Offloading Semantic Mashup by On-Path Routers for IoT Applications

摘要

Semantic mashup for IoT applications is one of the major functionalities provided by IoT middleware servers, such as in oneM2M architecture. However, the current centralized approach does have many issues that could jeopardize the network performance and clients' experience. This paper proposes to offload the semantic mashup at en-route routers. The future routers will have more intensive processing and computation capabilities. It is likely that the IP packet could be extended with more blocks to express client's customized requests. The paper proposes one of such extensions to incorporate the semantic mashup logic such that the ultimate target data may be returned by an on-path router. We model the transmission overhead performance and provide the numerical results to compare the existing and the proposed approaches. The proposed scheme shows dramatically improved performances with the feasibility to scale and implement.