The application of artificial intelligence technology in Internet of Vehicles(lov)has attracted great research interests with the goal of enabling smart transportation and traffic management.Meanwhile,concerns have be...The application of artificial intelligence technology in Internet of Vehicles(lov)has attracted great research interests with the goal of enabling smart transportation and traffic management.Meanwhile,concerns have been raised over the security and privacy of the tons of traffic and vehicle data.In this regard,Federated Learning(FL)with privacy protection features is considered a highly promising solution.However,in the FL process,the server side may take advantage of its dominant role in model aggregation to steal sensitive information of users,while the client side may also upload malicious data to compromise the training of the global model.Most existing privacy-preserving FL schemes in IoV fail to deal with threats from both of these two sides at the same time.In this paper,we propose a Blockchain based Privacy-preserving Federated Learning scheme named BPFL,which uses blockchain as the underlying distributed framework of FL.We improve the Multi-Krum technology and combine it with the homomorphic encryption to achieve ciphertext-level model aggregation and model filtering,which can enable the verifiability of the local models while achieving privacy-preservation.Additionally,we develop a reputation-based incentive mechanism to encourage users in IoV to actively participate in the federated learning and to practice honesty.The security analysis and performance evaluations are conducted to show that the proposed scheme can meet the security requirements and improve the performance of the FL model.展开更多
Panzhihua city （26°O5＇-27°21＇N, 101°OS＇- 102°15＇E）, located in a mountainous area, is one of the large cities in Sichuan province, China. A landslide occurred in the filling body of the easte...Panzhihua city （26°O5＇-27°21＇N, 101°OS＇- 102°15＇E）, located in a mountainous area, is one of the large cities in Sichuan province, China. A landslide occurred in the filling body of the eastern part of the Panzhihua airport on October 3, 2009 （hereafter called the lo.3 landslide）. We conducted field survey on the landslide and adopted emergency monitoring and warning models based on the Internet of Things （loT） to estimate the losses from the disaster and to prevent a secondary disaster from occurring. The results showed that four major features of the airport site had contributed to the landslide, i.e, high altitude, huge amount of filling rocks, deep backfilling and great difficulty of backfilling. The deformation process of the landslide had six stages and the unstable geological structure of high fillings and an earthquake were the main causes of the landslide. We adopted relative displacement sensing technology and Global System for Mobile Communications （GSM） technology to achieve remote, real-time and unattended monitoring of ground cracks in the landslide. The monitoring system, including five extensometers with measuring ranges of 200, 450 and 7oo mm, was continuously working for 17 months and released 7 warning signals with an average warning time of about 26 hours. At 10 am on 6 December 2009, the system issued a warning and on-site workers were evacuated and equipment protected immediately. At 2：20 medium-scale collapse monitoring site, which proved the reliability pm on 7 December, a occurred at the No. 5 justified the alarm and and efficiency of the monitoring system.展开更多
为及时隔离局域网内易受攻击的异常物联网设备,对网络管理员而言,具备高效的设备分类识别能力至关重要。现有方法中所选择的特征与设备关联性不高,且设备状态的差异会导致样本数据不平衡。针对上述问题,文中提出了一种基于流量和文本指...为及时隔离局域网内易受攻击的异常物联网设备,对网络管理员而言,具备高效的设备分类识别能力至关重要。现有方法中所选择的特征与设备关联性不高,且设备状态的差异会导致样本数据不平衡。针对上述问题,文中提出了一种基于流量和文本指纹的物联网设备分类识别模型FT-DRF(Flow Text-Double Random Forest)。首先设计特征挖掘模型,选取稳定的流统计数据作为设备流量指纹;其次基于HTTP,DNS和DHCP等应用层协议头部字段中的敏感文本信息生成设备文本指纹;在此基础上,对数据进行预处理并生成特征向量;最后,设计基于双层随机森林的机器学习算法对设备进行分类识别。对由13个物联网设备组成的模拟智能家居环境数据集和公共数据集进行有监督分类识别实验,结果表明,FT-DRF模型能够识别网络摄像头、智能音箱等物联网设备,平均准确率可达99.81%,相比现有典型方法提升了2%~5%。展开更多
基金supported by the National Natural Science Foundation of China under Grant 61972148.
