Section: New Results


Anonymous networks monitoring

Participants : Thibault Cholez [contact] , Isabelle Chrisment, Olivier Festor.

In 2015, we pursued our collaboration with Juan Pablo Timpanaro a former team's PhD student and published a new paper [47] on the I2P anonymous network (http://i2p2.de ). More precisely, we monitored I2P's decentralised directory, known as the netDB, and produced two contributions. On the one hand, we conducted arguably the first churn study of the I2P network, showing that I2P users are more stable than non-anonymous peer-to-peer users. On the other hand, we analysed the design of the netDB and compared it against the popular KAD design, demonstrating that the former is more vulnerable to different attacks, specially to Eclipse attacks, which can be mitigated by applying some safer design choices of the latter. We lately showed the positive impact on performance of including KAD's DHT configuration into the netDB in terms of bandwidth, storage and messages overhead.

Smartphone usage monitoring

Participants : Vassili Rivron [contact] , Mohammad Irfan Khan, Simon Charneau [Inria] , Isabelle Chrisment.

In [39] we presented some results from our study based on a combination of crowdsending and survey. We discussed some technical problems we faced and some lessons learned during our crowdsensing experiment. Furthermore we showed how information regarding social context can be used for better interpretation of crowdsensed data. Next we selected some questions from the multiple choice survey questionnaire and combined the responses with crowdsensed data to analyze users’ perception about their smartphone usage and discussed cognitive factors associated with reporting information on questionnaires. Moreover we showed that combining sensing with survey can improve both the techniques and the combination has important use cases such as helping users to have a better understanding and control of their technology usage.

Active Monitoring

Participants : Abdelkader Lahmadi [contact] , Jérôme François, Valentin Giannini, Frederic Beck [LHS] , Bertrand Wallrich [LHS] .

The main motivation of this work was to assess the exposition of industrial systems in the Internet, especially by measuring how many SCADA systems are accessible. To do so, we built an IPv4 methodology which is able to scan the entire IPv4 address space by maximizing the distance between consecutive IP addresses. It thus avoids colateral effect of overloading targeted networks and being blacklisted. We thus extend the Zmap tool (zmap.io ) by also including other functionalities such as distributed scans, indexation and visualisation of the results [63] . First experiences have been performed and are under evaluation.

Sensor networks monitoring

Participants : Rémi Badonnel, Isabelle Chrisment, Olivier Festor, Abdelkader Lahmadi [contact] , Anthea Mayzaud.

This year, our work on security-oriented monitoring has been centered on building a distributed architecture that supports passive monitoring in the Internet of Things using the RPL protocol [37] . A particular interest has been given to advanced metering infrastructure (AMI) networks, where higher order devices are expected to form the backbone infrastructure, to which more constrained nodes would connect. Our distributed architecture exploits the capabilities of these higher order devices to perform network monitoring tasks, and takes benefits from properties inherent to that protocol, such as DODAG building and multi-instance routing mechanisms, in order to passively monitor the environment with a minimal impact on constrained nodes.

We have also consolidated our taxonomy on security attacks in these networks [8] . In addition, we have pursued our work on topological inconsistency attacks [9] . It is evident from the experiments that we have conducted that mitigating such attacks is critical to avoid channel congestion and high resource usage. Our initial adaptive threshold (AT) strategy to mitigate the effects of such attacks has been further improved. The new strategy dynamically takes into account network characteristics in order to infer an appropriate threshold for counteracting these attacks.