In recent years, the rapid development of smart terminals and wireless networks has led to an exponential growth in the number of terminal devices and data volumes in the power IoT. These data resources have become important assets for power enterprises, significantly enhancing the smart sensing, internal control capabilities, and customer service efficiency of the power grid. However, as critical national infrastructure, power data is vulnerable to cyber-attacks and theft. If leaked, it could cause significant security risks and economic losses. Therefore, power enterprises must strengthen data security protection to address the security issues in data exchange, sharing, and mining. A cloud-edge collaborative intelligent grid data security sharing scheme is proposed, combining zero-knowledge proof technology with a secret sharing scheme to achieve fine-grained access control and privacy protection for user access requests. Finally, the simulation results show that the total time required to perform distributed data access verification in this paper is no more than 48 ms, indicating that the algorithm has a good performance.

With the vigorous construction of the digital new power system, the traditional power communication network is gradually transforming into a more robust, resilient, and multi-service-bearing all-time and all-domain communication network. To address the issue of difficult-to-guarantee deterministic latency for time-sensitive services due to the simultaneous access of multiple types of terminal services and massive data transmission in the all-time and all-domain communication network, a deterministic latency guarantee technology is proposed for all-time and all-domain communication networks based on Time-Sensitive Networking(TSN). Firstly, based on the analysis of the characteristics, importance, periodicity, and latency requirements of various types of services in the all-time and all-domain power communication network, the corresponding service models are established and prioritized. Then, a traffic scheduling mechanism based on TSN perception and shaping is proposed to ensure that the calculation of the gate control schedule meets the deterministic low-latency transmission of time-sensitive flows within the transmission cycle. To achieve the goal of minimizing end-to-end latency, a combined algorithm based on genetic algorithm and tabu search algorithm is employed to calculate the gate control list of time-sensitive flows. The average end-to-end latency is reduced by 15% compared with single optimization algorithm, and the latency jitter of time-sensitive flows is controlled at about 2 s. This improves the scheduling performance and provides strong support for the stable and safe operation of the all-time and all-domain power communication network.

In the downlink collaborative transmission process of the power Internet of Things(IoT), it is necessary to optimize the effective capacity of the system while ensuring low-latency communication, in order to continuously provide high-quality services to users. A Quality of Service(QoS)-Guaranteed strategy is proposed based on Remote Radio Heads(RRH) Collaboration(QG-RRHC). A two-layer power IoT network model based on Orthogonal Frequency Division Multiple Access(OFDMA) is designed. By introducing the theory of effective capacity, a distributed RRH collaborative transmission scheme for QoS guarantee is studied, which collaboratively serves the downlink data transmission on different subcarriers. Additionally, a heuristic optimization algorithm based on Lagrange duality is proposed to solve the joint optimization problem. Simulation results demonstrate that compared with other benchmark algorithms, the proposed strategy can significantly enhance the system's effective capacity and achieve the performance close to the optimal.

With the continuous expansion and increasing complexity of modern power grids, there is a need for a technology that can integrate and process multi-source information to meet the demands of large-scale and highly complex information processing. This is essential to enhance the efficiency and security of power grid operations. To this end, a multi-source information fusion method for power grid operation based on fuzzy sets is designed. The collection of multi-source information for power grid operation is implemented through various sensors, including electric sensors, pressure sensors, and humidity sensors. The data collection methods of these sensors inevitably introduce noise into the collected data. To address this, wavelet denoising methods are employed to reduce noise and extract effective information from the power grid operation data. A multi-source information fusion method combining the fuzzy similarity matrix in fuzzy set theory and the Dempster-Shafer(D-S) evidence theory is designed to achieve the fusion of multi-source information in power grid operation. Experimental test results indicate that as the number of data types increases, the maximum confidence level of this method is in a growth phase. The maximum confidence level of multi-source information fusion reaches 0.94, demonstrating that the fusion results are reliable and applicable to the fusion of various types of data. After adding noise levels of 5 dB, 10 dB, 15 dB, 20 dB, and 25 dB, the maximum confidence level of multi-source information fusion using the designed method only experiences a minimal decrease. This indicates that the method has good robustness in multi-source information fusion. Additionally, the high information entropy values suggest that the fused information is richer in content.

