Sonntag, 4. März 2018

Science of Security (SoC) for Information Hiding – Some Short Notes

In recent years, my research focus shifted slightly towards the domain which is called science of security (SoC). In particular, I studied which scientific fundamentals are lacking in the research discipline we refer to as network information hiding (NIH). The core part of NIH is about hiding something in something else, and detecting it's presence. A well-known major component of NIH is steganography, the ancient art of hiding a message in a cover object. For instance, a secret message can be hidden in a digital image. However, steganography also works with network traffic to create network covert channels and is a major part of today's NIH research.

Topic 1: Hiding Patterns
In recent years, especially after finishing my doctorate, I discovered several terminological inconsistencies in NIH. In many discussions with experts, I tried to unify the majority of these inconsistencies –– surely, there are some left to be unified in the future. One major publication of us was to present so-called hiding patterns. Patterns are re-occurring design principles. In the NIH domain, hiding patterns describe techniques that hide data. In comparison to the earlier (fine-granular) distinctions, these hiding patterns describe the idea of how data is hidden and thus can serve as an umbrella for several hiding techniques. It is like having all the buildings of a city and then categorizing them using the patterns `hospital', `residential building', `factory building' etc. For instance, there are several hiding techniques that embed secret data in unused areas of network data. These hiding techniques have different names and are designed for several different network protocols. However, they all share the same core idea, which is reflected by a pattern that we call `reseved/unused pattern'. In our works, we went through scientific publications published since the late 1980's to end up with a comprehensive collection of the available NIH hiding techniques. We defined our hiding patterns by searching for similarities in these hiding techniques. Then, we put all patterns that we found in a hierarchy to form a taxonomy. While doing that, it turned out that patterns can be derived from each other, i.e. sub-patterns of patterns exist. The pattern-based taxonomy provides a clear view on the complex world of hiding techniques.

Topic 2: A Method to Describe New Hiding Techniques
The next thing I noticed was that there existed no unified description of hiding methods. This means that there were tons of papers who explained new hiding methods in unique ways, rendering these methods hardly comparable. In a survey paper, we could finally show these differences and their extend. This means also that we could identify which details are covered or missed by the majority of papers. As a solution, we proposed a new unified description method. If new hiding methods will be described using this unified description method, they will – automatically – be much easier to compare.

Topic 3: Creativity Framework
In another work, we answered the question of how the novelty of a proposed hiding method can be evaluated in the scientific review process. It turned out that Psychology's understanding of creativity can be of help in such a situation. Creativity depends mostly on applicability and novelty of an idea. In our journal article, we proposed a creativity framework to evaluate the novelty of new hiding techniques. Mostly based on the question of whether they represent a new pattern or a variation of an old pattern and how applicable a new hiding technique is, the new research contribution can be sorted into a category of novelty. Our framework helps to prevent terminological inconsistencies by limiting the creation of too many new (but redundant) patterns. Of course, our approach will only work if our creativity framework will be applied by the scientific community.

Topic 4: NIH and Science 2.0
Finally, one recent paper deals with the question of whether NIH can profit from Open Science (or Science 2.0).

-More work of this kind is currently in progress or already under review. Feel free to contact me if you are interested in joining my work in this area.-

Keine Kommentare:

Kommentar veröffentlichen