Ethereum, currently the most widely utilized smart contracts platform, anchors the security of myriad smartcontracts upon its own robustness. Its foundational peer-to-peer network facilitates a dependable node connect...Ethereum, currently the most widely utilized smart contracts platform, anchors the security of myriad smartcontracts upon its own robustness. Its foundational peer-to-peer network facilitates a dependable node connectionmechanism, whereas an efficient data-sharing protocol constitutes as the bedrock of Blockchain network security.In this paper, we propose NodeHunter, an Ethereum network detector implemented through the application ofsimulation technology, which is capable of aggregating all node records within the network and the interconnectednessbetween them. Utilizing this connection information, NodeHunter can procure more comprehensive insightsfor network status analysis compared to preceding detection methodologies. Throughout a three-month period ofunbroken surveillance of the Ethereum network, we obtained an excess of two million node records along with overone hundred million node acquaintances. Analysis of the gathered data revealed that an alarming 49% or more ofthese node records were maliciously forged.展开更多
Information diffusion in online social networks is induced by the event of forwarding information for users, and latency exists widely in user spreading behaviors. Little work has been done to reveal the effect of lat...Information diffusion in online social networks is induced by the event of forwarding information for users, and latency exists widely in user spreading behaviors. Little work has been done to reveal the effect of latency on the diffusion process. In this paper, we propose a propagation model in which nodes may suspend their spreading actions for a waiting period of stochastic length. These latent nodes may recover their activity again. Meanwhile, the mechanism of forwarding information is also introduced into the diffusion model. Mean-field analysis and numerical simulations indicate that our model has three nontrivial results. First, the spreading threshold does not correlate with latency in neither homogeneous nor heterogeneous networks, but depends on the spreading and refractory parameter. Furthermore, latency affects the diffusion process and changes the infection scale. A large or small latency parameter leads to a larger final diffusion extent, but the intrinsic dynamics is different. Large latency implies forwarding information rapidly, while small latency prevents nodes from dropping out of interactions. In addition, the betweenness is a better descriptor to identify influential nodes in the model with latency, compared with the coreness and degree. These results are helpful in understanding some collective phenomena of the diffusion process and taking measures to restrain a rumor in social networks.展开更多
A novel image sequence-based risk behavior detection method to achieve high-precision risk behavior detection for power maintenance personnel is proposed in this paper.In this method,the original image sequence data i...A novel image sequence-based risk behavior detection method to achieve high-precision risk behavior detection for power maintenance personnel is proposed in this paper.In this method,the original image sequence data is first separated from the foreground and background.Then,the free anchor frame detection method is used in the foreground image to detect the personnel and correct their direction.Finally,human posture nodes are extracted from each frame of the image sequence,which are then used to identify the abnormal behavior of the human.Simulation experiment results demonstrate that the proposed algorithm has significant advantages in terms of the accuracy of human posture node detection and risk behavior identification.展开更多
通过研究网络的拓扑结构可以探索到丰富的知识,特别是网络中节点的邻居可以形成不同的邻居结构,而不同的结构蕴含着不同的意义,进而也有着不同的影响。实际上,邻居结构与节点的交互行为之间是互相影响、互为因果的。对三种最为普遍的邻...通过研究网络的拓扑结构可以探索到丰富的知识,特别是网络中节点的邻居可以形成不同的邻居结构,而不同的结构蕴含着不同的意义,进而也有着不同的影响。实际上,邻居结构与节点的交互行为之间是互相影响、互为因果的。对三种最为普遍的邻居结构进行分析,并提出结合深度学习的网络邻居结构影响力模型DNSI(neighbor structure influence based on deep learning)。通过对图片格式的网络数据提取特征,DNSI可以得到三种邻居结构影响力。分别在几个真实世界网络数据集上进行节点属性预测、类别中心度度量和用户行为预测等任务,实验结果表明该模型在绝大多数情况下具有优越性。展开更多
基金the National Key Research and Development Program of China(No.2020YFB1005805)Peng Cheng Laboratory Project(Grant No.PCL2021A02)+2 种基金Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies(2022B1212010005)Shenzhen Basic Research(General Project)(No.JCYJ20190806142601687)Shenzhen Stable Supporting Program(General Project)(No.GXWD20201230155427003-20200821160539001).
