Stefan Lindskog

On the functional relation between security and dependability impairments

Problems related to security and dependability/ reliability are still treated separately in many contexts. It has been shown that there is a considerable conceptual overlap, however, and an integrated framework to the two disciplines has already been suggested. This paper shows that there is also a conceptual overlap of impairments from these areas and suggests an integrated approach that clarifies the functional relation between these, both from dependability and security viewpoints. The overall objective is to arrive at a general and clear-cut framework that would describe how trustable (dependable, secure) a system is, regardless of the reason for its not being totally trustable. For example, it should be possible to treat a system failure caused by an intentional intrusion or a hardware fault using the same methodology. A few examples from real-world situations are given to support the suggested approach.

A Content-Independent Scalable Encryption Model

Various user surveys have shown that security will be a key business enabler for applications of future communication systems. In these, security requirements will vary heavily—from very low to extremly rigid. The aim of this paper is to demonstrate how security could be treated in a more flexible and dynamic way. A content-independent model for scalable encryption, which is based on a partial encryption paradigm, is proposed and described. The main advantage with this model is that the users are able to make tradeoffs between security and performance. In the paper, we descibe the design and implementation as well as an initial performance analysis of the proposed model.