Cognitive Networks: Towards Self-Aware Networks

12.3. Threat Model

Since cognitive networks are constantly evolving to ensure trust and performance, technologies developed in the intrusion detection area can be of help to solve some of these problems. Conversely, the inherently changing aspects of cognitive networks create new challenges for such technologies.

When looking at a security problem, it is necessary to assess the threat before prevention, detection and reaction mechanisms can be proposed. One possible approach to determining threats against cognitive networks is to recognize that the field of autonomic networking is also looking at the perspective of having systems and networks that include self-protecting and self-healing functionalities. A general description of the security aspects of autonomic networking can be found in ^[]; thereafter, we will attach ourselves to the impact, target and attacker aspects of the threats.

12.3.1. Model and Constraints

We consider that a cognitive network consists of two main parts: content and transport. Content is related to the services that the user obtains from the network, such as exchanging mail or having voice conversations with other users. Transport is related to the way this content is transferred from one point of the network to another. Note that in many cases, administrative information will also be transported by the network; thus, for network operators, the management content must ensure that the user content properties are preserved.

We do not believe that threats ...