With the rapid development of power Internet of Things (IoT) scenarios such as smart factories and smart homes, numerous intelligent terminal devices and real-time interactive applications impose higher demands on computing latency and resource supply efficiency. Multi-access edge computing technology deploys cloud computing capabilities at the network edge; constructs distributed computing nodes and multi-access systems and offers infrastructure support for services with low latency and high reliability. Existing research relies on a strong assumption that the environmental state is fully observable and fails to thoroughly consider the continuous time-varying features of edge server load fluctuations, leading to insufficient adaptability of the model in a heterogeneous dynamic environment. Thus, this paper establishes a framework for end-edge collaborative task offloading based on a partially observable Markov decision-making process (POMDP) and proposes a method for end-edge collaborative task offloading in heterogeneous scenarios. It achieves time-series modeling of the historical load characteristics of edge servers and endows the agent with the ability to be aware of the load in dynamic environmental states. Moreover, by dynamically assessing the exploration value of historical trajectories in the central trajectory pool and adjusting the sample weight distribution, directional exploration and strategy optimization of high-value trajectories are realized. Experimental results indicate that the proposed method exhibits distinct advantages compared with existing methods in terms of average delay and task failure rate and also verifies the method’s robustness in a dynamic environment.

Multi-frequency Global Navigation Satellite System (GNSS) positioning offers significant advantages in positioning accuracy, convergence speed, and carrier-phase ambiguity resolution. However, in real-world operating environments, partial signal outages across frequencies frequently occur, resulting in incomplete multi-frequency observations. Such observation deficiencies weaken the functional and stochastic models and consequently degrade positioning performance. To address this issue, this paper proposes a multi-frequency GNSS incomplete observation repair method based on epoch-differenced observation constraints. A single-satellite, epoch-differenced observation model is constructed and constrained by a rigorous variance-covariance formulation, enabling the recovery of missing measurements while preserving a consistent and reliable precision characterization. To ensure carrier-phase continuity, a robust cycle-slip detection scheme is integrated into the recovery process. The recovered observations are subsequently incorporated into extra-wide-lane combination models to enhance single-epoch positioning capability. Experimental validation using 24-h BDS-3 and Galileo datasets demonstrates that the proposed method achieves a precision of 1-2 mm at a 1 s sampling interval under both ionosphere-ignored and ionosphere-estimated strategies. When the sampling interval is increased to 30 s, neglecting ionospheric variation introduces residuals of approximately 3-5 mm, whereas explicit ionospheric estimation maintains millimeter-level accuracy. In a 197.2 km baseline experiment, excluding one to three frequencies per satellite degrades extra-wide-lane positioning accuracy from 0.137/0.144/0.369 m (N/E/U) to 0.172/0.191/0.606 m. After applying the proposed recovery method, the positioning performance is restored to a level highly consistent with that obtained using complete observations, with deviations limited to 0.007 %, 0.0002 %, and 0.049 % of the original solution in the north, east, and up components, respectively. These results confirm that the proposed method effectively preserves ambiguity fixability and positioning accuracy under incomplete-frequency conditions, providing practical and reliable support for high-precision GNSS applications in challenging environments.

The emergence of wireless networks will empower the smart factory to achieve the next level of efficiency, connectivity, and flexibility while contributing to the development of a sustainable ecosystem. However, the coexistence of cellular networks and WLAN in factories requires an efficient access service to ensure seamless coverage and provide high reliability for mobile services, such as motion control, automated guided vehicle control, and so on. The achievement of a balance between technological advancements and environmental considerations is essential for the long-term sustainability of smart factories. In this paper, we propose an intelligent multi-criteria access selection algorithm, called MASPC, which is the integration of an improved analytical hierarchy process-entropy weight method and deep reinforcement learning, to tackle the problems regarding access failure, ping-pong effect, and low utilization rate of spectrum resources and different sustainability constraints. Simulations show that the proposed algorithm can accurately make network selection decisions with a comprehensive consideration of network condition, service characteristics, user preferences, and sustainability constraint, and also depict significant performance improvement in terms of minimizing unnecessary handover, radio link failure, and throughput compared to other conventional access services.

