Low earth orbit (LEO) satellite constellations are evolving to support on-board processing, reconfigurable functions, satellite-terrestrial cooperation, and the coordinated execution of multiple tasks, including communication, sensing, navigation, and control.However, LEO satellite networks are constrained by bottlenecks such as inadequate signal processing and computing capabilities of individual LEO satellites and constrained inter-satellite link capacity, necessitating a deep integration with terrestrial mobile communication networks that include extensive cloud-based processing and computing capabilities.Therefore, an innovative architecture for an integrated satellite-terrestrialbased 6G cloud-fog collaborative self-organizing network was proposed and its protocols were discussed, achieving the deep integration of LEO fog-based satellite networks and terrestrial cloud-based communication networks.In addition, key technologies such as integrated communication, sensing, navigation, and control waveform, as well as resource management based on satellite onboard fog processing, along with mobility management and intelligent information processing based on cloud-fog collaboration, were elaborated in detail.In the end, insights into future directions for standardization and hardware experimentation were presented.

To solve the network access problem of low data rate and delay-insensitive services in remote areas, a widearea Internet of things (IoT) information transmission method based on mobile piggyback was proposed.Mobile carriers were used to provide senseless access support for IoT devices and expand network coverage.In areas without base station signal coverage, the mobile carrier established a connection with the sensor, obtained sensor data and stored it.When 5G access conditions were met, the sensing information was offloaded to achieve information access to the network.A data caching strategy based on cost function was proposed to improve the efficiency of information piggybacking.The coverage performance of the mobile piggyback scheme was analyzed and a hardware implementation scheme based on LoRa was designed.Simulation results show that the wide-area IoT information transmission method based on mobile piggyback can improve the efficiency of IoT information piggyback, and expand the coverage area of 5G network by 4.98 times.Field tests verify the effectiveness of the proposed method, and mobile piggyback nodes can achieve communication coverage with a radius of 4 km.

In the industrial Internet of things, the generation of tasks is observed to exhibit both continuity and periodicity, along with stringent latency requirements.These characteristics posed challenges to system’s cost-efficiency.To address these challenges, a cost minimization resource allocation algorithm based on the urgency of tasks was proposed.By employing a genetic algorithm, the task offloading strategy and the system’s resource allocation strategy were optimized.For offloaded tasks, they were scheduled according to their level of urgency.Additionally, the optimal transmission power for each task was calculated to meet latency constraints.Simulation results indicate that the proposed algorithm effectively reduces the overall energy cost of the system.

Accurate channel estimation is a critical component in the design of 5G OFDM communication system receivers, since it can significantly reduce the bit error rate (BER), thus improving wireless communication efficiency and quality.Channel estimation methods based on least square (LS) and minimum mean square error (MMSE) effectively utilize the system’s sparsity, but LS algorithms face low computational precision, while MMSE algorithms suffer from high computational complexity.To promote the estimation accuracy, practitioners have presented several deep learning-based channel estimation methods.However, existing methods often split complex matrices into real and imaginary parts, failing to adequately capture the complex characteristics of the channel, leading to distortion in the estimated channel matrix.A complex-valued generative adversarial network (GAN) model that could fully extract the complex features of the signals was proposed, enabling accurate estimation of the channel matrix for the physical downlink shared channel (PDSCH) in the 5G new radio (NR) standard.To validate the effectiveness of the proposed method, the proposed method was compared with LS algorithms, actual channel estimation, super-resolution neural networks, and residual neural network channel estimation methods.Results show that when the mean square error between the estimated channel matrix and the true channel matrix is 0.01, the proposed method-based communication system has a signal-to-noise ratio (SNR) that is 5 dB higher than existing ones.

