Proceedings of the 13th International Conference on Computer Engineering and Networks

Edge devices of health monitoring systems are constantly generating a large amount of data. Because each piece of data is accompanied by a signature to verify its authenticity, there is an urgent need to reduce the space occupied by the signatures. In this paper, we introduce a new health monitoring system model using locally verifiable aggregate signatures to meet the need. The locally verifiable aggregate signature can compress multiple signatures into a single aggregated signature and recover all the original signatures from the aggregation. It not only reduces the space for storing signatures but also reduces the authentication cost, especially for verifying the authenticity of a single piece of data. Based on the RSA signature proposed by Seo (Information Sciences 2020), we present a concrete locally verifiable aggregate signature scheme from the RSA assumption, instead of other strong assumptions. It is proven that our scheme is secure in the standard model.

This paper presents an approach utilizing static and dynamic analysis techniques to identify malicious Android applications. We extract static features, such as certificate information, and monitor real-time behavior to capture application properties. Using machine learning, our approach accurately differentiate between benign and malicious applications. We introduce the concept of “Multi-dimensional features”, combining static and dynamic features into unique application fingerprints. This enables us to infer application families and target groups of related malware. Tested on a dataset of 8000 applications, our approach demonstrates high detection rates, low false positive and false negative rates. The results highlight the effectiveness of our comprehensive analysis in accurately identifying and mitigating Android malware threats.

The mismatch point elimination algorithm is a commonly used method in the field of computer vision and image processing to deal with the presence of mismatches or outliers in matched point pairs. These mismatch points may be caused by noise, occlusion, illumination changes or image distortion. In this paper, we first explain why there is a need to eliminate the mismatch points and the current state of research, and then introduce various types of feature points and describe the extraction methods of various feature points. Next, we review several methods of false match feature point elimination, such as geometric consistency verification-based methods, graph optimization-based methods, motion statistics-based methods, and learning-based methods, analyze their advantages and disadvantages as well as make comparisons, and give an outlook on future research directions. In the conclusion, we summarize the full paper and discuss the application trends of the mismatching feature point elimination algorithms. The purpose of this paper is to provide readers with a clearer and deeper understanding of false match feature point elimination algorithms, and hopefully give some reference significance to later researchers.

Federated learning is a privacy-preserving solution for distributed machine learning, allowing participants to solve machine learning problems collaboratively without transmitting their local data to a central server. Instead, they exchange model parameters to achieve the desired outcomes. However, recent scholarly research has revealed several challenges in the traditional federated learning framework. This paper aims to address the issues of communication efficiency, privacy leakage, and client selection algorithms within the federated learning paradigm while exploring potential future research directions.

In the 5G and 5G+ scenario, a large number of devices access to the network and a large number of different services are demanded by different vertical industries. To maintain the QoS and satisfy such a lot of demands, MEC (mobile edge computing) deployment and IDN (Intent-Driven Network) scheduling are necessary. MEC could upload the third-part App to the cloud servers, which could save the calculate force in the local UE (user equipment). And IDN could identity and integrate UE intents in natural language and translate them into cloud-network scheduling policy to implement them and manage cloud-network resources. ONAP as the platform for orchestrating, manages and automating network and edge computing services. This paper introduces the intent instance management model of ONAP and IDN, introduces the MEC, containers and slices management and designs intent-and-resource -weighted algorithms to make the policies and ensure the QoS.

The emergence of blockchain-based anonymous and encrypted digital currencies has brought with it a rapid increase in financial crimes. However, the regulation and detection of financial crimes requires the detection of abnormal transactions in the scenario of blockcain-based anonymous and encrypted digital currencies, where traditional methods are not applicable. In this paper, we propose an abnormal transaction node detection method on bitcoin based on outlier ranking of transaction communities. The public key addresses in bitcoin transactions are merged according to whether they belong to the same user in order to form a user transaction graph, which is used as the input of our method. This graph is then divided into smaller communities. The abnormal transaction nodes are detected by ranking each node with its inter/intra-community link outlier value. By conducting experiments on a subset of bitcoin transactions, it shows that the proposed method is able to effectively detect known abnormal nodes involved in financial crimes.

SLAM (hereinafter referred to as SLAM) refers to the autonomous mobile carrier in the unknown environment, through the data information obtained by its own sensor to achieve its own positioning, in addition to the technology can also continuously build and update the map in the process of carrier movement. Visual SLAM is a technology that uses visual sensors as input and uses dense perception of the surrounding environment to achieve SLAM function. Compared with the traditional SLAM method, visual SLAM can retain the semantic information in the environment while ensuring the accuracy, so as to expand the function of the carrier. This paper first introduces the milestone methods in the field of visual SLAM in chronological order, then introduces the standard flow of visual SLAM, and finally introduces the advantages and several typical excellent algorithms of visual SLAM that integrates point-and-line features.

