Towards Cooperative Semantic Computing: A Distributed Reasoning Approach for Fog-Enabled SWoT

Abstract

The development of the Semantic Web of Things (SWoT) is challenged by the nature of IoT architectures where constrained devices are connected to powerful cloud servers in charge of processing remotely collected data. Such an architectural pattern introduces multiple bottlenecks constituting a hurdle for scalability, and degrades the QoS parameters such as response time. This hinders the development of a number of critical and time-sensitive applications. As an alternative to this Cloudcentric architecture, Fog-enabled architectures can be considered to take advantage of the myriad of devices that can be used for partially processing data circulating between the local sensors and the remote Cloud servers. The approach developed in this paper is a contribution in this direction: it aims to enable rule-based processing to be deployed closer to data sources, in order to foster the implementation of semantic-enabled applications. For this purpose, we define a dynamic deployment technique for rule-based semantic reasoning on Fog nodes. This technique has been evaluated according to a strategy improving information delivery delay to applications. The implementation in Java based on SHACL rules has been executed on a platform containing a server, a laptop and a Raspberry Pi, and is evaluated on a smart building use case where both distribution and scalability have been considered.

Quality of service‐aware approaches in fog computing

“…By semantic support, along with available contextual information, QoS solutions can be enhanced. Applying semantic computing and context‐aware computing for developing QoS‐aware solutions in fog system might be an interesting direction for future studies. Applying properties and information on Online Social Networks (OSNs) for QoS provisioning in FC is an exciting direction for future works. OSN can be used in the QoS field to predict service and resource requests in fog systems .…”

EDR: A generic approach for the distribution of rule-based reasoning in a Cloud–Fog continuum

“…The proposed approach depends on the observation representation, hard-coded in the current implementation, which is why standard vocabularies have been used as much as possi-ble. In future versions, EDR will be updated to be compliant with the updated version of SSN, SSN/-SOSA 11 [? ].…”

Distributing Reasoning on WoT Edge Architectures

“…Typical Internet of Things (IoT) infrastructures rely on sensors and actuators to measure and influence physical phenomena, but also on cloud servers to collect and process data [1]. Such centralized infrastructures can cause latency and bandwidth consumption, as well as bottlenecks and single points of failure [2].…”