文摘Ethereum, currently the most widely utilized smart contracts platform, anchors the security of myriad smartcontracts upon its own robustness. Its foundational peer-to-peer network facilitates a dependable node connectionmechanism, whereas an efficient data-sharing protocol constitutes as the bedrock of Blockchain network security.In this paper, we propose NodeHunter, an Ethereum network detector implemented through the application ofsimulation technology, which is capable of aggregating all node records within the network and the interconnectednessbetween them. Utilizing this connection information, NodeHunter can procure more comprehensive insightsfor network status analysis compared to preceding detection methodologies. Throughout a three-month period ofunbroken surveillance of the Ethereum network, we obtained an excess of two million node records along with overone hundred million node acquaintances. Analysis of the gathered data revealed that an alarming 49% or more ofthese node records were maliciously forged.
基金supported by the National Natural Science Foundation of China(Grant Nos.61401015 and 61271308)the Fundamental Research Funds for the Central Universities,China(Grant No.2014JBM018)the Talent Fund of Beijing Jiaotong University,China(Grant No.2015RC013)
文摘Information diffusion in online social networks is induced by the event of forwarding information for users, and latency exists widely in user spreading behaviors. Little work has been done to reveal the effect of latency on the diffusion process. In this paper, we propose a propagation model in which nodes may suspend their spreading actions for a waiting period of stochastic length. These latent nodes may recover their activity again. Meanwhile, the mechanism of forwarding information is also introduced into the diffusion model. Mean-field analysis and numerical simulations indicate that our model has three nontrivial results. First, the spreading threshold does not correlate with latency in neither homogeneous nor heterogeneous networks, but depends on the spreading and refractory parameter. Furthermore, latency affects the diffusion process and changes the infection scale. A large or small latency parameter leads to a larger final diffusion extent, but the intrinsic dynamics is different. Large latency implies forwarding information rapidly, while small latency prevents nodes from dropping out of interactions. In addition, the betweenness is a better descriptor to identify influential nodes in the model with latency, compared with the coreness and degree. These results are helpful in understanding some collective phenomena of the diffusion process and taking measures to restrain a rumor in social networks.
基金supported by the project“Research and application of key technologies of safe production management and control of substation operation and maintenance based on video semantic analysis”(5700-202133259A-0-0-00)of the State Grid Corporation of China.
文摘A novel image sequence-based risk behavior detection method to achieve high-precision risk behavior detection for power maintenance personnel is proposed in this paper.In this method,the original image sequence data is first separated from the foreground and background.Then,the free anchor frame detection method is used in the foreground image to detect the personnel and correct their direction.Finally,human posture nodes are extracted from each frame of the image sequence,which are then used to identify the abnormal behavior of the human.Simulation experiment results demonstrate that the proposed algorithm has significant advantages in terms of the accuracy of human posture node detection and risk behavior identification.
文摘通过研究网络的拓扑结构可以探索到丰富的知识,特别是网络中节点的邻居可以形成不同的邻居结构,而不同的结构蕴含着不同的意义,进而也有着不同的影响。实际上,邻居结构与节点的交互行为之间是互相影响、互为因果的。对三种最为普遍的邻居结构进行分析,并提出结合深度学习的网络邻居结构影响力模型DNSI(neighbor structure influence based on deep learning)。通过对图片格式的网络数据提取特征,DNSI可以得到三种邻居结构影响力。分别在几个真实世界网络数据集上进行节点属性预测、类别中心度度量和用户行为预测等任务,实验结果表明该模型在绝大多数情况下具有优越性。