As the aging population grows, health issues have attracted widespread attention, especially among older adults. We explored the effects and simulation prediction of network services based on a multi-dimensional analysis of the health status of older adults, which could help older adults to better manage and evaluate their health. The study was conducted as follows. Based on the China Family Panel Studies (CFPS) data, a chi-square test was used to screen out 16 indicators with significant effects on the self-rated health (SRH) of older adults. To eliminate selection bias between samples, a propensity score matching (PSM) was used to explore the potential impact of network services on SRH of older adults. A multi-agent simulation model was constructed to examine SRH effects and then compare the health both before and after using network services based on the AnyLogic platform. The XGBoost algorithm was used to build a prediction model for assessing SRH of older adults. The experimental results show that network services have a positive effect on SRH of older adults using the multiple PSM methods, improving SRH of older adults by 14.1%. Meanwhile, the multi-agent simulation proves that network services can improve the health status of older adults. It also proves that the XGBoost algorithm has better accuracy, specificity, and running time than the other compared algorithms, and can meet the prediction needs of this paper. This study may enrich and expand the theoretical framework of health influence mechanism and simulation prediction studies.

With the continuous deepening of information technology construction and the surge in demand for telecommuting, traditional security protection measures are difficult to cope with complex network environments. Solving security issues such as telecommuting based on Zero Trust architecture become a focus of attention. The core of a Zero Trust architecture is “continuous verification, never trust”, which means that by default, both internal and external personnel, terminals, and businesses are considered untrustworthy, and their access to the network and business resources will be continuously verified and evaluated. The paper first expounds the historical evolution, basic characteristics, and key technologies of Zero Trust, and then proposes a telecommuting security solution based on Zero Trust architecture. The solution can effectively solve problems such as identity trustworthiness discrimination, device trustworthiness discrimination and behavior trustworthiness discrimination, to achieve secure and reliable business access for telecommuting workers and intelligent terminals. The solution has reference significance for further optimization and implementation of Zero Trust framework in relevant application scenarios.

Previous prefetching schemes have been found to be very effective at enhancing the performance of computers. However, speculative prefetching requests can have negative effects on computers, such as increased memory bandwidth consumption and cache pollution. To address the deficiencies of previous prefetching schemes, we propose the Reinforcement Learning Based Offset Prefetching Scheme (RLOP), an offset prefetching scheme based on reinforcement learning. As with previous offset prefetching schemes, RLOP evaluates multiple offsets and enables offsets that qualify to issue prefetching requests. RLOP, however, selects appropriate prefetch offsets through reinforcement learning, and the reinforcement learning reward scheme determines the goal of the prefetcher; we divide the rewards into four different rewards—accurate and timely prefetch, accurate but delayed prefetch, inaccurate prefetch, and no prefetch operation—and by increasing or decreasing the reward value, we facilitate or inhibit RLOP from future environments to collect such rewards, which enables or inhibits RLOP from collecting such rewards, which enables We evaluated and contrasted RLOP with various advanced data prefetchers and demonstrated that our scheme resulted in a 25.26% increase in system performance over systems without data prefetchers and a 3.8% increase over the previous best performing data prefetcher.

The War Potential Network (WPN) is critical infrastructure determining national security. With the recent trend of increasingly tense international situation, frequent occurrences of cyber-attacks, and the proliferation of new intelligent endpoint devices in WPN, the importance of Continuous Auditing (CA) for intelligent endpoints in WPN has become increasingly significant. Several researches have focused on the accuracy of CA. However, the information in WPN intelligent endpoint devices might have sensitive information. Some laws require computer systems to not disclose data containing national secrets, while certain legal regulations demand the protection of personal privacy. In order to meet compliance requirements, specific technologies have to be implemented in CA, while there are existing research gaps in this field. To fill the gap, this research proposed a compliance-enhancing approach based on Locality-Sensitive Hashing (LSH) and clustering method to enhance compliance in CA. In this approach, auditing nodes gathers encoded data which cannot be read by human, while can be analyzed by algorithms to conduct CA. To quantitatively evaluate this approach, this research also introduced an inference attacking method in WPN scenario as threat model. The research also evaluated the influence of the capability of the auditing object and the correctness of the auditing result, to prove our compliance-enhancing approach can achieve relatively good performance in different evaluation dimensions.

