The article deals with the problem of defining areas of cyberspace where the state has the right to apply various methods of protection. The "digital border" is a key concept for ensuring "digital sovereignty". The "digital border" is also a necessary condition for ensuring "digital sovereignty". Different definitions of the "digital boundary" are considered in the article. For each type of "digital boundaries", the reasons for their occurrence, the technical and technological methods for their construction, and the effectiveness for ensuring "digital sovereignty" are described. Providing "digital sovereignty" involves using of authentication, authorization and audition. All these functions are used differently for various types of "digital" boundaries. A model of a distributed multi-level digital boundary is proposed. This model supposes the control of information flows at all levels of cyberspace. Technical and economic constraints are proposed to be solved through the distribution of information flow control functions over the network. Technological basis of such digital boundary can be made by software-defined networks, which allow to concentrate all tasks of analysis and network management in one center. Moreover, information collected in that center can be used during investigations of various incidents if they fall within the sphere of state responsibility.