With growing demand and increasing concern for energy sustainability, smart grids (SGs) have emerged as a promising solution by integrating information and communication technologies to enhance the efficiency, reliability, and flexibility of power systems. While SGs enable real-time monitoring, they also introduce new security risks, particularly for endpoint and edge devices such as smart meters and inverters. Although earlier attacks primarily targeted centralized systems, recent studies have highlighted vulnerabilities on the consumer side, especially in the context of MadIoT-style attacks (MadIoT, short for Manipulation of Demand via IoT, refers to a class of coordinated attacks exploiting high-wattage IoT devices to destabilize power grids). This paper analyzes the attack surfaces of near-field communication network (NFN) protocols and devices within SGs, with a focus on widely adopted public protocols. We propose mitigation strategies to address these risks, including a reverse engineering-based edge device firmware emulation and execution method, a large language model-based protocol analysis approach, and a fuzzing-based malicious behavior simulation technique in a NFN. In our experiments, the proposed AFL-Netzob framework discovered 6 vulnerabilities across 3 firmware samples and achieved up to a 2× improvement in fuzzing efficiency compared to Boofuzz. These results demonstrate the practical effectiveness and general applicability of our framework in real-world smart grid scenarios.

With the deep integration of edge computing, 5G and Artificial Intelligence of Things (AIoT) technologies, the large-scale deployment of intelligent terminal devices has given rise to data silos and privacy security challenges in sensing-computing fusion scenarios. Traditional federated learning (FL) algorithms face significant limitations in practical applications due to client drift, model bias, and resource constraints under non-independent and identically distributed (Non-IID) data, as well as the computational overhead and utility loss caused by privacy-preserving techniques. To address these issues, this paper proposes an Efficient and Privacy-enhancing Clustering Federated Learning method (FedEPC). This method introduces a dual-round client selection mechanism to optimize training. First, the Sparsity-based Privacy-preserving Representation Extraction Module (SPRE) and Adaptive Isomorphic Devices Clustering Module (AIDC) cluster clients based on privacy-sensitive features. Second, the Context-aware Incluster Client Selection Module (CICS) dynamically selects representative devices for training, ensuring heterogeneous data distributions are fully represented. By conducting federated training within clusters and aggregating personalized models, FedEPC effectively mitigates weight divergence caused by data heterogeneity, reduces the impact of client drift and straggler issues. Experimental results demonstrate that FedEPC significantly improves test accuracy in highly NonIID data scenarios compared to FedAvg and existing clustering FL methods. By ensuring privacy security, FedEPC provides an efficient and robust solution for FL in resource-constrained devices within sensing-computing fusion scenarios, offering both theoretical value and engineering practicality.

Download This Paper Open PDF in Browser Add Paper to My Library Share: Permalink Using these links will ensure access to this page indefinitely Copy URL Copy DOI

With the Beidou navigation system's fast expansion in China, it is popular in military and civilian aspects. However, since the satellite orbit operates at an extremely high position and there is energy loss during the propagation process, the receiver only picks up a very faint signal, which makes the Beidou receiver very vulnerable to interference. The interference of the receiver is divided into natural interference and human interference, of which the human interference is particularly serious. Deception is commonly used in human interference. The deception interference detection technology in Beidou navigation system is studied in this research. Firstly, the signal in the signal capture stage is detected by multi-peak detection algorithm to determine the signal type. If it cannot be determined, the signal is detected by the half-peak full-width algorithm, so as to determine the signal type. In the stage of signal tracking, the Doppler shift of the spoofing signal is applied to determine whether the signal is spoofed or not. When the spoofing signal forwarding delay is set to 0.5 and 1 chip respectively, the full width of half peak is 8.56 and 11.35 after fitting the main peak. If the half-peak full width exceeds the normal navigation signal, it indicates spoofing interference. The constructed model can effectively track downspoofing signals and improve the Beidou navigation system’s detection performance.

The increase of terminal devices in the automated power distribution system expands the attack surface for malicious adversaries, which can severely impair the reliability and stability of the power grids. As cyber-attacks are empowered by diverse and elaborate techniques, relying solely on traditional passive defense methods is insufficient to address the evolving security threats. It is imperative to develop adaptive and active approaches to defend against the cyber-attacks aimed at terminal devices in the power distribution system. Motivated by this idea, we focus on the network defense decision-making approach. Effective defense strategies not only mitigate the risk of disruptive cyber-attacks, such as data breaches and system compromises, but also ensure the continuous operation of essential services in the power system. Though researchers have made efforts to explore game-theoretic defense-decision methods for power systems, there are still shortcoming in defining rewards and state transitions. In this paper, we utilize the stochastic game to model the confrontation between the attacker and defender. We design a hybrid reward function to describe the rewards in the two-player game from various perspectives. We also introduce a deep Q-network to learn the high-dimensional state transitions and achieve the Nash equilibrium. Experimental results based on the simulated network are presented to demonstrate the effectiveness of the proposed approach.