Expensive and relatively rigid specialized processing devices are used to perform audio and video processing functions in traditional audio and video systems. These devices not only have limited configurability but also lack openness to the users. In this paper, an embedded streaming terminal is implemented by embedded chips, combining software and hardware. It integrates audio and video capture, encoding/decoding, transmission, and image processing functions. The encoding/decoding, transmission, and fusion processing functions are standardized and quantifiable, forming a versatile processing terminal. It is constructed through quantity stacking and Ethernet switch cascading, enabling functions such as video switching, video splicing, video overlay, PIP (Picture-in-Picture), PBP (Picture-by-Picture), and seamless zooming. The terminal exhibits excellent performance in terms of encoding/decoding latency, bitrate control, and overlay channels, ensuring the flexibility and scalability of audio and video business systems.

In the current digital and networked era, the demand for digital copyright transactions is growing, and the traditional centralized registration method has problems such as high costs and long time limits for copyright registration, and difficulties in copyright protection. The technical features of blockchain technology, such as decentralization, immutability of information, openness and anonymity, offer new opportunities for digital copyright protection. We design a digital copyright transaction scheme based on blockchain technology. Firstly, we use Java to build an underlying platform that simulates the operation of the blockchain data layer, consensus layer and contract layer, on top of which we design a distributed underlying blockchain platform that can securely generate blocks, generate transactions, validate transactions, add transactions into blocks, broadcast blocks, validate blocks and add blocks into the blockchain, and use an interface to manipulate the database for the trading functions to be implemented by the platform, and a visual interface is provided to facilitate users’ transactions of digital copyright content. The results show that our platform can effectively address trust, intermediary and execution problems in digital copyright transactions and help reduce copyright infringement.

The Internet has penetrated into various fields of human production and life. While enjoying Internet technology, people have to face various problems brought about by the Internet, among which network security issues are particularly prominent. The network security situation assessment summarizes, filters and analyzes security events generated by devices, builds suitable mathematical models based on security indicators and assesses the level of security threats to the entire network system, thereby analyzing and capturing the overall security status of the network. This paper analyzes the relevant research at home and abroad, and selects Elman neural network model, intuitionistic fuzzy set model and hidden Markov model for network security situation assessment. The result is compared with the expert assessment, and the advantages and disadvantages of the different models are analyzed in conjunction with relevant model theory. It is found that the network security situation assessment model more suitable for the current network environment is the intuitionistic fuzzy set model.

In this paper, a millimeter-wave wireless localization model assisted by IRS is developed to explore the wireless localization problem under NLOS conditions. By deriving and analyzing the Fisher information matrix and Cramer-Rao lower bounds(CRLB), the accuracy index and optimization criteria are quantified, and a phase shift optimization method based on the semi-positive definite relaxation method is proposed to achieve the optimization of the localization performance of the system as a result. Simulation results show that the algorithm proposed in this paper can improve the performance of the system significantly compared with the benchmark solution.

Driven by the double carbon target, the energy revolution is imperative, and traditional single-energy power stations are gradually being transformed into a new system form with new energy complementary types, integrating digitalization and intelligence. China is promoting the development of multi-energy complementary tidal power stations, which incorporate and complement the use of green renewable energy sources such as light, wind, and tidal energy in an efficient manner. On this basis, multi-energy complementary tidal power stations should also combine the current digital, intelligent, networked, and platform-based technology features with building an intelligent platform to solve the new energy power station grid connection and other problems, thus improving the pressure on electricity and achieving clean use. In addition, China's rapid technological and social development, energy and enterprise transformation, and upgrading are imminent, so high efficiency, high integration, and friendly power stations have become an important trend for future development, with broad development prospects.

In order to solve the problem of economic loss caused by malicious use of underlying network resources by network attackers, this paper proposes a resource security management mechanism based on dynamic key and blockchain in the network slicing environment. First, a resource security management architecture based on dynamic key and blockchain in the network slicing environment is designed. The architecture includes two modules: the bottom network provider and the service provider. Each module consists of three sub modules: the certification center, the resource management center, and the resource settlement center. The overall mechanism of resource security management based on dynamic key and blockchain in the network slicing environment includes three steps: the service provider obtains the permission to apply for resources, the service provider applies for and obtains resources, and the underlying network provider completes the resource cost settlement. Through the analysis of resource security performance from three aspects of preventing man in the middle attack, preventing data tampering, and data leakage, we can see that the resource security management mechanism proposed in this paper can better solve the problem of economic loss caused by malicious use of underlying network resources by network attackers.

In order to solve the problem of low security in transactions between multi service providers and multi bottom network providers, this paper proposes a secure and efficient network slice resource access control mechanism in a distributed environment. First, a secure and efficient network slice resource access control architecture in distributed environment is designed. The framework consists of multiple domains. The roles of each domain include multiple service providers, multiple underlying network providers, authentication servers, and resource use license servers. The overall mechanism of secure and efficient network slice resource access control in a distributed environment includes six steps: the service provider applies for identity authentication to the authentication server, the service provider applies for resources to the resource use license server, the resource use license server returns the list of underlying network providers to the service provider, the underlying network provider allocates resources to the service provider, the service provider applies for resources from a resource use license server outside the domain, and the resource use license server outside the domain returns a remote list of underlying network providers to the service provider. Finally, the mechanism in this paper has good performance.

