Vehicular edge computing (VEC) involves deploying edge servers in vehicles moving along highways, to provide additional computing resources for low latency and high reliability in computational tasks. However, the dynamic nature of VEC environments, particularly regarding the task offloading of mobile vehicles acting as mobile edge servers, requires further research. To address this challenge, we propose a novel delay optimization algorithm based on deep reinforcement learning in VEC scenarios. It considers intelligent vehicles as mobile service nodes, collaborating with fixed roadside unit service nodes to form an integrated task offloading model. This model facilitates computational task offloading through vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications. It accounts for the sojourn time routing planning of vehicles and supports divisible tasks, thereby improving offloading efficiency. Then, an Asynchronous Advantage Actor-Critic (A3C) algorithm is used to achieve an effective computational offloading strategy. Simulation results demonstrate that the algorithm significantly outperforms existing methods in terms of the total system delay, while providing a more efficient and effective solution for task offloading in VEC scenarios.
In the wide-spreading netted ISAC systems, time-division pulse sequences have become essential for augmenting station diversity and facilitating collaborative netted sensing applications. This paper introduces an optimization model for pulse sequence (PS) optimization, designed to meet sensing requirements without compromising communication. The numerical results confirm the effectiveness of this method, enhancing surveillance capabilities while adhering to the intricate constraints of new radio (NR) communication requirements.
In this work, we propose a novel direction-of-arrival (DOA) estimation algorithm for sparse linear array via Vandermonde decomposition reconstruction. Unlike virtual array interpolation algorithms, the suggested method performs interpolation directly on the physical array, that is, Nyquist spatial filling. By utilizing the Vandermonde decomposition of the covariance matrix of a uniform linear array (ULA), this filling process is formulated as a structured matrix completion problem via matrix factorization, where the factor matrix is encouraged to exhibit a Vandermonde structure. Subsequently, an iterative approach is developed to solve the resultant problem using the alternating direction method of multipliers (ADMM). Finally, DOAs are retrieved from the reconstructed covariance matrix using subspace-based algorithms. Simulation results demonstrate the superiority of our algorithm over the existing methods.
This paper proposes a simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS)-assisted symbiotic radio (SR) system, in which the transmission functionality of STAR-RIS is used to transmit extra Internet of Things (IoT) data and the reflection functionality of STAR-RIS is adopted to enhance the downlink transmission. STAR-RIS utilizes ON-OFF keying for IoT data modulation, which allows the IoT receiver to decode data using a low-complexity energy detector. The transmit beamforming at the base station and the transmission and reflection beamforming at the STAR-RIS are jointly optimized to maximize the weighted sum-rate (WSR) of downlink users, subject to the received power constraint of IoT transmission. Considering the practical coupled phase shift constraint of STAR-RIS, an efficient penalty dual decomposition-based algorithm is proposed to solve the formulated highly non-convex problem. Based on the obtained solution, the optimal IoT receiver and the theoretical bit error rate are further derived. Simulation results verify the effectiveness of the proposed STAR-RIS-assisted SR system.
With the development of information technologies such as artificial intelligence, cloud computing, and the Internet of Things, Iot devices play an important role in people's lives. As a hardware carrier for carrying devices and transmitting sensitive information, Internet of Things devices may be out of control and leak sensitive information if they are attacked by networks. Therefore, they have urgent security needs in terms of confidential transmission of device data, dynamic command control, real-time picture sharing, and data collaborative analysis. In recent years, the frequent security incidents of Internet of Things devices have made people realize the importance of data security of Internet of Things devices. Aiming at the problems such as low key distribution efficiency and difficulty in distributed multi-party collaborative key distribution in cloud-edge collaborative scenario of existing key distribution methods used by Iot devices, this paper proposes a multi-party collaborative key distribution method based on attribute encryption. The method uses attribute encryption algorithm and national secret SM4 algorithm to achieve secure and efficient distribution of session keys in the environment of multi-user request data, ensure the data security of Iot devices, and achieve fine-grained control of user key distribution. Compared with the existing key distribution methods based on asymmetric cryptographic algorithms, this paper verifies that this method has certain advantages in the efficiency of key distribution, and verifies the effectiveness of this method through simulation experiments, and evaluates the performance.
