Secure data flow in networks and in the Internet of things

点击翻译按钮获取中文摘要

Organizations and data networks, such as those found in the Cloud and the Internet of Things, often have complex and evolving data flow paths, which may be important to control, for secrecy (also called confidentiality) and privacy. Supposing that certain entities contain some data, and are connected in certain ways to other entities by communication channels, where can the data end up? Given certain restricted data, what are the entities to which they are privy? How can we configure communication channels so that only desired data flows are possible? How can data flows be controlled, by the owners of the data or by security administrators in the organizations, according to requirements, such as network or organizational policies? How can we deal with the continuous updates and transformations of the data flow structure, as well as changing requirements? The applicant has recently identified new principles for data flow security. These principles make it possible to answer, theoretically and practically, and by using efficient algorithms, important questions such as: a) given an existing data communication channel structure in a network, what are the most appropriate areas in the network where secret data should be placed; b) given certain secrecy requirements, how to configure channel structures capable of enforcing the requirements. This configuration leads to Multilevel models which are generalizations of the traditional ones, but which are more flexible and can be proved to be necessary and sufficient to guarantee data secrecy. What has been done is only a beginning, because very flexible and dynamic solutions are needed, namely for the Internet of Things or the Cloud. Many types of environments exist, and each environment has its particular secrecy needs: hospital networks, home networks, e-commerce networks, industrial networks, intelligent environments etc. Beyond this, today's data networks are worldwide, and characterized by rapid mutability. Some mutations can be inconsequential, others will need various degrees of security assurance. Developing these ideas is a main goals of this research project. As an application of these results we are planning to show how to use the principles that we are developing in the field of data secrecy for legal applications, namely smart contracts in the IoT. Blockchain implements data authentication, but not data flow constraints. We plan to show how to assure desired data flow properties in smart contracts. This research will be important for developers of IoT systems, as well as for developers of software tools to help them. The research outcomes will be principles and prototype tools to help such specialists in their work. This research is important to Canada, since our country is positioning itself as a provider of Internet services, in areas such as e-commerce, medical systems, transportation systems, etc.