In the context of utilizing 5G wireless technology for facilitating the timing service of the industrial Internet of Things (IoT), achieving precise clock synchronization while considering the balance between 5G wireless resource utilization and network performance becomes imperative. Firstly, an incomplete observation clock synchronization model is established. Subsequently, the Kalman filter algorithm is employed to determine the boundedness of clock synchronization error, enabling the formulation of an optimization model for 5G wireless resource allocation aimed at ensuring clock synchronization accuracy. Furthermore, a two-step optimization framework is introduced, employing a clustering algorithm and Lyapunov method, combined with multi-agent deep reinforcement learning, to effectively address the proposed problem. Simulation results corroborate the efficacy and superiority of the proposed model and methodology in achieving a joint optimization of clock synchronization accuracy and throughput.

The COVID-19 epidemic is rampant, affecting the normal life of people all over the world. It also has a sustained impact on the global economy, bringing global crises and challenges. In order to effectively prevent and control the epidemic situation and curb the further spread of COVID-19, the identification of close contacts has become a current research hotspot. The low accuracy, insufficient user privacy protection ability, and limited effectiveness have become important issues to be solved. In this paper, we propose a solution based on P-RAN, fog computing and blockchain to support accurate, safe and effective COVID-19 close contact identification. A three-layer accurate close contact identification architecture (ACCCD) is established by combining P-RAN, fog computing and blockchain. P-RAN and fog computing are combined to collect close contact information, and blockchain technology enables reliable, safe and automatic close contact information management and close contact track tracking. Based on ACCCD architecture, the blockchain network deployment scheme and accurate close contact identification process are designed, and the implementation potential and future challenges of this solution are analyzed and explained in detail. The application of ACCCD architecture in specific scenarios is described, and we design simulations to verify the effectiveness of the solution.

With evolution of 5G networks and IoT technology, enormous amount of network resource demands are raising. How to effectively promote the interconnection and comprehensive sharing of resources which belongs to different resource providers to meet the demands of diversified services is a critical issue. Considering the superior advantages of blockchain technology in distributed systems, we propose a blockchain-based decentralized solution for network resource sharing and define a trusted reputation system in this paper. A two-layer distributed network resource sharing architecture are proposed. Based on this architecture, a reputation system are designed. Based on the behavior of network resource provider during, the sharing reputation are quantified to provide a reference for network resource requester to choose suitable resources. Then a reputation-based shard parallel consensus algorithm are developed. Finally, the simulation experiments are designed to analyze the performance of reputation system. The results show that with the support of the reputation system, the resource sharing system will develop healthily.

Computational offloading is considered a promising emerging paradigm for addressing the limited resources of edge devices in expanding power grids. However, with the advancement of intelligent technologies such as digitalized power grids, applications often consist of several interdependent subtasks, forming interconnected automated workflows. This paper focuses on the computational offloading technique within task-dependent workflows. It proposes a multi-objective optimization problem for offloading, considering both time and energy consumption. The model takes into account the constraints of task duration, communication capacity, and computational capacity. Additionally, a predictive-guided a predictive-guided multi-objective reinforcement learning algorithm based on Pareto optimization (MORLBP) is introduced. This algorithm combines the principles of multi-objective optimization, Pareto optimality theory, and deep reinforcement learning. It utilizes the quality of the Pareto front as a metric and is compared against NSGA-II and MOPSO algorithms. The proposed algorithm’s effectiveness and advancement are validated through simulations, demonstrating its efficiency and innovation in tackling the multi-objective offloading problem within task-dependent workflows.

Network traffic prediction has always been an important research topic, frequently employed in intelligent network operations for load awareness, re-source management, and predictive control. Most existing methods adopt a centralized training and deployment approach, neglecting the involvement of multiple parties in the prediction process and the potential for training prediction models using distributed methods. This study introduces a novel wireless traffic prediction framework based on split learning, addressing the limitations of existing centralized methods. The proposed framework enables multiple edge clients to collaboratively train high-quality prediction models without transmitting large amounts of data, thus mitigating latency and privacy concerns. Each participant trains a dimension-specific prediction model using its local data, which are then aggregated through a collaborative interaction process. A partially global model is trained and shared among clients to tackle statistical heterogeneity challenges. Experimental results on real-world wireless traffic datasets demonstrate that our approach outperforms state-of-the-art methods, showing its potential and accuracy in Internet traffic prediction.