In response to the challenges posed by high power consumption and the logistical difficulties of battery replacement in conventional underwater sensor nodes, we came up with a design that uses backscatter communication. The underwater sensor node comprises a piezoelectric transducer and node control circuit, which underwent simulation to validate its operational capabilities. A MCU governs the switching of the piezoelectric transducer between absorption and reflection states: during absorption, the transducer generates electrical energy to sustain the node; during reflection, the node transmits signals. Simulation results indicate that under 100kHz incident sound waves, the energy harvesting circuit in absorption mode reliably yields a stable 2.2V voltage, ensuring uninterrupted power for the MCU during sensor data retrieval. During the reflection state, the node modulates sensor data onto the reflected signal via backscatter technology, transmitting this signal back to the hydrophone. Operating autonomously without an external power source, this node supports prolonged underwater deployment, thus offering a viable solution for low-power underwater IoT applications.
The current communication signal modulation recognition based on end-to-end deep learning methods is susceptible to signal noise, resulting in poor recognition performance under low signal-to-noise ratio(SNR) conditions. To address this issue, this paper presents a modulation recognition convolutional neural network that combines soft threshold and hole convolution(STCNN). Firstly, adaptive filtering of useless features such as noise is achieved through a soft threshold layer and attention mechanism. Secondly, by using dilated convolutional networks to extract multi-scale time-frequency domain features of signals, the number of model parameters is reduced while the accuracy of signal classification is improved. Finally, simulation analysis was conducted based on the RML2016.10a dataset, and the highest recognition rate of 62.7% was achieved for 11 modulation signals under different SNR conditions, which is 10.9%, 4.5%, and 1.6% higher than that of CNN, LSTM, and SCRNN algorithms, respectively. Furthermore, the number of parameters is reduced by 98.8%, 82.5%, and 91.2%, respectively, which is simulated to verify the effectiveness and recognition performance of the proposed method.
In this paper, we study the downlink transmission of spatial non-stationary extremely large-scale multiple-input multiple-output in the near-field. The line-of-sight channel with uniform spherical wave is assumed in typical mmWave/THz communications, and the sub-connected hybrid beamforming is employed to reduce the energy consumption. Based on these assumptions, we first study the user-subarray pairing. Concretely, the optimal allocation method to maximize the sum-rate is presented, and the signal-to-leakage-ratio based greedy method is presented to maximize the minimal rate. Furthermore, we present a heuristic two-stage hybrid BF optimization with emphasis on the digital beamforming calculation based on the zero-forcing principle. Finally, the effectiveness of the proposed method is verified by computer simulations.
2D Human Pose Estimation plays a crucial role in analyzing performance in fitness activities. Current single-stage methods suffer from the lack of interaction between classification and regression branches, large network parameter sizes, and poor detection accuracy due to strong short-distance dependencies between keypoints. To address these issues in fitness, a lightweight dynamic task alignment framework based on Yolov8-pose is proposed. In order to enhance classification and regression alignment in single-stage networks, a dynamic task alignment detection head is proposed by leveraging label assignment strategies and learning task interaction features. To mitigate the issue of information loss caused by the unidirectional propagation in Yolov8-pose, the backbone is replaced with RevCol to enhance feature retention. Additionally, efficient self-adjusting weighted downsampling module is designed to retain more useful information. Furthermore, the C2f module in downsampling is enhanced with Context-Guided Blocks, integrating local and global feature fusion. Experimental results on a self-created fitness action dataset show that, compared to Yolov8n-pose, our proposed algorithm reduces parameters by 60.8%, decreases computational cost by 33.7%, and improves average detection accuracy by 1.89%.
Artificial intelligence-generated content (AIGC) and intelligent computing applications have grown quickly in recent years, necessitating the inexpensive and effective transfer of large volumes of training data across wide area networks (WAN). Deterministic networking (DetNet) technology for wide-area remote direct memory access (RDMA) access has recently drawn interest in the context of intelligent computing center connectivity. In IP-routed data center networks, RDMA is implemented via the RoCEv2 (RDMA over converged Ethernet v2) protocol, which depends on priority-based flow control (PFC) to create a Lossless network. However, PFC can cause poor application performance due to issues such as network congestion caused by burst traffic, PFC deadlocks, and so on. To overcome these problems, we propose an improved PFC-based flow control method called Reduce Speed Priority-based Flow Control (RPFC). The content is the data sending rate of the upstream device predicted by the downstream device. When burst traffic occurs, the system calculates the difference between the predicted value and the bandwidth, encapsulates the difference into pause frames, and instructs the upstream device to adjust the rate to maintain network stability. The downstream device side’s estimate of the data transmission rate is crucial to the RPFC mechanism. Initially, we examined data transmission rate prediction using existing timing models. Following multiple tests on real-world data, we found that directly predicting the data sending rate was ineffective. Firstly, we investigated the prediction of data transmission rate using existing timing models, and by conducting extensive experiments on a real data set, we found that direct prediction of data sending rate is not effective. To realize RPFC, we propose a prediction algorithm that combines LSTM and Attention mechanisms for time window aggregation grading (TWAG-LSTMA).Experiments show that, by conducting a large number of experiments on a real dataset, our model obtains a good effect.
