ü H V E R S I 0 A D
Malaga, 26 April 2010
Evaluation of Thesis Proposal
'Privacy preserving protocols for wireless sensor networks' Author: Jiri Kur
The deployment of pervasive computing technologies demands higher levels of privacy because individuals seem to be eager to exchange information with others, though do not pay much attention to the security risks involved. Hence, the design of new technologies must take privacy concerns into account, and wireless sensor networks is one of the most challenging scenarios from this perspective. This is precisely the interesting area where the thesis focus.
The PhD student is fully aware of the most relevant results which have appeared in the literature during the last few years in the field of privacy protocols for wireless sensor networks. He has gathered information concerning different types of contextual privacy, both the location privacy of senders and receivers and temporal privacy.
Also, he is aware of the privacy problem due to the use of the static naming of nodes. In general, the applicant seems to have a good understanding of the techniques proposed in these works and that privacy is an application-dependent problem that might have different implications depending on the context or the nature of the information to be protected.
Since WSNs are very resource limited technologies, it is interesting to have different modes of operation in order to reduce the amount of energy waste due to the performance of privacy-aware protocols. However, shifting between different modes of operation might be also an indicator to the attacker of abnormal circumstances occurring the nearby the network. If the WSN enters a privacy-preserving mode, an attacker might infer that an important asset has approached the network. Even in the case of existing various modes of operation with different levels of protection, the use of any of them can be treated as a clear indicator of the importance of the resource being monitored. Thus, the shifting between different modes is an issue that should be tackled with attention.
Also, the type of adversary that the network must face to is a fact to take into consideration. Traditionally, in privacy-aware protocols for WSNs, the attacker has been considered to be passive which are limited to overhearing the communications. However, security literature in WSNs usually deals with more severe adversaries which
are able to perform more sophisticated attacks. Considering the presence of these new types of attackers is an interesting and appealing area of research for the privacy protocols. This might be reflected in the definition of the metrics for the evaluation of the protocols with respect to the level of privacy protection offered.
According to the proposed working plan, the ordering of the objectives is rather coherent. On the one hand, the work load might be insufficient for the completion of the various objectives (2 years), taking into account that four different application scenarios are considered for each of the proposed objectives. However, on the other hand, the application scenarios that are presented have several features in common between each other. In particular, scenarios 1 and 4 look very similar and the same occurs between scenarios 3 and 4.
Summarizing, the document includes a plan that shows a very good potentiality of the PhD work, while the PhD student seems to be very well aware of the previous work in the literature and to be well prepared for the work still to be done. Hence, I fully recommend that the proposal is accepted.
Javier Lopez
Computer Science Department University of Malaga