In the computing network convergence environment, the deployment and scheduling of the service function chain put forward high requirements for network flexibility, and the traditional network architecture is difficult to sup-port highly flexible network scheduling functions. SRv6 has become an important way to implement service function chain technology because of its segment routing function and protocol expansion ability. This paper first introduced the technologies such as SRv6 and P4 and many tools used in the implementation process. Then, the detailed design of SRv6 routing module based on P4 is introduced in detail, including the key technologies such as using P4 Match-Action abstraction to identify and parse SRv6 messages, add and delete SRv6 header, segment routing forwarding behavior based on SRH, as well as the automation and parameterization design of SRv6 segment routing to facilitate deployment and verification experiments, and the tracking packet path analysis method in virtual environment based on the secondary encapsulation of Mininet log function. Finally, the function of the designed SRv6 routing module is demonstrated by simulation experiments.

SDN technology brings the advantages of improving resource utilization and management efficiency to the network, and at the same time, it poses new challenges to the reliable operation of the network. In order to solve the problem of data forwarding error caused by network equipment attack, this paper proposes a multi-party cooperative routing management mechanism for power communication network. In order to achieve a reliable and efficient routing mechanism, a routing management platform architecture is designed according to the characteristics of the network. The architecture includes four types of devices: blockchain, centralized management center, SDN controller and repeater. In the routing table generation stage, a blockchain-based routing table audit mechanism is proposed. In the routing table execution phase, a routing table detection mechanism based on active detection is proposed. In the performance analysis, from the two dimensions of blockchain management routing table, centralized management center and SDN controller collaboration, it is verified that the mechanism in this paper has good performance in improving routing reliability and efficiency.

With the rapid development of the Internet of Things, numerous security and trust risks, including a single point of failure for authorities, data tampering, unauthorized users decrypting data, and challenges with searching encrypted data, have emerged on the centralized Internet of Things data sharing platform. As a novel technology, blockchain offers the benefits of decentralization, tamper-proofing, and trusted data sharing. Therefore, this article suggests a blockchain-based searchable encryption access control method. Users can simply search the necessary data using keywords provided they comply with the access rules. To ensure the privacy of the data, the cloud will not get any privacy-related information. The simulation demonstrate that our study has a lower time overhead than the current investigations. Data encryption takes 33% less time than standard access control methods, decryption takes 5% less time, and searching takes 75% less time.

In order to solve the problems of the uncertainty of end-to-end delay caused by the separation of scheduling and routing in the existing scheduling strategies in time-sensitive networks and the poor convergence of traditional genetic algorithms, a traffic scheduling method based on particle swarm optimization algorithm (PSO) combined with genetic algorithm and route planning was proposed. The method combines the inherent characteristics of time sensitive traffic and allows for flexible allocation of traffic routing. On this basis, a scheduling constraint model with no-wait constraints and risk balanced routing is established to guarantee low latency requirements for time sensitive traffic scheduling in both temporal and spatial dimensions. The experimental results show that the proposed method shows good convergence in different topology scenarios. By using the route planning strategy and the improved genetic algorithm, the total transmission time of the time-triggered flow is reduced by about 7.92% compared with the traditional genetic algorithm.

The traditional federated learning framework heavily relies on a single central server, which leads to problems such as single-point failures and malicious attacks. The new-type power system brings diverse collaborative business needs of “generation-transmission-distribution-storage”. With the significant increase of sensing terminals of new-type power devices, the security protection of data generalization becomes more and more crucial, and the energy consumption of devices has become a critical bottleneck for current federated learning tasks. The DAG structure has inherent decentralization and asynchronous characteristics, which can greatly accelerate the speed of global aggregation in federated learning, and the complexity of the DAG network can ensure the security and reliability of the model. In this paper, we propose a DAG-based federated learning framework for energy-constrained new-type power systems. In order to solve the problems of energy loss and training delay in DAG-based federated learning, a resource allocation algorithm based on multi-objective differential evolution is proposed. The algorithm aims to consider the impact of device energy consumption on federated learning performance, so as to minimize the completion time and energy loss of federated learning tasks under the constraint of expected learning accuracy of edge devices in the smart grid.

With the accelerated transformation of the development of new power systems, the data interaction between the data center and the edge of the business system is increasingly frequent, which challenges the reliability and privacy security of the data in the interaction process. The existing data interaction schemes based on data anonymity usually rely on the trusted third party (TTP). However, the TTP has the risk of a single point of failure and internal leakage. At the same time, the terminal can easily be hijacked and become a medium to disrupt system decisions maliciously dur to the lack of security protection of edge networks. To solve these problems, this paper proposes a reliable data exchange scheme based on the oblivious transfer (OT) technology in the smart grid. Experiments show that this scheme can achieve the protection of user privacy without excessive overhead while getting rid of the dependence on the TTP.

