A platform for aggregate computing over LoRaWAN network

Recent technological developments led to increased computational and networking capabilities of everyday objects. This situation resulted in an increase in number of devices embedded in cyber-physical systems. In order to simplify the design and management of pervasive and heterogeneous systems like these, there is need for new high-level paradigms able to capture concerns like heterogeneity and location of the devices. Aggregate computing is one of these: it proposes to describe the global behaviour of a system by managing global spatio-temporal data structures, and abstracting details of its physical network, as topology and communication technology. A related problem with the design of complex pervasive systems is verifying their behaviour in a real scenario, because it is generally expensive, complicated, and not always possible in practice. A partial solution to the problem is testing this kind of systems using simulations. Even though simulations execute a system model, it should be noted that such model is only a system abstraction; however they can still provide reliable insights on the system behaviour and performance. In the Internet-of-Things context, an emergent enabling communication technology for situated devices is LoRaWAN. LoRaWAN is a network protocol that allows long range communications and low energy consumption, at the cost of limited data rate. There are currently no platforms for aggregated languages that support their execution over LoRaWAN networks. Moreover nowadays there are no simulators supporting real simulation of aggregate system over LoRaWAN networks: however there are simulators supporting aggregate applications or LoRaWAN networks. The contribution of this thesis is to provide a platform that supports the LoRaWAN abstractions as backend of an aggregate computing system, and join it to the existing DingNet simulator achieving a platform allowing aggregate applications simulations over realistic LoRaWAN networks.