文摘The application of artificial intelligence technology in Internet of Vehicles(lov)has attracted great research interests with the goal of enabling smart transportation and traffic management.Meanwhile,concerns have been raised over the security and privacy of the tons of traffic and vehicle data.In this regard,Federated Learning(FL)with privacy protection features is considered a highly promising solution.However,in the FL process,the server side may take advantage of its dominant role in model aggregation to steal sensitive information of users,while the client side may also upload malicious data to compromise the training of the global model.Most existing privacy-preserving FL schemes in IoV fail to deal with threats from both of these two sides at the same time.In this paper,we propose a Blockchain based Privacy-preserving Federated Learning scheme named BPFL,which uses blockchain as the underlying distributed framework of FL.We improve the Multi-Krum technology and combine it with the homomorphic encryption to achieve ciphertext-level model aggregation and model filtering,which can enable the verifiability of the local models while achieving privacy-preservation.Additionally,we develop a reputation-based incentive mechanism to encourage users in IoV to actively participate in the federated learning and to practice honesty.The security analysis and performance evaluations are conducted to show that the proposed scheme can meet the security requirements and improve the performance of the FL model.
基金supported by the National Science Fund for Distinguished Young Scholars of China (Grant No. 40125015)a Research Project of the State Key Laboratory of Geohazard Prevention and Geoenvironment Protection (Grant No. SKLGP2010Z002)+1 种基金the Science and Technique Plans for Sichuan Province, China (Grant No. 2011SZ0182 and NO. 2013SZ0168)the Fundamental Science on Nuclear Waste and Environmental Security Laboratory (Grant No. 12zxnp04)
文摘Panzhihua city （26°O5＇-27°21＇N, 101°OS＇- 102°15＇E）, located in a mountainous area, is one of the large cities in Sichuan province, China. A landslide occurred in the filling body of the eastern part of the Panzhihua airport on October 3, 2009 （hereafter called the lo.3 landslide）. We conducted field survey on the landslide and adopted emergency monitoring and warning models based on the Internet of Things （loT） to estimate the losses from the disaster and to prevent a secondary disaster from occurring. The results showed that four major features of the airport site had contributed to the landslide, i.e, high altitude, huge amount of filling rocks, deep backfilling and great difficulty of backfilling. The deformation process of the landslide had six stages and the unstable geological structure of high fillings and an earthquake were the main causes of the landslide. We adopted relative displacement sensing technology and Global System for Mobile Communications （GSM） technology to achieve remote, real-time and unattended monitoring of ground cracks in the landslide. The monitoring system, including five extensometers with measuring ranges of 200, 450 and 7oo mm, was continuously working for 17 months and released 7 warning signals with an average warning time of about 26 hours. At 10 am on 6 December 2009, the system issued a warning and on-site workers were evacuated and equipment protected immediately. At 2：20 medium-scale collapse monitoring site, which proved the reliability pm on 7 December, a occurred at the No. 5 justified the alarm and and efficiency of the monitoring system.
文摘为及时隔离局域网内易受攻击的异常物联网设备,对网络管理员而言,具备高效的设备分类识别能力至关重要。现有方法中所选择的特征与设备关联性不高,且设备状态的差异会导致样本数据不平衡。针对上述问题,文中提出了一种基于流量和文本指纹的物联网设备分类识别模型FT-DRF(Flow Text-Double Random Forest)。首先设计特征挖掘模型,选取稳定的流统计数据作为设备流量指纹;其次基于HTTP,DNS和DHCP等应用层协议头部字段中的敏感文本信息生成设备文本指纹;在此基础上,对数据进行预处理并生成特征向量;最后,设计基于双层随机森林的机器学习算法对设备进行分类识别。对由13个物联网设备组成的模拟智能家居环境数据集和公共数据集进行有监督分类识别实验,结果表明,FT-DRF模型能够识别网络摄像头、智能音箱等物联网设备,平均准确率可达99.81%,相比现有典型方法提升了2%~5%。