Compared with traditional Ethernet, FlexE network has the advantages of flexible and variable multi-grain rates, decoupling from the optical transmission capability, IP and optical convergence networking, and enhanced QoS for multi-service bearers. In this context, an algorithm (JLRB algorithm) is designed in this paper for FlexE networks with high traffic services. This paper first introduces the FlexE network technology and the research background, and then two factors: load balancing and risk balancing to be considered and optimized by the algorithm are analyzed. In the next section, the expressions of each parameter in the algorithm are defined and the specific procedure of the algorithm is designed. In the simulation experiments, the performance of the proposed algorithm is tested and compared with the comparison algorithm (LRWS algorithm). The superiority of the designed algorithm is proved by the comparison experiments with the comparison algorithm.

Network fault prediction is one of the important means to ensure network security and stable operation. Efficient fault prediction can improve the ability of operation and maintenance personnel to deal with faults and reduce losses caused by faults. In edge scenarios, device resources may be limited by hardware resources, such as storage space, memory, processing power. They may also be limited by unstable network connections, such as limited bandwidth, high packet loss rate, and large delay. This paper investigates the research status of network fault prediction at home and abroad. Currently, commonly used network fault prediction methods include methods based on statistics, based on machine learning, and based on deep learning. These network fault prediction methods can learn the characteristics of network faults and have achieved good results in network fault prediction tasks. However, the methods based on neural networks have a large computational resource overhead and are easily limited by device performance in edge scenarios. The methods based on statistics and machine learning have low cost but low accuracy. In this paper, an edge side network fault prediction model based on improved BiLSTM is designed, and improve the continuous distillation technology to design Stage Continuous Knowledge Distillation (SCKD). The simulation experiments prove that the student model performs similarly to the teacher model in terms of accuracy and F1-Score, and has lower memory usage and parameter volume.

SDN technology realizes the separation of network control and data forwarding, and improves the efficiency and resource utilization of network communication. However, the current research lacks the research on network testing mechanism in SDN environment from the perspective of service quality, resulting in higher probability of network service problems and longer service repair time. In order to solve this problem, this paper first designs a network test architecture oriented to service quality under SDN environment. Secondly, the power communication network test mechanism oriented to service quality under SDN environment is designed. This mechanism includes six steps: the business management system discovers business exceptions, the test management global center triggers the test, the test management global center sends the test task to the test domain management center, the test domain management center sends the test request to the SDN domain controller, the SDN domain controller sends the flow table and feeds back the results to the test domain management center, and the test management global center collects the test domain management center results and analyzes them. Finally, the test mechanism proposed in this paper is analyzed from the two dimensions of usability and enforceability, which verifies that this mechanism has good performance.

In the context of service slicing, network resource allocation has become a research focus. To reduce the energy consumption of power communication networks, this paper proposes a service slice resource allocation algorithm based on node capability in power communication network. This algorithm adopts a strategy of simultaneous node mapping and link mapping for resource allocation. Use breadth first algorithm for each virtual node to allocate resources for the virtual node and its connected links. When allocating resources to virtual nodes, based on the historical mapping experience of the underlying network, priority is given to opening the underlying node that has been mapped the most times. When allocating resources for virtual links, the shortest path algorithm is used to select the underlying links that have already been mapped to the underlying links from multiple paths, thereby reducing the energy consumption of the underlying link resources. By comparing existing algorithms, it has been verified that this algorithm saves energy consumption on underlying network resources and improves the success rate of virtual network mapping.

In the dynamic environment of big data and cloud computing, image feature classification has become a key factor spanning various fields. Ensuring the security, privacy, and computational efficiency of image data, while minimizing the processing of image data and maintaining the effectiveness of encrypted classification, is a significant challenge. In this paper, we propose a new method, Homomorphic Encryption Image Classification Evaluation (HEICE), for secure image classification. This method leverages the power of Convolutional Neural Networks (CNNs) and the security of Homomorphic Encryption (HE) to perform image classification on encrypted data. Each model uses different activation functions: square function, polynomial approximation of ReLU, polynomial approximation of Sigmoid and Tanh, and a piecewise linear approximation. These modified models are then used to test encrypted images, and the results are compared with the baseline. This method allows us to evaluate the performance of different activation functions when processing encrypted data and to choose the most suitable model for image classification, i.e., the classification model with the square function as the activation function. Our method provides a systematic approach to address the challenge of ensuring model performance while maintaining data security in image classification. This comparison validates the effectiveness of our method in achieving the dual objectives of maintaining data privacy and achieving accurate image classification.