5G has become an important driving force for the digital transformation of the power industry. Especially today, with the increasing security risk of big data, the safe and available power 5G business plays an important role in improving power business innovation and user experience. This paper analyzes the demand level differences of different businesses for big data security in the application scenario of power 5G business, and establishes the SLA classification and classification model. On this basis, the power 5G business objective is divided into three processes: business perception, business execution and business SLA evaluation, which are assigned to the corresponding slices respectively. A power 5G slice perception collaborative optimization model is proposed and solved iteratively by multi-objective particle swarm optimization algorithm. Simulation experiments show that compared with the traditional 5G application mode, the proposed scheme can search and optimize the network resource allocation through the cooperation between slices, effectively schedule the slice resources and improve the operation efficiency and performance of power 5G service.

室内定位服务是电网系统运行的关键环节之一，提供基于位置的服务可以有效帮助运检、安监和基建等电力系统运作流程。提出一种基于深度学习的视频定位系统，采用YOLO目标检测算法和深度简单在线实时跟踪算法，实现了对于室内人员的精准定位。在YOLO主干网络部分中加入了CBAM（Convolutional Block Attention Module）注意力机制模块以及GhostBottleneck轻量化模块，验证了不同模型类型下的检测效果并实现了性能的优化。提出了一种基于无损卡尔曼滤波（Unscented Kalman Filter, UKF）的北斗伪卫星系统与视频定位的融合方法，可以实现2种定位结果的融合。实验表明，基于深度学习的视频定位与北斗伪卫星的融合室内定位方法解算精度较高、稳定性较强，可以满足电力系统的室内定位需求。

Due to the particularity of power monitoring system, power communication and data network, the civil public communication information network is difficult to meet the requirements, so the construction of power wireless private network becomes increasingly important. In order to ensure that the power wireless private network can continue to provide basic services in the complex electromagnetic environment, anti-interference measures need to be considered. Therefore, this paper proposes an EMI suppression method based on independent component analysis algorithm for power wireless private network. This method first separates the background noise, then detects the EMI, and finally suppresses the EMI, and verifies the effectiveness of the method through experiments.

Aiming at the efficiency and credibility of data feed in power blockchain application system,the trusted data feed technology based on 5G and Oracle mechanism is studied. Firstly,the data feed method in the blockchain system is discussed,and the Oracle data feed method applicable to the power 5G blockchain is analyzed. Secondly,the trusted data feed technology of the power 5G business system based on distributed Oracle is proposed,and the overall system architecture based on cloud-edge-terminal integration is designed,as well as the workflow of data source registration,evaluation and on chaining. Among them,the distributed data collection and sharing are realized through the blockchain system,and the data source evaluation is realized based on threshold signature algorithm and verifiable random function. The evaluation algorithm is run by the edge Internet of Things (IoT) agent node to ensure the security and availability of the system. Through the division and optimization of 5G slice tasks,the efficient transmission of authentication data and business data is achieved. Finally,the proposed scheme is deployed in the electricity 5G blockchain based power information acquisition system,and experimental verification is carried out from the aspects of communication performance,business performance,resource utilization,etc. The results show that the average data transmission delay of the proposed scheme under different load stress tests is about 10 ms,and the bit error rate and packet loss rate are less than 0.9%. In the case of 100-level concurrent service requests,the performance of the scheme is improved by more than 80% compared with the existing schemes,so it has good feasibility and promotion value.

The establishment of failure risk assessment on the secondary system in smart substation can help power gird company timely arrange maintenance when the failure risk of equipment accumulates to a certain extent,thus effectively preventing systematic failure. In this paper,a probabilistic failure model is established through Weibull probability distribution function from a perspective of reliability engineering. According to the type tests and various daily operation statistics of secondary equipment,the regression equation calculated by the average rank method and least square estimation can obtain the model parameter estimation. By doing so,the equipment failure model close to the real operation condition can be obtained. Moreover,the relationship between the failures of the secondary system and a single equipment is established through the fault tree. Therefore,the risk probability distribution function of top event in secondary system can be calculated quantitatively,which can provide auxiliary decision for risk warning and state overhaul. Finally,it is verified by the numerical results that the proposed secondary equipment failure model can provide a reference for accurate time arrangement for system maintenance.

该文开发了相控阵天线波束赋形及信道测量实验平台,包含相控阵天线波束赋形系统和信道采集与处理系统.波束赋形技术通过将射频信号加权处理形成特定指向的窄波束,精确对准目标用户,辅助计算机技术和波束扫描跟踪快速算法等知识,可实现天线波束宽度和指向控制以及波束扫描和波束跟踪功能.信道采集和处理系统对接收到的射频信号变频至中频信号,上位机控制采集卡不失真连续信道采集,并根据发射已知的信道探测序列同步提取特定方位的信道数据,观察波束赋形技术对信道的影响.各功能均在上位机控制协调下完成,经验证达到了预期效果.

