D6.3 Auto-Discovery space deployment validation report
The presented document is the deliverable “D6.3 – Auto-Discovery space deployment validation report” of the VICINITY project (Grant Agreement No.: 688467), funded by the European Commission’s Directorate-General for Research and Innovation (DG RTD), under its Horizon 2020 Research and Innovation Programme (H2020). This deliverable contributes to reach Milestone 7 (MS7 - First integrated system prototype available) by validation of the Auto-Discovery platform deployment within the task T6.3.
The document presents the process, the results and the quality and performance feedback for the validation of the auto-discovery functionality of the VICINITY platform. The key enabler for achieving smooth integration of heterogeneous IoT systems, platforms, and devices is the interoperability at the semantic level. In the VICINITY platform the semantic interoperability relies on the VICINITY Agents at the client/node side and the Semantic Discovery and the Agent Configuration Platform (SDACP) at the server/cloud side. SDACP in turn consists of a semantic triplestore and the service that provides the API for manipulating IoT object descriptions stored in the triplestore.
During the validation process we proved that:
- Auto-discovery process of the devices described in the VICINITY ontology is fully functional and accessible for the IoT platform adapters connected to the VICINITY Agent.
- Current implementation of the SDACP platform uses a Semantic graph database GraphDB for storing semantic triplestores. The GraphDB is able to manage the appropriate level of load generated by dozens of IoT objects per adapter nd dozens of adapters per one VICINITY Agent.
- The core functionality of the Agent, which is responsible for updating the semantic repository, can perform more than 10.000 DIFF statements per second and more than 15.000 IoT objects per second.
- The Semantic repository that is available via the Semantic discovery, and configuration service, offers an average processing time for one insert operation of 0,09s, for one update operation 0,79s, and one delete operation 0.02s . However, the current development and testing installation of SDACP is based on the free licensed version of the GraphDB store that is limited to two parallel client sessions. This setup can dramatically reduce the scalability, which can be further enhanced by unlimited commercial version of GraphDB.
This deliverable covers all individual modules that have been developed in WP3 - “VICINITY Server Implementation” and WP4 - “VICINITY Client Infrastructures Implementation”, with focus on the server and client side discovery components designed and implemented in the tasks T3.2 and T4.2. Problems identified are timely reported and solved.
The quality and performance feedback of this task’s tests is used in order to improve the auto-discovery functionality – its availability and ability to be also deployed to potential middle and large-scale IoT platforms.