Fluid antenna systems (FAS) signify a pivotal advancement in 6G communication by enhancing spectral efficiency and robustness. However, obtaining accurate channel state information (CSI) in FAS poses challenges due to its complex physical structure. Traditional methods, such as pilot-based interpolation and compressive sensing, are not only computationally intensive but also lack adaptability. Current extrapolation techniques relying on rigid parametric models do not accommodate the dynamic environment of FAS, while data-driven deep learning approaches demand extensive training and are vulnerable to noise and hardware imperfections. To address these challenges, this paper introduces a novel self-supervised learning network (SSNet) designed for efficient and adaptive channel extrapolation in FAS. We formulate the problem of channel extrapolation in FAS as an image reconstruction task. Here, a limited number of unmasked pixels (representing the known CSI of the selected ports) are used to extrapolate the masked pixels (the CSI of unselected ports). SSNet capitalizes on the intrinsic structure of FAS channels, learning generalized representations from raw CSI data, thus reducing dependency on large labeled datasets. For enhanced feature extraction and noise resilience, we propose a mix-of-expert (MoE) module. In this setup, multiple feedforward neural networks (FFNs) operate in parallel. The outputs of the MoE module are combined using a weighted sum, determined by a gating function that computes the weights of each FFN using a softmax function. Extensive simulations validate the superiority of the proposed model. Results indicate that SSNet significantly outperforms benchmark models, such as AGMAE and long short-term memory (LSTM) networks by using a much smaller labeled dataset. A key observation is that the proposed model is more effectively trained using a small unmasked ratio of known CSI. Specifically, the proposed SSNet trained using CSI of 10 % total ports outperforms that trained using CSI of 25 % and 50 % total ports. This is because using a smaller number of known CSIs during training, the proposed model is forced to learn more effective channel correlation for channel extrapolation at the expense of higher training complexities. Ablation experiments reveal substantial performance gains from the MoE module’s integration. Furthermore, zero-shot learning experiments show a moderate performance degradation of about 3-5 dB, underscoring the model’s robust generalization ability. Finally, the inference speed experiments illustrate that the proposed model outperforms the benchmark models dramatically at the expense of a slightly longer execution time of 1.13 ms, 2.9 ms, and 3.12 ms on NVIDIA RTX 4090, 4060, and 3060 graphics processing units (GPU)s, respectively.

Orthogonal Time Frequency Space (OTFS) modulation is expected to address the performance degradation of orthogonal frequency division multiplexing (OFDM) modulated signals, particularly due to issues like Doppler shifts in mobile communication environments. In this paper, we propose a lightweight deep learning-based framework for end-to-end joint channel estimation and data detection (JCEDD) in an OTFS communication system. To fully exploit the characteristics of OTFS modulation, we introduce a data padding preprocessing method and a slicing data augmentation technique. Furthermore, the performance of the proposed deep learning-based framework could be enhanced dramatically with only a small overhead compared to the superimposed pilot scheme. Ablation experiments demonstrate that the proposed data padding preprocessing method and the slicing data augmentation technique significantly improve the performance of the deep learning-based framework. Simulation results show that the proposed framework outperforms existing algorithms in terms of JCEDD performance, while maintaining a relatively low level of computational complexity.

As modern networks face increasing demands for superior service and management, Encrypted Traffic Classification (ETC) technology has become increasingly crucial. Considering that traffic data is easy to collect but hard to label, self-supervised ETC methods have attracted more and more attention. Compared to popular methods based on traffic images and text, traffic tables are simple to construct and more suitable for the flow-packet structure. However, existing methods have two problems: (1) The lack of data augmentation methods for tables weakens the performance of self-supervised learning. (2) Most methods only focus on single feature and cannot make full use of distinct features of traffic tables, such as temporal feature. To solve these problems, we propose a multi-feature fusion method based self-supervised learning approach for traffic tables. A new data augmentation method called Random Subsets Selection (RSS) is introduced alongside an effective fusion approach. In this way, temporal features can be successfully extracted and concatenated with the latent representations of input traffic tables. Experimental results from two open datasets and one self-collected dataset have shown that on imbalanced datasets, our method can effectively solve ETC problems even with a small number of labeled data. Empirically, both classification performance and processing speed are improved. Specifically, compared to the state-of-the-art tabular self-supervised learning method, our method achieves the better classification results on all datasets while the processing speed increases by almost two times, from 1.83 tables per second to 3.76 tables per second.

The emergence of 6th generation (6G) mobile networks brings new challenges in supporting high-mobility communications, particularly in addressing the issue of channel aging. While existing channel prediction methods offer improved accuracy at the expense of increased computational complexity, limiting their practical application in mobile networks. To address these challenges, we present LinFormer, an innovative channel prediction framework based on a scalable, all-linear, encoder-only Transformer model. Our approach, inspired by natural language processing (NLP) models such as BERT, adapts an encoder-only architecture specifically for channel prediction tasks. We propose replacing the computationally intensive attention mechanism commonly used in Transformers with a time-aware multi-layer perceptron (TMLP), significantly reducing computational demands. The inherent time awareness of TMLP module makes it particularly suitable for channel prediction tasks. We enhance LinFormer’s training process by employing a weighted mean squared error loss (WMSELoss) function and data augmentation techniques, leveraging larger, readily available communication datasets. Our approach achieves a substantial reduction in computational complexity while maintaining high prediction accuracy, making it more suitable for deployment in cost-effective base stations (BS). Comprehensive experiments using both simulated and measured data demonstrate that LinFormer outperforms existing methods across various mobility scenarios, offering a promising solution for future wireless communication systems.