With the large-scale development of renewable energy and the high proportion of massive terminals connected to the grid, the network load in the next generation smart grid will be further intensified, which brings unprecedented and great challenges to the power sensing network for real-time data collection and processing, and whole-domain information monitoring. At the same time, the sensor nodes have the problems of difficult energy replenishment as well as limited computational resources, and the traditional network structure will be difficult to meet the needs of the next generation grid, so it is of practical significance to study how to improve the energy efficiency of power sensor network. A Mobile Edge Computing(MEC) assisted computing offloading scheme for power sensor network is proposed to optimize the nodes' task processing latency and energy consumption under limited computational resources, by modeling the optimization problem as a Markov Decision Process(MDP) and solving the problem using Double Deep Q Network(DDQN) algorithm to minimize the total system overhead. Simulation results show that the proposed scheme outperforms the benchmark scheme in terms of delay, energy consumption and convergence performance.

With the rapid development of the energy Internet, new power services such as vehicle-network interaction have increasingly stringent requirements on service quality, which brings many challenges to the power terminal access network. To address the problem of resource waste and network performance degradation caused by the overlapping coverage of multiple communication technologies and a wide range of communication methods in terminal access network, a Communication Method Selection algorithm based on Random Forest(RF-CMS) is proposed, which intelligently categorizes a large number of diverse new electric power services and selects the most suitable communication methods for them through Random Forest mode. Then, the Multi-Agent Proximal Policy Optimization(MAPPO) algorithm is employed to dynamically allocate routes for the power services from the viewpoint of traffic loading and communication quality to ensure that various terminal service data (e. g., measurement information, control information) can be transmitted timely and accurately in the access network. The effectiveness of the proposed algorithm is validated by comparing it with the routing algorithm based solely on MAPPO in terms of average end-to-end delay and load balancing degree.

With the rapid development of Internet of Things(IoT), big data, and Artificial Intelligence(AI) technologies, massive amounts of data are being generated and utilized on an unprecedented scale. These data contain a large amount of sensitive information, and how to securely store sensitive data has become a realistic problem that needs to be solved. The existing data storage schemes usually focus on the direct protection of sensitive data, while ignoring the leakage risks associated with explicit and implicit associations between sensitive and non-sensitive data. The explicit and implicit relationships among data are deeply analyzed from the perspective of information entropy, and a method is proposed to quickly assess the explicit and implicit relationships and predict the leakage risk of sensitive data. By introducing the information Lift Ratio(LR) and the Probability of Information Control(PIC), the method can effectively identify the influence of non-sensitive data on the risk of sensitive data leakage. In the simulation experiments, the maximum single-attribute LR in the Statistical Property Dataset(SPD) is 0.308, and the joint-attribute LR can be up to 0.891, and the predicted value of the sensitive data leakage risk is significantly improved, up to 23.2%. The simulation results show that the method can effectively identify and cope with the security risks caused by explicit and implicit relationships, thus significantly improving the overall security level of sensitive data storage.

To conduct ultrafast application experiments such as terahertz-laser pump-probe experiments on the Terahertz Free Electron Laser Facility(THz-FEL), it is necessary to achieve high-precision synchronization at the femtosecond level between terahertz pulses and laser pulses. A technical method is proposed based on photoelectric detection and fast feedback from piezoelectric ceramics. Terahertz pulses are generated using the organic crystal N-Benzyl-2-methyl-4-nitroaniline(BNA), and the random jitter of the free electron laser pulses is simulated using an electrically controlled displacement stage. The piezoelectric fast feedback compensation system rapidly adjusts the optical path of the laser to reduce the jitter range, achieving high-precision real-time following of the terahertz pulses by the laser pulses.

There is a significant difference between the terahertz frequency and low frequency in the polarization scattering characteristics of targets. Taking dihedral angle as the research object, the influence of different surface roughness on the identification of target scattering structure and the integrity of profile is studied. Firstly, the rough model is geometrically constructed based on the linear filtering method. Then, the component interpretation is performed on the polarization image by Pauli polarization decomposition, and the integrity of target contour in radar image is quantitatively analyzed. Finally, the electromagnetic simulation results prove that the even component of dihedral angle decreases and the odd component increases with the increase of roughness, and the integrity of target profile is higher in radar image. Results show that the fusion of polarization interpretation and optical interpretation is an important approach for target recognition at terahertz frequencies.

In order to expand the application field of compact High Frequency Ground Wave Radar(HFGWR), research has been conducted on the scenario of single-station compact HFGWR detecting aircraft targets. By exploring the impact of high-speed targets on the range spectrum and Doppler spectrum, as well as the challenges faced in obtaining the speed and range information of high-speed targets, the constraints of the linear Frequency Modulated Interrupted Continuous Wave(FMICW) waveform parameter design have been derived. Corresponding waveform parameters have been designed for long and short coherent accumulation time, and numerical simulations have been carried out. The research shows that under the condition of long accumulation time, the Generalized Radon Fourier Transform(GRFT) method is employed to eliminate the migration of range and Doppler frequency, achieving effective accumulation of target energy and accurate estimation of motion parameters; under the condition of short accumulation time, the signal-to-noise ratio of the target is significantly improved by combining time-frequency analysis with the greedy algorithm to first estimate the motion parameters of the target, and then compensate for the slow time signal of the range. The feasibility of single-station compact HFGWR detecting high-speed targets has been verified through theoretical derivation and numerical simulation, and a theoretical foundation has been laid for subsequent field tests.