Compared with the traditional blockchain, the blockchain system based on directed acyclic graph (DAG) has higher throughput and greater storage pressure, and there is also redundancy of transaction data in the block, which is caused by concurrent block sending, that is, the same transaction may appear in different blocks, and different blocks are attached to the DAG at the same time, which aggravates the storage pressure. In this paper, to solve the above problems, we propose a method that can reduce data twice. The first data reduction is aimed at the redundancy in the block, which is in the blockchain system based on DAG. And the second data reduction is based on the user's experience and the first basis. The experimental results show that the proposed method can save 92.18% of the storage space and effectively alleviate the storage pressure.

A compact dual-band directional button antenna is proposed for miniaturized equipment and tight space in integrated power communication networks. In this design, a metamaterial lens consisting of $$2\times 2$$ 2 × 2 cells is applied to the radiation direction, enabling a compact and directional configuration. The antenna consists of a microstrip line feed, a monopole loaded by the dielectrics, and a metamaterial lens placed on the other side of the feed. The metamaterial unit cell has two different split resonant rings responsible for two different resonances, and the $$2\times 2$$ 2 × 2 cells are employed to form the desired lens and are symmetrical on the monopole. The results show that the antenna with lens has the similar properties with that of the reflector metamaterial.

Recently, the integrated power communication network has gained considerable attention, because of new differentiated business. To improve the energy efficiency (EE) for information acquisition services, reconfigurable intelligent surface (RIS) is proposed. Under the constraints of the minimum rate of each user, the maximum transmit power limit of the base station and the unit modulus constraint of the phase angle of RIS, this paper aims for maximizing EE in RIS-aided multiple-input-single-output (MISO) systems. Firstly, the maximum signal-to-interference-noise ratio (SINR), the total power consumption of the system and the phase matrix of RIS are analyzed and derived, and then the optimization problem is established with the beamforming and transmission power of the base station transmitting multi-user as variables. Thirdly, in order to solve the optimization problem, this paper proposes to use the maximum ratio to send pre-coding to maximize the SINR received by users, and uses an improved sine and cosine optimization algorithm to optimize the phase matrix of RIS. Finally, a scheme of alternating iterative optimization of phase matrix and power is designed. Simulation results show that the proposed algorithm is very effective in improving the energy efficiency of the system. Compared with the traditional relay amplification scheme, the improved SCA optimization scheme achieves about 30% improvement in energy efficiency, which confirms the feasibility of the proposed method in improving the energy efficiency of MISO system.

With the rapid growth of internet traffic, network congestion becomes more and more severe, which causes massive packets loss. The reliability and delay of data transmission needs to be guaranteed in real-time applications such as financial transactions and cloud games. Traditional transmission strategies use packet retransmission to ensure data reliability, but the retransmission causes extra delay. The extra delay reduces the quality of service (QoS) for deterministic services. For the above problem, this paper proposes a deterministic network (DetNet)-oriented multipath transmission strategy in the software-defined network (SDN) architecture. The architecture introduces the packet replication and elimination function (PREF) of DetNet to achieve reliable transmission. The strategy establishes a path optimization model with transmission delay and packet loss rate, and then solve the model by the Q-learning algorithm. The delay and packet loss rate of the link construct the reward function. The Q-value table enables to obtain the best combination of paths, and it is gained by the reward function. Simulations show that our strategy have lower packet loss rate and delay jitter than the traditional single-path transmission strategy and the multi-path transmission strategy.

Aiming at the problem of communication redundancy and delay for grid information authentication in the existing consortium blockchain network, a new architecture for optimizing the broadcast path by controlling the topology of the consortium blockchain by software defined networking (SDN) is proposed. First, the SDN controller can collect the information and status of nodes in the consortium blockchain network in real time, and make topology adjustments accordingly. Secondly, some existing practices reduce the redundancy and delay brought by the Gossip protocol, but also bring serious computational burden to the nodes. Therefore, our proposed architecture is divided into two layers. The task of adjusting the topology structure is assigned to the upper-SDN controller layer, and the normal operation of the consortium blockchain network is assigned to the lower-blockchain layer, which not only solves the problem of limited computing power of grid information authentication nodes, but also makes the consortium blockchain network flexible, stable and easy to expand. Finally, the simulation results show that the topology adjustment made by SDN effectively reduces the communication redundancy and delay brought by the Gossip protocol in the consortium blockchain network and improves the consensus efficiency.

Tor, as a widely used anonymous communication system, is frequently employed by some users for illegal activities. Snowflake server as a plugin that enables users to connecting to the Tor network, allowing users to evade surveillance by connecting to the Tor network through it. Since Snowflake hides user traffic within regular WebRTC, it becomes challenging for authorities to differentiate and regulate, posing significant difficulties in monitoring efforts. To address these issues, this paper proposes a feature extraction method based on traffic statistical characteristics and a Snowflake traffic identification model based on MLP. We collected traffic datasets in Docker environment, extracted variable-length DTLS handshake sequences, and employed the feature extraction method to extract their statistical characteristics, including packet length, session duration, and average time between sending two packets, among other features. The MLP-based Snowflake traffic identification model can determine whether the traffic belongs to the target traffic based on these features. Moreover, this method can accurately identify traffic even when the traffic fields change. Experimental results demonstrate that this method achieves a 99.83% accuracy rate in identifying Snowflake traffic. Additionally, even when the data distribution in the dataset is altered, although the method requires more training iterations, it still achieves a 99.67% accuracy rate.