The integrated deployment of 5G and time sensitive network (TSN) is an important foundation for realizing wireless industrial Internet and flexible manufacturing.How to guarantee the high-quality transmission simultaneously of diverse industrial services and mobile services is a key issue to be solved for 5G TSN integration.Therefore, the types of industrial services were divided according to the differences in requirement characteristics, a slice scheduling strategy for industrial Internet was designed to satisfy the requirements of periodic time-critical services and sporadic burst services, and a resource partitioning algorithm based on uniform distribution was proposed.By controlling resource reservation and resource preemption, the proposed method can optimize the utilization of wireless resources and guarantee the delay of industrial traffic and the transmission rate of mobile services.Finally, the effectiveness of the proposed method in improving the service requirements satisfaction and resource utilization was verified by simulation experiments.

There are many ways to generate adversarial samples for synthetic aperture radar (SAR) images at present, but some problems such as large amount of perturbation of adversarial samples, unstable training, and unguaranteed quality of adversarial samples still exist.To solve the above problems, a SAR image adversarial sample generation model was proposed.The model was based on the AdvGAN model architecture.Firstly, according to the characteristics of the SAR images, an adaptive threshold segmentation method based on the enhanced Lee filter OTSU was designed.The mask extraction module composed of equal modules, this method produced a smaller amount of disturbance, and the structural similarity (SSIM) with the original sample reached that more than 0.997.Secondly, the improved relativistic average GAN (RaGAN) loss was introduced into AdvGAN, and the relative mean discriminator was used to make the discriminator rely on both real data and generated data during training, which improved the stability of training and the attack effect.Experiments were compared with related methods on the MSTAR dataset.Experiments show that the attack success rate of SAR image adversarial samples generated by this method is increased by 10%～15% than that of traditional methods when attacking defense models.

In the multilayer graph convolution calculation, each node is usually represented as a single vector, which makes the high-order graph convolution layer unable to distinguish the information of different relationships and sequences, resulting in the loss of information in the transmission process.To solve this problem, a heterogeneous graph neural network algorithm based on meta-path convolution was proposed.Firstly, the feature transformation was used to adaptively adjust the node features.Secondly, the high-order indirect relationship between the nodes was mined by convolution within the meta-path to capture the interaction between the target node and other types of nodes under the element path.Finally, the reciprocity between semantics was explored through the self-attention mechanism, and the features from different meta-paths were fused.Extensive experiments were carried out on ACM, IMDB and DBLP datasets, and compared with the current mainstream algorithms.The experimental results show that the average increase of Macro-F1 in the node classification task is 0.5%～3.5%, and the ARI value in the node clustering task is increased by 1%～3%, which proves that the algorithm is effective and feasible.

The use of frequency measurements to achieve target positioning is characterized by low cost and high reliability.Using only frequency difference of arrival (FDOA) measurements, a precise localization method for static target position was proposed.To address the highly nonlinear of the established frequency measurement equation, it was transformed into a pseudo-linear equation in matrix form by introducing auxiliary variables.Then the non-convex weighted least squares (WLS) problem was relaxed into a semi-definite programming (SDP) problem by using the semi-definite relaxation (SDR) method, so as to further accurately estimate the position of unknown variables.Finally, the root mean square error (RMSE) of the proposed method was analyzed to verify its performance.The simulation results show that the performance of the adopted semi-definite relaxation method is able to reach the Cramer-Rao lower bound (CRLB) at lower Gaussian noise levels and the algorithm is highly robust to geometry.In addition, its RMSE performance is better than that of the two-stage weighted least square (TSWLS) method for a smaller number of sensors.

Aiming at poor performance of physical layer security (PLS) of decode-and-forward cooperative nonorthogonal multiple access (NOMA) systems and its performance deterioration in the presence of multi-eavesdropper, the PLS model of multi-relay selection NOMA wireless systems using modify-and-forward (MF) protocol was proposed.The approximate expressions of secrecy outage probability(SOP) and asymptotic SOP of multi-relay selection MF-NOMA wireless systems with non-colluding or colluding eavesdropping scenarios were derived by using GaussChebyshev Quadrature theorem.The simulation results verify the accuracy of PLS performance analysis of the above MF-NOMA wireless systems.The results also indicate that the PLS performance of multi-relay selection MF-NOMA wireless systems will be improved when the number of relays increases or the number of eavesdroppers decreases, and that colluding eavesdropping is more detrimental to secure transmission of multi-relay MF-NOMA wireless systems than non-colluding eavesdropping.