Based on the new threats posed by Artificial Intelligence(AI) to electromagnetic spectrum security and the current development status and trends of anti-jamming satellite communication systems, an intelligent anti-jamming satellite communication system architecture based on AI is proposed, which features 'distributed perception+hierarchical decision-making+unified action+quasi-real-time assessment'. The corresponding unit composition architecture of the satellite payload, terminals, and ground stations, and the hardware and software architecture required for engineering implementation are introduced as well in order to adapt to the increasingly intelligent interference situation. In light of the characteristics of extensive use of deep neural networks in the proposed intelligent anti-jamming satellite communication system architecture, the working mechanism of the system architecture is studied, the basic workflow and preliminary deployment methods are put forward. The relevant supporting technologies are analyzed and sorted out, the disadvantages of traditional technologies in areas such as interference perception, anti-jamming decision-making, anti-jamming waveform reconstruction, and quasi-real-time anti-jamming effectiveness assessment, are discussed. The current research status based on AI technology and the key issues that need to be focused on and resolved in the future are concluded. This is intended to provide some reference for the construction of AI-based anti-jamming satellite communication systems.

In response to the application requirements of drone communication systems, a novel metamaterial antenna is designed. This antenna employs periodic metamaterial unit cells as the radiating structure, achieving excellent performance such as broadband and high directivity through efficient manipulation of electromagnetic waves. The antenna consists of a 3×3 rectangular metamaterial array as the radiating element, fabricated using 0.15 mm thick copper foil attached to a 7.8 mm thick polymethylmethacrylimide(PMI) foam substrate. A coplanar waveguide with a coupled aperture feeding structure is utilized to reduce the antenna profile. The antenna measures 188 mm ×111 mm×7.8 mm, with a profile height of 0.034L (where L is the wavelength corresponding to the lowest operating frequency), and weighs only 52 g. It is mounted within an aluminum housing. Simulation and test results indicate that the antenna has a Voltage Standing Wave Ratio(VSWR) of less than 2, an impedance bandwidth of 28.4%(1.30~1.73 GHz), and a gain of 8.1~10.0 dBi within the operating frequency band.

With the rapid development of miniaturized implantable Radio-Frequency(RF) antennas, wireless sensing technologies for human body monitoring have ushered in new opportunities for development. Based on the structure of dental implants, a Dielectric Resonator Antenna(DRA) is proposed. This antenna exhibits good simulation and measurement results, with a resonance at -15.6 dB at 3.82 GHz, and a radiation efficiency of over 97%. When the input power is below 22.8 mW, the Specific Absorption Rate(SAR) meets the requirements for human health and safety. On this basis, the antenna is optimized into a hybrid dielectric resonator antenna with embedded metal. This optimized antenna has higher gain and stronger directivity in its far-field radiation pattern, which can enhance the signal transmission link of the antenna after it is implanted in the oral cavity. Both proposed antennas are small in size and have excellent performance, and they are expected to be implanted in the human oral cavity based on dental implant technology to monitor the oral environment.

The data volume of millimeter-wave security imaging is relatively small, and the data volume for uncommon types of dangerous objects is even smaller, with the data exhibiting a long-tail distribution. Security inspection requires fast detection speed, and existing deep learning methods are not yet fully applicable to millimeter-wave security imaging datasets. To meet the needs of daily security inspection, a method based on the YOLOv5 algorithm is proposed for feature extraction of human hidden targets in millimeter-wave security radar imaging. Firstly, the YOLOv5 algorithm framework is studied, and a Focal CIoU loss function is proposed to re-weight the samples and reduce the long-tail effect. Next, the dataset is processed by cropping the targets and randomly pasting them into existing images to expand and balance the data volume among different categories, achieving the purpose of resampling. Finally, the Squeeze-and- Excitation Network(SENet) attention mechanism is introduced to improve the accuracy of target recognition. The validation results show that the mean Average Precision (mAP) of human hidden targets using the proposed method reaches 85.4%, which is a 4.7% improvement compared to the original YOLOv5 algorithm. This performance meets the detection requirements for daily usage scenarios.