As machine learning has gradually become an important technology in the field of artificial intelligence, its development is also facing challenges in terms of privacy. This article aims to summarize the attack methods and defense strategies for machine learning models in recent years. Attack methods include embedding inversion attack, attribute inference attack, membership inference attack and model extraction attack, etc. Defense measures include but are not limited to homomorphic encryption, adversarial training, differential privacy, secure multi-party computation, etc., focusing on the analysis of privacy protection issues in machine learning, and providing certain references and references for related research.

Metaverse is the advanced form of Internet technology development and the expansion and extension of cyberspace in the era of digital economy. In the meta-universe, data, computing power, algorithms and other elements integrate and promote each other, giving birth to new business models and application scenarios, changing people's production and life style and social governance model. In this paper, the concept, characteristics and development history of the meta-universe security and forensics are described, and the application scenarios and technical framework of the meta-universe are introduced. At the same time, this paper takes the forensics of meta-universe devices smart bracelet and smart TV as an example, analyzes the current problems of meta-universe security and forensics, and puts forward corresponding countermeasures for the current security problems of meta-universe forensics.

At present, smart contracts cannot guarantee absolute security, and they have exposed many security issues and caused incalculable losses. Due to the existence of these security vulnerabilities, researchers have designed many detection and classification tools to identify and discover them. In this article, we present a classification of smart contract security vulnerabilities based on a large number of detailed articles. Then, we introduce the latest smart contract vulnerability detection methods, summarize the process model of detection tools based on artificial intelligence methods, and compare and analyze various detection tools. Finally, we provide an outlook on future research directions based on the current status of smart contract security.

With the booming development of blockchain technology, blockchain-based data transactions have been applied in many fields such as finance, healthcare and logistics. It can help users to realize data transactions and management more conveniently, securely, transparently and efficiently. However, there is a certain problem of identity privacy leakage when data transactions are conducted on blockchain. Therefore, the issue of user identity privacy protection has become the core issue of data transactions on the blockchain, which is crucial to the sustainable development and wide application of the blockchain. This paper discusses the privacy protection in the process of data transactions on blockchain in terms of user identity anonymity, introduces and analyzes in detail the current research status and implementation technologies for realizing identity anonymity on blockchain, explains the threats and challenges for realizing identity anonymity, analyzes the existing problems, and gives an outlook and summary of the future research directions for realizing identity anonymity on blockchain.

In this information age, real currency is transitioning to digital currency. Blockchain technology has introduced smart contracts and distributed ledger technology, leading to new research directions in various fields including finance, the Internet of Things, and energy. Central Bank Digital Currency (CBDC) based on blockchain has been incorporated into the technical choices of central banks in various countries. This paper examines the current research and implementation of central bank digital currency by various central banks worldwide. It provides an overview of CBDC's definition, operational status, and technical characteristics. Additionally, this paper highlights the challenges and issues associated with CBDC systems in the current environment. These problems include security issues, performance issues, privacy protection issues, and legal issues. In this regard, some literature has proposed solutions, which are analyzed and summarized in this paper. We classify and summarize the existing problems and propose effective solutions, including using fragmentation technology to increase transaction throughput and implementing corresponding regulatory measures to strengthen supervision.

Smart contract is an automated contract system based on blockchain technology, which is self-executing, tamper-evident and decentralized. The writing and analysis of smart contracts still face several challenges, including complex programming languages and potential security vulnerabilities. Natural Language Processing (NLP) as a discipline that studies the interaction between natural language and computers, can provide strong support for the development and analysis of smart contracts. This paper explores the cross-application of blockchain, smart contracts and NLP. First, this paper introduces the basic principles of blockchain technology and the concept of smart contracts. Then it points out the problems in the development process of smart contracts, and focuses on the analysis and summary of the relevant research results of NLP technology in the generation of smart contract code and annotation generation, and summarizes and analyzes the important role of NLP technology on the efficiency of smart contract development, the correctness, reliability, readability, and maintainability of the code. Secondly, for the security of smart contracts, the research related to smart contract vulnerability detection using NLP technology is summarized. Finally, the advantages, challenges and future development directions of combining natural language processing with blockchain smart contracts are pointed out to provide reference and inspiration for research and application in related fields.