In the process of locating faults caused by branch line breaks in distribution networks, it is difficult to determine the range of wire breakage due to differences in fault interference branch node identification methods.The relative error of fault location results under different sampling rates is relatively large.Therefore, a accurate location method for branch line break faults in distribution networks considering system operation mode was proposed.The traveling wave transmission path within the power supply system was analyzed based on the topology of the distribution network.A connection matrix for branch nodes was constructed in the distribution network based on the operating mode of the power supply system.Identify the branch nodes affected by wire breakage faults and determine the scope of wire breakage occurrence.Determine the fault occurrence interval based on the characteristic parameters of wire breakage faults.By combining the time difference between zero mode traveling wave and line mode traveling wave, fault location can be obtained, achieving precise positioning of branch line breakage faults in distribution networks.The example analysis results show that after obtaining a high-precision traveling wave signal waveform, the relative error of fault localization results at different sampling rates is reduced by 48% and 12%.

In order to solve the problems of high cost, low efficiency, and weak intelligence in the traditional operation and maintenance mode of optical transport networks, the research status and challenges of knowledge graph technology in the telecom industry were analyzed.Firstly, an intelligent operation and maintenance system architecture for optical transport networks based on knowledge graphs were proposed.Key technologies such as network state perception, knowledge graph construction, and open search capabilities were emphatically studied.Secondly, the application of intelligent cutover in a city branch of China Telecom was taken as an example, the application effect was analyzed, and the test showed that the cutover efficiency was improved by about 3 times.Finally, the development trend of intelligent operation and maintenance of optical transmission networks based on knowledge graphs was discussed.

With the continuous development of communication technology and the accelerated arrival of enterprise digitization, the telecom business support system is also evolving and changing.In order to realize the goal of highquality, high-efficiency and sustainable evolution, telecom operators need to establish an effective and adaptive to the market demand of business support system hierarchical evolution target system.Based on the analysis of the basic characteristics of the current telecom business support system and the demand of operation services, a framework of L-PS-M (level-process-system-measurement) hierarchical evolution target system for telecom operation service support was proposed.The framework firstly constructed a three-dimensional cube mapping structure formed by the three dimensional factors of end-to-end business process, system support capability, and hierarchical goal.Secondly, a subdivided implementation path of the operation service support system was given.Thirdly, telecom carriers could carry out the planning and optimization of the system support capability according to the actual demand in a targeted way through the implementation path, so as to improve the efficiency of the operation and system production, and further clarified the evolution of the business support system, which was applicable to telecom operations.It further clarified the evolution direction of the business support system.

The core value of 5G in the field of digital distribution network is to give play to the technical advantages of 5G in ultra-high reliability and low latency, and provide high reliability and low latency communication for the control related business of distribution network.In order to provide a systematic theoretical analysis scheme for the latency and reliability of 5G communication deployment applications in the distribution network, a typical architecture of 5G communication in the distribution network in combination with digital distribution network services was given, the latency model for 5G communication in the distribution network was deeply analyzed.The end-to-end latency analysis for 5G frame structures of four operators in China was conducted with 5G reliability values of 3GPP Rel-15 and Rel-16 through 5G link level simulation.At last, a joint analysis method of time latency and reliability of 5G communication in distribution network was proposed.

Performance tests help to verify whether a business cluster can meet its access scale requirements, but current tools have problems such as evaluating proper parameters and requiring manual inputs, low performance, and difficulty in automation.A comprehensive quantitative request rate control mechanism and concurrent design approach based on coroutines were proposed, in contrast to JMeter’s reliance on threads and locks.Furthermore, an architecture design for its corresponding TaaS service and a process automation scheme were introduced.This scheme automatically perceives responses, adjusts access rates accordingly, and schedules additional server executions using real business data, significantly enhancing operational efficiency and result intuitiveness.