The orthogonal time frequency space (OTFS) modulation has garnered significant attention due to its potential to combat the frequency Doppler effect in high-mobility scenarios, especially in the low earth orbit (LEO) satellite communication. However, facilitating the OTFS in multiple-input and multipleoutput (MIMO) satellite communication requires accurate channel state information, which is a challenging task. To this end, we propose a novel deep-learning-based framework to enhance the channel estimation accuracy in a MIMO satellite communication network by exploiting the channel correlation of the OTFS-MIMO channels. Through extensive simulations, our proposed framework shows an impressive capability to predict MIMO-OTFS channels with remarkable accuracy enhancement. The effect of dataset and data distribution on the channel estimation accuracy and generalization are also revealed.

In this paper, a novel DRL-based model (VWP, VAE-WGAN-PPOE) is proposed to solve the problem of long training time and unsatisfactory training effect in the end-to-end autonomous driving. The model is optimized from feature extraction and algorithm decision. In feature extraction, we encode the input video by combining variational auto encoder (VAE) with wasserstein generative adversarial network (WGAN). The state dimension is reduced and the problem of mode collapse and gradient disappearance caused by generative adversarial network (GAN) training is solved. In decision algorithm, we formulate a new reward function by analyzing the factors affecting driving performance. Furthermore, we propose an enhanced algorithm PPOE based on the proximal policy optimization (PPO). In the CARLA simulator, compared with CNN and ResNet34, the convergence speed of the DRL model based on VAE-WGAN increases by 26.1% and 20.3%, the navigation task completion rate increases by 18.5% and 9.2%, and the collision rate decreases by 13.6% and 9.4%. Compared with deep deterministic policy gradient (DDPG) decision algorithm, the convergence speed of the DRL model based on PPOE increases by 23.3%, the navigation task completion rate increases by 5.0% in sunny days and 8.4% in severe weather, the collision rate decreases by 3.5% in sunny days and 6.6% in severe weather. Extensive experiments show that the proposed model enables the agent to drive safely along the navigational route in the complex environment with pedestrian and vehicle interaction, even in severe weather.

Machine analysis of underwater images is essential to most underwater applications. However, both the limitation of communication bandwidth and underwater degradation bring much difficulty to accurate machine recognition at the end system. Few existing underwater compression methods consider unique underwater prior knowledge to better serve for machine vision under low bit-rates. To address this problem, we propose a novel underwater image compression framework for machines, which utilizes underwater priors to contrastively enhance degraded features by contrastive learning and efficiently compress machine-friendly features under low bit-rates. A dataset is built to provide positive and negative samples for contrastive learning based on machine analysis performance. At the encoder side, a feature extractor and a feature encoder are employed to extract machine-related features and compress them into compact representations. To alleviate the effect of underwater degradation on machine vision, a prior-guided contrastive feature enhancement module is proposed to learn more machine-friendly features based on positive and negative samples from our dataset. Then a feature refinement block is designed to remove channel-wise redundancy and focus spatial-wise importance based on high similarity of machine-related features and characteristics of underwater images. More compact representations are obtained without degrading analysis performance under low bit-rates. At the decoder side, both machine-friendly features and image are reconstructed to support different types of analysis tasks. Experimental results demonstrate the superiority of our framework in machine vision tasks compared with traditional compression methods and learned-based methods. Besides, our method still preserves basic capability of human perception.

Underwater enhanced images (UEIs) are affected by not only the color cast and haze effect due to light attenuation and scattering, but also the over-enhancement and texture distortion caused by enhancement algorithms. However, existing underwater image quality assessment (UIQA) methods mainly focus on the inherent distortion caused by underwater optical imaging, and ignore the widespread artificial distortion, which leads to poor performance in evaluating UEIs. In this paper, a novel mapping-based underwater image quality representation is proposed. We divide underwater enhanced images into different domains and utilize a feature vector to measure the distance from the raw image domain to each enhanced image domain. The length and direction of the vector are defined as the enhancement degree and enhancement direction of the image. We construct a best enhancement direction and map other vectors to this direction to obtain the corresponding quality representation. Based on this, a novel network, called generation-based joint luminance-chrominance underwater image quality evaluation (GLCQE), is proposed, which is mainly divided into three parts: bi-directional reference generation module (BRGM), chromatic distortion evaluation network (CDEN), and sharpness distortion evaluation network (SDEN). BRGM is designed to generate two reference images about the unenhanced and the optimal enhanced versions of input UEI. In addition, the distortions in the luminance and chrominance domains of the UEI are analyzed. The luminance and chrominance channels of images are separated and input to SDEN and CDEN respectively to detect different distortions. A multi-scale feature mapping module is proposed in CDEN and SDEN to extract the feature representation of quality in chrominance and luminance of these images respectively. Moreover, a parallel spatial attention module is designed to focus on distortions in structural space by utilizing the different receptive fields of the convolution layer, due to the diverse manifestations of structural loss in the image. Finally, the mapped features extracted by two collaborative networks help the model evaluate the quality of underwater images more accurately. Extensive experiments demonstrate the superiority of our model against other representative state-of-the-art models.

