Section: New Results

Interferences and failures management

Participants : Nathalie Mitton, Viktor Toldov, Valeria Loscri, Simon Duquennoy.

In the recent years, the Machine-to-Machine (M2M) paradigm together with the integration of wireless sensors networks with the generic infrastructure via $6LoWPAN$ require the implementation of ad hoc communication protocols at the Medium Access Control layer, that do not depend on pre-existing infrastructure. Channel hopping concept has more and more gained consensus as a viable and effective solution for wireless MAC layer coordination with time-synchronized channel hopping (TSCH). In [24] we propose a decentralized multichannel MAC coordination framework (DT-SCS) leveraging the concept of pulse-coupled oscillators at the MAC layer. In DT-SCS, nodes randomly join a channel and are automatically spread across the available channels. The nodes then achieve PCO-based coordination via the periodic transmission of beacon packets at the MAC layer. As such, for channels with an equal number of nodes, DT-SCS converges to synchronized beacon packet transmission at the MAC layer in a completely uncoordinated manner. In order to combat the well-know phenomenon of Cross-Technology Interference (CTI) a cross-layer mechanism, CrossZig, has been implemented in [39], based on the exploitation of information at the physical layer in order to detect the presence of CTI in a corrupted packet.

A different perspective of the interference management has been considered in [47] and [41], where a novel solution to allow to secondary users the access of allocated spectrum has been proposed. The study has been based on the major consideration that a big bottleneck in cognitive radio systems is based on finding the best available channel as fast as possible.

A totally different approach to face the enormous quantity of data generated by IoT devices, is to try to reduce the sending of useless data, based on the adoption of effective predictive approaches.

In [50] we have considered the concept of high spatio-temporal correlated data and we have proposed a Belief Propagation (BP) algorithm to derive methods to drastically reduce the number of transmitted messages, by keeping an high accuracy in terms of global information.

Together with interference management approaches it is also important to figure out tools to support network operator for mitigation of the impact of failures on their infrastructures. The need of advanced Network Planning and Management Tool (NPMT) has been considered in [30].