Automatic modulation recognition (AMR) plays a crucial role in non-cooperative communication environment for identifying modulation types of received radio signals. Recently, the achievements of deep learning (DL) have sparked significant interest in applying DL to the field of AMR. However, existing DL-based AMR methods only use image modal or sequence modal as input, which cannot leverage sufficient information of the signal in complicated electromagnetic environment characterized by scarce labeled sample and multipath fading. To overcome this limitation, we explore different modal representations of the signal to fully exploit their complementary information and propose a hybrid modal contrast and fusion method for automatic modulation recognition (HMCF-AMR). It consists of two stages: 1) modal-level feature contrast for self-supervised pre-training and 2) modal-level feature fusion for supervised fine-tuning. In modal-level feature contrast, sequence encoder and image encoder are designed to extract multi-scale features of the modulated signal from the image modal and sequence modal. Meanwhile, a multi-task collaborative pre-training method combining generative and contrastive learning is achieved to enhance and align different modal representations. In modal-level feature fusion, an attentional feature fusion mechanism integrates the features learned from different modal to further improve modulation recognition performance and online learning is implemented by fine-tuning to handle different scenarios. Simulation results show that our proposed HMCF-AMR outperforms other baseline models in both adequate-sample and few-shot scenarios and demonstrates greater robustness in complicated multipath fading channels.

Recent deep learning-based face recognition(FR) methods have demonstrated remarkable performance in high-resolution (HR) or down-sampled low-resolution (LR) tasks. However, these methods often exhibit disappointing speed-accuracy trade-offs when deployed on real-world LR scenarios due to limited model generalization. To this end, we propose a lightweight Face Transformer framework for real-world low-resolution face recognition(LRFR) named LLRFaceFormer. Firstly, we propose a Transformers-as-convolutions (TaC) network using a Transformer layer to replace the matrix multiplication in the standard convolutional process. This hybrid approach combines the strengths of Transformers and CNNs, allowing the TaC network to extract sufficient effective identity information via a global receptive field while adaptively discarding redundant homogeneous identity information on constructed LR faces. Secondly, we propose a Transformer-specific adaptive average procedure that incorporates tensor shape operations and a depthwise(DW) convolution. This procedure enables the LLRFaceFormer framework to focus on different regions of the input images. We also introduce an identity-aware simulator that generates real-world-like blurred LR scenarios during the training process to reduce the distribution discrepancy between the training LR faces and the testing real-world LR faces. The identity-aware simulator is simultaneously trained with the FR network with a cooperative training strategy. Further experiments illustrate the significantly superior speed-accuracy trade-offs over existing LRFR methods with state-of-the-art(SOTA) performance on several LRFR benchmarks.

In this paper, a novel transmissive reconfigurable meta-surface (RMS) transceiver enabled multi-tier computing network architecture is proposed for improving computing capability, decreasing computing delay and reducing base station (BS) deployment cost, in which transmissive RMS equipped with a feed antenna can be regarded as a new type of multi-antenna system. We formulate a total energy consumption minimization problem by a joint optimization of subcarrier allocation, task input bits, time slot allocation, transmit power allocation and RMS transmissive coefficient while taking into account the constraints of communication resources and computing resources. This formulated problem is a non-convex optimization problem due to the high coupling of optimization variables, which is NP-hard to obtain its optimal solution. To address the above challenging problems, block coordinate descent (BCD) technique is employed to decouple the optimization variables to solve the problem. Specifically, the joint optimization problem of subcarrier allocation, task input bits, time slot allocation, transmit power allocation and RMS transmissive coefficient is divided into three subproblems to solve by applying BCD. Then, the decoupled three subproblems are optimized alternately by using successive convex approximation (SCA) and difference-convex (DC) programming until the convergence is achieved. Numerical results verify that our proposed algorithm is superior in reducing total energy consumption compared to other benchmarks.

Abstract Recently, 3D Convolutional Neural Networks (3D CNNs) have attracted extensive attention in extracting spatial and temporal features in videos for their efficient feature extraction ability. However, it also brings enormous model parameters by training very deep 3D CNNs. Here, a novel network named spurious‐3D Residual Attention Networks (S3D RANs) is proposed for video‐based action recognition, which has the powerful capacity to learn collaborative spatiotemporal features. In particular, by leveraging the merits from 2D Convolutional Neural Networks (2D CNNs) and 3D CNNs, 2D CNNs are applied rather than 3D CNNs on frames of the single view of volumetric videos data to learn temporal motion features directly. Furthermore, view and channel‐wise attention mechanism submodules are employed in the residual unit to learn the importance of each view for action recognition and guide the network to pay more attention to the more useful information for action recognition. Experimental results on UCF‐101, HMDB‐51 datasets demonstrate that our S3D RANs have higher accuracy and lower model complexity than existing works.

Developing resource allocation algorithms for the uplink sparse code multiple access (SCMA) scheme to satisfy multiple objectives is challenging, especially where users are randomly distributed. In this paper, we aim to address this challenge by developing a joint resource allocation method as a multi-objective optimization (MO) problem to maximize the average sum rate and fairness among users as key and sub-key objectives, respectively. For this purpose, the exact analytical expressions for the average sum rate and users' individual rate are extracted based on an adaptive weighted graph matrix (AWGM). An AWGM matrix beneficially replaces the factor graph and the power allocation matrices to simplify the MO problem based on the asymmetric modified bipartite matching (AMBM) algorithm. The power allocation strategy is utilized during the optimal resource assignment process using the AMBM algorithm. After the AMBM process, we propose a low-complexity four-step algorithm to obtain the AWGM. The simulation results show that our proposed method can compromise and improve the multiple objectives' performance and guarantees a stable range of network performance at different times.

Heterogeneous network is a novel network architecture proposed in long-term-evolution, which highly increases the capacity and coverage compared with the conventional networks. However, in order to provide the best services, appropriate resource management must be applied. In this paper, we consider the joint optimization problem of user association, subchannel allocation, and power allocation for downlink transmission in multi-cell multi-association orthogonal frequency division multiple access heterogeneous networks. To solve the optimization problem, we first divide it into two subproblems: 1) user association and subchannel allocation for fixed power allocation and 2) power allocation for fixed user association and subchannel allocation. Subsequently, we obtain a locally optimal solution for the joint optimization problem by solving these two subproblems alternately. For the first subproblem, we derive the globally optimal solution based on graph theory. For the second subproblem, we obtain a Karush-Kuhn-Tucker optimal solution by a low complexity algorithm based on the difference of two convex functions approximation method. In addition, the multi-antenna receiver case and the proportional fairness case are also discussed. Simulation results demonstrate that the proposed algorithms can significantly enhance the overall network throughput.

This paper investigates the joint transmitter and receiver optimization for the energy efficiency (EE) in orthogonal frequency-division multiple-access (OFDMA) systems. We first establish a holistic power dissipation model for OFDMA systems, including the transmission power, signal processing power, and circuit power from both the transmitter and receiver sides, while existing works only consider the one side power consumption and also fail to capture the impact of subcarriers and users on the system EE. The EE maximization problem is formulated as a combinatorial fractional problem that is NP-hard. To make it tractable, we transform the problem of fractional form into a subtractive-form one by using the Dinkelbach transformation and then propose a joint optimization method, which leads to the asymptotically optimal solution. To reduce the computational complexity, we decompose the joint optimization into two consecutive steps, where the key idea lies in exploring the inherent fractional structure of the introduced individual EE and the system EE. In addition, we provide a sufficient condition under which our proposed two-step method is optimal. Numerical results demonstrate the effectiveness of proposed methods, and the effect of imperfect channel state information is also characterized.