文中对LoRaWAN单元内节点之间的数据速率公平性进行了研究.LoRaWAN数据提取速率不公平的原因在于:(1)分配给节点的数据速率不合理;(2)距离问题,LoRa/LoRaWAN表现出捕获效果,仅提取冲突信号中的较强信号.为解决上述问题,文中提出一种公平自适应数据速率分配和功率控制方案.通过在一个单元内部署每个数据速率的最公平比率并控制传输功率,降低捕获效果,无论其与网关之间的距离如何,都可以进行公平的数据提取.数值仿真结果表明,与现有的最新技术相比,文中方法在数据提取速率上实现了更高的公平性.同时,该方法可以不使用过高的传输功率来维持节点的生命周期,从而降低整个传输系统的整体能耗,符合绿色通信的理念.

The machine learning algorithms applied in multipath components (MPCs) clustering should be evaluated by an appropriate performance index. In this paper, a novel performance evaluation index for MPCs clustering is proposed, which is based on the density of clusters instead of cluster distance. The proposed index improves the traditional S_Dbw validity index by taking into account the intra-cluster density obtained according to the Graham scanning method and Green's formula, which can be applied to MPCs with arbitrary distribution characteristics. The proposed index is used to evaluate the performance of different machine learning algorithms, such as K-means and Gaussian mixture model (GMM) algorithms, which shows a more accurate result than other traditional indexes. In addition, the utility of the proposed index is verified by the measured MPCs data involving both delay and angle information. The evaluation result shows that the variational Bayesian GMM (VB-GMM) algorithm outperforms the K-means and expectation maximization GMM (EM-GMM) algorithms in MPCs clustering.

Fifth-generation (5G) communication network puts forward a demanding requirement for spectrum resources. To improve spectral efficiency, a novel uplink cooperative nonorthogonal multiple access (NOMA) model is proposed. In particular, the proposed scheme enables the relay node to perform a cooperative transmission and uplink transmission simultaneously during the cooperative phase, at the expense of a slight decrease in the signal reception reliability. Moreover, the optimal power allocation strategy of relay node is given. Relay selection criteria is also presented. We analyze the outage probability and outage throughput in both ideal and non-ideal successive interference cancellation (SIC) conditions. Simulation results confirm that the proposed scheme can obtain higher throughput than other cooperative NOMA schemes.

This paper presents a suitability evaluation approach to assess whether the LTE-based wireless private network satisfies the communication needs of smart grid applications. This approach can obtain different evaluation indexes including reliability, delay and throughput when the business attributes are used as the initial inputs by analyzing the MAC and PHY layer models and coverage characteristics. Then combined with the business analysis model, the applicability of wireless private network can be evaluated according to different power business needs. In order to verify the use of the proposed evaluation method, a simulation analysis of the suitability of the LTE 1.8GHz wireless private network is conducted for the businesses of power distribution and utilization.

With the development of digital wireless communication technol-ogy, the wireless signal identification has been suffering from increasingly complex electromagnetic environment and higher spectrum utilization. In this paper, we propose a wireless signal identification method based on interference cleaning and convolutional neural network (CNN) in 230MHz Band. The method firstly analyzes the received signal in time domain, building feature data sets combined with amplitudes, phases, in-phase components and orthogonal components. The method then generalizes singular value decomposition(SVD) and subspace division to preserve signal subspace, eliminate noise subspace and interference compress subspace. Finally, it utilizes the data set to train the CNN and make the wireless signals' identification through the well-trained the CNN. The experimental results with different kinds of modulation show that this method can achieve high recognition accuracy and strong anti-noise ability.

Cellular Internet of Things (cIoT), which will create a huge network of billions or trillions of Things communicating with one another, are facing many technical and application challenges. In particular, in this paper, we have proposed a mechanism based on D2D multicast group communication. The goal of us is to enhance the coverage ability of the cellular network system which is used in IoT. Firstly, we propose a D2D clustering algorithm with the consideration of energy consumption and the outage of the system. Secondly, we want to improve the throughput of the whole system. We decomposed this optimization problem into two sub problems: power control and channel assignment. Referring to the Genetic Algorithm, we get the optimal transmit power of the whole devices. Then, we use the Greedy Algorithm to obtain the perfect channel allocation. Finally, the simulation results demonstrated the efficiency of the communication mechanism proposed by us.