In the Internet-of-Things, an increasing number of heterogeneous devices will be co-located
in the same area. Current networking approaches are not designed to support efficient direct communication between co-located devices from different networks. To optimize these
heterogeneous networks, ad hoc cooperation between co-located devices of different network
technologies is needed.
To this end, this project develops methodologies for cross-network, cross-technology
negotiation. To optimize the network performance of co-located networks in a global way, colocated devices take into account the network requirements of all devices when negotiating about cooperation opportunities (in the form of sharing network resources). As part of this project, the following challenges are tackled. (i) A network service architecture is developed, in which network functionality (such as interference avoidance, dynamic transmission power
control, shared routing, different routing protocols, etc.) can be added to devices at run-time.
(ii) Distributed network monitoring algorithms are developed. These algorithms are capable
of detecting co-located (heterogeneous) networks and can monitor how the activation of
network services influences the performance of multiple co-located networks that use
different communication technologies. (iii) Finally, negotiation algorithms based on
reinforcement learning are developed to select the optimal network services that should be
activated, based on the network requirements of the collaborating devices.
Ultimately, this project leads to autonomous algorithms capable of negotiating about optimal network optimization solutions that take into account the heterogeneity and dynamic
requirements of co-located devices.