2. Service discovery in IoT.est
• It is the most important component for the
knowledge management module in the IoT.est
project.
• It provides a number of functionalities
essential for other work packages in service
creation, composition, testing and
deployment.
– semantic annotation for IoT services
– service description data distribution/storage
– IoT service lookup and discovery 2
3. Proposed solution
• Our solution: geospatial indexing + distributed
semantic repository;
• Performing semantic search on semantic repositories
ensures the accuracy: support more meaningful and
expressive queries;
• Geospatial indexing (e.g., R-Tree) ensures the search
efficiency: efficiently reduce the search scope during
discovery.
3
4. Update operations in the discovery
platform
• updating R-Tree is an expensive operation;
• however, since the leaves in the R-Tree are semantic
repositories (e.g., Gateways), even the services
change frequently, the R-Tree may not need to be
updated.
• most changes are constrained within the managing
gateways:
– Update messages do not need to propagate
– significantly reduces computational complexity of the
update operations in dynamic environments.
4