic vector cyclic shift keying (MU-OFDM-CVCSK) system for channels impaired by multi-user interference (MUI), Gaussian, and impulsive noise. The system’s co-designed transceiver uses cyclic shifts of a single reference signal for energy efficiency, while a novel receiver combines low-rank matrix factorization with a truncated-quadratic loss function to suppress both MUI and noise. In particular, the transmitted signal’s cyclical structure enables iterative reference refinement, further enhancing performance. We provide a comprehensive analysis of the algorithm’s convergence, complexity, energy efficiency, power spectral density, peak-to-average power ratio, and derive the bit error rate (BER) expressions. Simulations demonstrate superior BER performance compared to benchmark schemes in challenging interference and noise conditions.

Xinwei Yang, Zhaofeng Liu, Chen Huang, Jiashuai Zhang, Tong Zhang, Yifan Zhang, Wenqiang Lei. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

In traditional multi-user orthogonal frequency division multiplexing differential chaos shift keying (MU-OFDM-DCSK) systems, the noisy correlation of reference signals with overlapped information-bearing signals from multiple users degrades the bit error rate (BER). This paper devises a MU-OFDM-DCSK receiver that enhances the BER performance for uplink transmission. In our design, the challenges of reducing multi-user interferences (MUIs) and noise in both reference and information-bearing chaotic signals are addressed via leveraging the low-rank structure of the received signal matrix. By exploiting low-rank matrix factorization, a novel objective function is constructed. Minimization of this function, which corresponds to a least squares optimization, can effectively decode the transmitted data bits, even in the presence of additive white Gaussian noise (AWGN) and MUIs. The BER of our scheme is analyzed and verified in both AWGN and multipath Rayleigh fading channels. Theoretical developments including uniqueness of matrix factorization and algorithm convergence as well as complexity, are also provided. Simulation results demonstrate that our designed receiver yields smaller BER than benchmark schemes.

In order to meet increasing demands of higher energy efficiency and better reliability performance for the multiuser Internet of things (IoT) secure communications, in this paper, we design a reliable and energy efficient superposition modulation (SM) and singular value decomposition (SVD) detection based multi-user orthogonal frequency division modulation-based differential chaos shift keying (SM-MU-OFDM-DCSK) transceiver for IoT devices. At the transmitter, we propose to superimpose and overlap the information-bearing chaotic signals for higher energy efficiency. Since the chaotic modulated signals of each user share identical reference chaotic signals, the transmitted superimposed signal matrices have the low-rank property. Then at the receiver, we apply the SVD-aided detection to recover these low-rank signals, which can suppress the noise to attain better reliability performance. Moreover, we prove that the proposed design can achieve the maximum likelihood detection performances. Furthermore, we derive the theoretical bit rate, energy efficiency and bit error rate expressions over additive white Gaussian noise (AWGN) channel. Simulation results are then provided to validate the theoretical derivations. Subsequently, the simulated performances over AWGN and fading channels are investigated to show that higher energy efficiency and better reliability performances than benchmark schemes can be achieved in industrial IoT scenarios.

The task of underwater image enhancement aims to improve the quality of the images and promote the visualization effect. Current methods still face some challenges, which include the difficulty to model the real ocean environment and the lack of simultaneous exploitation and modeling of both global and local information in images. Particularly, in resource-constrained underwater scenarios, models need to enhance images on embedded devices in real-time, which usually requires them to strike a balance between the inference speed and accuracy. In this paper, we put forward a real-time underwater image enhancement network (RUE-Net). Compared to previous advanced models using single branch structure or traditional U-shaped networks, it consists of parallel Receptive Field Enhancement (RFE) Module and Fine Grain Detail (FGD) Module, which perform parallel modeling and fusion of image information at global and local scales. Furthermore, dense connections are established in each stage throughout the whole model. Compared with current models, RUE-Net shows the best performance in the image enhancement task in LSUI, UFO-120, and EUVP datasets. Furthermore, based on the RUE-Net, we introduce a simultaneous enhancement and super-resolution (SESR) multi-task model. The experiments prove RUE-Net has an excellent capability for image restoration and super-resolution reconstruction in UFO-120 and USR-248 datasets. RUE-Net is also deployed on the Jetson Orin Nano, which achieves an impressive inference speed of up to 51 frames per second (FPS), demonstrating its value in underwater robot with limited computing capabilities. The code is available at https://github.com/GuocunWang/RUE-Net.