We propose a novel monostatic multiple-input multiple-output (MIMO) integrated sensing and communication (ISAC) system to provide high resolution 4D information of sensing targets with a low proportional communication capacity loss. This system colocates the ISAC transmitter and radar receiver, using unified orthogonal frequency division multiplexing (OFDM) signals for both communication and sensing tasks. We introduce a frequency resource allocation strategy that sets aside sparsely distributed blocks of subcarriers for ISAC and reserves the remaining subcarriers for conventional communication tasks. The cyclic-shift-based port virtualization technique is utilized on ISAC subcarrier blocks to synthesize a large 2D virtual array, enhancing angle resolution for sensing, and broadcast data with spatial diversity. Additionally, to ensure both sensing accuracy and high communication data rates, we present an ISAC subcarrier block selection method that optimizes the sparse placement of blocks through specific design constraints. Numerical examples demonstrate that our proposed monostatic MIMO ISAC system effectively balances communication and sensing performance, showing significant potential for future sixth-generation (6G) applications.
This paper first incorporates the concept of simultaneous wireless information and power transfer (SWIPT) into simultaneous transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) empowered rate-splitting multiple access (RSMA) system to improve the energy efficiency (EE). On the basis of the power limit at the BS, a multi-user resource optimization problem of maximizing EE is studied by jointly optimizing the transmission-reflection coefficients (TRCs) at the STAR-RIS, the active beamforming vectors at BS, the power splitting (PS) ratios, and the common message rates. To solve the non-convex optimization problem, we decompose the optimization problem into three sub-problems: amplitude coefficient optimization, phase optimization, and common rate PS ratio optimization. Two optimization schemes are proposed: fractional programming and semidefinite relaxation (FP-SDR) based scheme and successive convex approximation and FP (SCA-FP) based scheme. By iteratively optimizing the three sub-problems using alternating optimization (AO), we can obtain the suboptimal solution to original optimization problem. Simulation results show that the STAR-RIS-RSMA scheme with SWIPT can achieve better EE performance than other multiple access technologies.
In this paper, a three-dimensional (3D) time-varying channel model is proposed for unmanned aerial vehicles (UAVs) air-to-air (A2A) wireless channels based on geometric channel model theory in terahertz (THz) band. In this proposed channel model, the scattering fading and reflection fading on rough surfaces of propagation environments, and the atmospheric molecules absorption attenuations are considered in THz band. Moreover, the statistical properties of the proposed channel model, including path loss, time autocorrelation function (T-ACF) and Doppler power spectrum density (PSD), have been derived and analyzed with the several important UAV-related parameters and different carrier frequencies (i.e. millimeter wave (mm-wave) and THz bands). Finally, the correctness of the proposed channel model has been verified via simulation, and some useful observations are provided for the system design of THz UAV-based A2A wireless communication systems.
This paper proposes a method that combines compressed sampling and deep learning to reduce computational complexity while ensuring the performance of detecting direct sequence spread spectrum (DSSS) signals. We consider two scenarios: one where the prior condition of the spreading code sequence period is known, and the other where this prior condition is unknown. By exploiting signal sparsity in the correlation domain, we use random measurements to reduce dimensionality before feeding data into a convolutional neural network (CNN) for detection. Extensive simulation experiments validate the feasibility of this method for direct detection of DSSS signals under low signal-to-noise ratio (SNR) conditions. The approach accomplishes signal detection tasks without signal reconstruction.
Network architecture plays a critical role in achieving the objectives of IMT-2030 and beyond. RAN serves as the foundation for diverse capabilities of 6G. This paper analyzes the evolution of RAN from 3G to 5G NR and proposes a novel RAN architecture for 6G named Unified Access and Control RAN (UACRAN). UACRAN facilitates efficient integration of communication, sensing capabilities, computing, and big data technologies in a cost-efficient manner with reduced complexity. Following an analysis of usage scenarios and requirements of IMT-2030 and beyond, the paper outlines the design principles of 6G RAN, the UACRAN architecture and its underlying concept and its key enabling technologies. UACRAN offers significant advantages: low complexity, lean design, compatible and low-cost. Finally, the paper identifies research areas for academia and industry to explore further.