High data throughput is desired in the wireless communication system design. Rate adaptation is an efficient way to update the data rate in the dynamic wireless environment. Conventional rate adaptation algorithms rely on the feedback of acknowledgment/negative acknowledgment (ACK/NACK) messages or signal to noise ratio (SNR). Existing rate adaptation algorithms can not achieve satisfactory transmission rates in time-varying environments. In this paper, we model the rate selection problem as a multi-armed bandit (MAB) problem and propose an online learning rate adaptation algorithm that learns the channel status from both RSSI and ACK/NACK signals. Compared with existing rate adaptation algorithms, the proposed algorithm can adapt to the time-varying channel better and achieve near-optimal transmission rate performance.

Device‐to‐device (D2D) communications are widely seen as an efficient network capacity scaling technology. The co‐existence of D2D with conventional cellular (CC) transmissions causes unwanted interference. Existing techniques have focused on improving the throughput of D2D communications by optimising the radio‐resource management and power allocation. However, very little is understood about the impact of the route selection of the users and how optimal routing can reduce interference and improve the overall network capacity. In fact, traditional wisdom indicates that minimising the number of hops or the total path distance is preferable. Yet, when interference is considered, the authors show that this is not the case. In this study, they show that by understanding the location of the user an interference‐aware‐routing (IAR) algorithm can be devised. They propose an adaptive IAR algorithm that on average achieves a 30% increase in hop distance, but can improve the overall network capacity by 50% whilst only incurring a minor 2% degradation to the CC capacity. The analysis framework and the results open up new avenues of research in location‐dependent optimisation in wireless systems, which is particularly important for increasingly dense and semantic‐aware deployments.

The Clifford Fourier transform (CFT) can be applied to both vector and scalar fields. However, due to problems with big data, CFT is not efficient, because the algorithm is calculated in each semaphore. The sparse fast Fourier transform (sFFT) theory deals with the big data problem by using input data selectively. This has inspired us to create a new algorithm called sparse fast CFT (SFCFT), which can greatly improve the computing performance in scalar and vector fields. The experiments are implemented using the scalar field and grayscale and color images, and the results are compared with those using FFT, CFT, and sFFT. The results demonstrate that SFCFT can effectively improve the performance of multivector signal processing.

Energy‐efficient wireless communication is important for wireless devices with a limited battery life and cannot be recharged. In this study, a bit allocation algorithm to minimise the total energy consumption for transmitting a bit successfully is proposed for a two‐way orthogonal frequency‐division multiplexing relay system, whilst considering the constraints of quality‐of‐service and total transmit power. Unlike existing bit allocation schemes, which maximise the energy efficiency (EE) by measuring ‘bits‐per‐Joule’ with fixed bidirectional total bit rates constraint and no power limitation, their scheme adapts the bidirectional total bit rates and their allocation on each subcarrier with a total transmit power constraint. To do so, they propose an idea to decompose the optimisation problem. The problem is solved in two general steps. The first step allocates the bit rates on each subcarrier when the total bit rate of each user is fixed. In the second step, the Lagrangian multipliers are used as the optimisation variants, and the dimension of the variant optimisation is reduced from 2 N to 2, where N is the number of subcarriers. They also prove that the optimal point is on the bounds of the feasible region, thus the optimal solution could be searched through the bounds.

Network lifetime (NL) is a critical metric in the design of energy-constrained wireless sensor networks (WSNs). In this paper, we investigate a joint optimal design of the physical, medium access control (MAC) and routing layers to maximize NL of a multiple-sources and single-sink (MSSS) WSN with energy constraints. The problem of NL maximization (NLM) can be formulated as a mixed integer-convex optimization problem with adoption of time division multiple access (TDMA) technique. When the integer constraints are relaxed to take real values, the problem can be transformed into a convex problem and the solution achieves the upper bounds. We provide an analytical framework for the relaxed NLM problem of a WSN in general planar topology. We first restrict the topologies to the planar networks on a small scale, including triangle and regular quadrangle topologies. In this special case, we employ the Karush-Kuhn-Tucker (KKT) optimality conditions to derive analytical expressions of the globally optimal NL, which take the influence of data rate, link access and routing into account. To handle larger scale planar networks, an iterative algorithm is proposed using the D&C approach. Numerical results illustrate that the proposed algorithm can be extended to the large planar case and its performance is close to globally optimal performance.