In this paper, we devise a sparse-coded orthogonal frequency division multiplexing (OFDM) differential chaos shift keying (DCSK) communication system based on low-rank matrix recovery which can handle Gaussian background noise and outlier-contaminated symbols simultaneously. As the noise-free OFDM-DCSK symbol matrix has rank 1, we exploit the vector outer product for its modeling, while sparse coding is also applied to reduce the transmission energy. To demodulate information bits from the sparse-coded signal, we formulate an objective function which consists of a sum of Frobenius norm for rank-1 matrix recovery and$\ell _{0}$-norm for identifying the possibly outlier-contaminated symbols, with a self-adaptive weight parameter. The resultant optimization problem is solved iteratively via block coordinate descent, and the Laplacian kernel with the Silverman’s rule is adopted for outlier detection. Theoretical analysis including convergence of the objective function, bit error rate (BER), energy efficiency and computational complexity, are provided. Simulation results show that the proposed system has comparable mean square error and BER performance with the$\ell _{p}$-norm minimization based matrix recovery approach at$p=2$in additive white Gaussian noise, and is superior to that of$p=1$in Middleton class A noise, even when sparse coding is applied. Moreover, compared with other binary DCSK systems, our system achieves higher energy efficiency thanks to the sparse coding.

In this paper, we propose a novel two-dimensional frequency and time diversity aided orthogonal frequency division multiplexing based differential chaos shift keying (OFDM-DCSK) system. Our aim is to provide secure and robust transmissions for practical wireless systems, where perfect channel state information (CSI) may not be available at the receiver, by exploiting the natural high security of chaotic sequences and frequency diversity gains brought by the frequency hopping (FH). In our design, the information bits are firstly modulated by chaotic chips, then non-repetitive FH operations are performed on both reference chips and chaotic modulated symbols. After the inverse fast Fourier transform (IFFT), the non-repetitive reference chips and chaotic modulated symbols are respectively transmitted over different subcarriers. Subsequently, the receiver recovers the information using the received reference chaotic chips which naturally embed the channel frequency response (CFR) of all subcarriers. We then analyze the energy and spectral efficiencies, derive the bit error rate (BER) and information leakage expressions, and provide a detailed complexity analysis of the proposed scheme. Simulation results verify the effectiveness of our derivations and demonstrate that the proposed system achieves better BER and security performances compared with the benchmark system, especially when the CSI is imperfect or unknown.

In this paper, we present a frequency hopping (FH) orthogonal frequency division multiplexing-aided differential chaos shift keying (OFDM-DCSK) modulation system for cognitive radio (CR) systems operating over non contiguous bands. Our aim is to provide reliable transmissions for CR transceivers which may hardly get the exact and perfect channel state information (CSI). In our design, we utilize the natural high security of chaotic sequences and the frequency diversity brought by the FH. The information bits are firstly modulated by chaotic chips, then both reference chips and chaotic modulated symbols are carried out FH operations and hopped non-repetitively. After performing the inverse fast Fourier transform (IFFT) over available non-contiguous spectrum bands, the resultant OFDM symbols are transmitted and the non-repetitive reference chips are respectively transmitted over different subcarriers. Subsequently, the receiver would retrieve the information using the received reference chaotic chips which naturally embed the channel frequency response (CFR) of all subcarriers. The bit error rate (BER) expressions are derived and simulations results verify the effectiveness of our derivations, which demonstrate that the presented system can achieve better BER than counterpart systems, especially when the CSI is not exact or unknown.

In this paper, we proposed a novel carrier interferometry (CI) code index modulation aided orthogonal frequency division multiplex (OFDM) differential chaotic shift keying (DCSK) communication system. In our design, we utilize the orthogonality property of CI codes matrix, and propose to transmit more information bits using the index of the CI codes matrix. The input information stream is divided into two subsets. One subset is modulated by the chaotic sequence and then spread by the information-bearing CI codes matrix which carries another information subset. At the receiver, the non-coherent chaotic demodulation and the maximum likelihood detection are respectively performed on these two subsets to retrieve the information. Since more bits can be transmitted while not requiring extra chaotic sequences and remaining the same overall energy, the energy efficiency is improved. Additionally, thanks to the spreading in the frequency domain achieved by the usage of CI codes, the peak to average power ratio (PAPR) can be suppressed. Moreover, theoretical performances including the energy efficiency, the bit error rate (BER) expressions are analyzed. Simulation results are provided to validate the theoretical analysis, and demonstrate the satisfactory PAPR and BER performances achieved by our design.