The Terahertz band supports ultra-wide bandwidth and ultra-high-speed transmission rates. Hybrid dynamic array-of-subarrays (DAoSA) beamforming in massive multiple-input multiple-output (mMIMO) systems offers advantages such as high beamforming gain and low hardware complexity. It is a crucial factor in achieving milli-degree accuracy in estimating the direction-of-arrival (DOA). However, in non-stationary noise environments, the various subcarriers of ultra-wideband signals may encounter beam-squint issues, which present significant challenges for DOA estimation. Addressing this challenge, this paper investigates the DOA estimation problem in Terahertz mMIMO systems employing DAoSA architecture. The Pre-Whitening noise and Beam-squint Canceller method for Multiple Signal Classification (PWBC-MUSIC) has been proposed. Initially, it employs pre-whitening of the array data’s covariance matrix using an estimated noise covariance matrix. Subsequently, it employs a linear transformation matrix to mitigate beam-squint. Furthermore, the paper includes derivation of the Cramér-Rao lower bound (CRLB) to substantiate the algorithm’s performance. The simulation results demonstrate that the proposed algorithm can effectively suppress the interference of non-stationary noise, particularly at low signal-to-noise ratios, and can achieve milli-degree level DOA estimation under ultra-wideband conditions.
This paper considers the joint design of transmit waveforms and receive filters for orthogonal frequency division multiplexing (OFDM) multiple-input-multiple-output (MIMO) dual-function radar communication (DFRC) systems under satellite-unmanned aerial vehicle (UAV) framework. We use the signal-to-clutter-plus-noise ratio (SCNR) as the design metric and introduce dimensionality reduction processing to ensure the target detection performance and reduce computational complexity, respectively. Meanwhile, the multi-user interference (MUI) is constrained to maintain the quality of service for communication. Additionally, we also impose a similarity constraint on the designed waveforms to obtain a good ambiguity function. Finally, an iterative algorithm based on cyclic optimization and semi-definite relaxation (SDR) is presented to tackle the joint optimization problem. The simulation experiments have verified that the designed waveforms and filters can achieve sub-optimal clutter suppression performance, while ensuring low symbol error rate and satisfying achievable communication rate.
In recent years, Large Language Models (LLMs) have been widely used in various fields, including personalized education, data analysis, disease diagnosis, and engineering design. These advancements have opened new possibilities for wireless communication engineering. In this paper, we propose an LLM-based human-machine collaborative framework to generate a simulation platform for the communication system. The proposed framework effectively combines human experience with the powerful generative capabilities of LLMs through well-designed prompt engineering techniques, enhancing the design of wireless communication systems. Specifically, the proposed prompt engineering framework directs the LLM in tasks such as requirement elicitation, system modeling, and code generation for different modules of wireless communication systems. Parallel tests on the commercially mature LLMs like GPT-3.5 and Claude 3 further demonstrate that our approach can improve the efficiency, quality, and reliability of the design process.
Multiple-input multiple-output (MIMO) has made significant contributions to the improvement of system performance. To support increasing connection density, it is promising to deploy more antenna elements with closer spacing on the MIMO array, which is called holographic multiple-input multiple-output (HMIMO). Due to the inevitable coupling effects in HMIMO systems, it needs to determine whether digital pre-coding can bring performance gains as it did in MIMO systems. In this paper, the system performance of HMIMO systems is investigated based on different digital precoding schemes. Moreover, a radio channel model is utilized to analyze the signal propagation process of HMIMO systems and the antenna coupling effects are considered. Six common digital precoding schemes for HMIMO systems are proposed and the sum rate based on these digital precoding schemes is analyzed. The results show that when the antenna array element density increases, the coupling effect plays a significant role and cannot be eliminated by digital precoding. Hence, digital precoding schemes cannot offer performance improvements in HMIMO systems.
With the rapid development of wireless communications, the number of mobile devices is increasing at an unprecedent speed, leading to a shortage of spectrum resources. Dynamic spectrum allocation (DSA) is an effective way to alleviate the scarcity of spectrum resources. DSA relies on spectrum sensing, which aims to detect unoccupied frequency bands. Traditional spectrum sensing methods only consider the scenario of a single sensing node, which can only monitor a limited geographical scope. In order to monitor a large geographical range, it is necessary to consider the distributed sensing architecture. Due to the varying hardware accuracy and sensing environment, there exists a heterogeneity among devices in a distributed system. To address this issue, this paper proposes a new distributed spectrum sensing architecture. The proposed architecture consists of several sensing nodes, each of which is equipped with a convolutional neural network (CNN) in order to identify whether the monitored spectrum is occupied. The shallow layers of these CNNs are demanded to be the same, while the deep layer of each CNN is independently trained using each node's local training data. The proposed method can significantly enhance the sample efficiency while enabling each CNN to be well-adapted to the local SNR (signal-to-noise ratio). Simulation results demonstrates the efficiency of the proposed method.
