Enabling Efficient Data Transport for ICN-based In-Network Computing

Information-centric networking (ICN), originally innovated for efficient content dissemination, has recently been studied to provide end users with in-network computing services. However, existing schemes with the passive communication style suffer from the problem of a larger number of request messages (interests) and consequently, higher latency. This is because of the insufficient interplay between data transport and construction of appropriate chains of the required in-network functions. In this study, we introduced an efficient data transport scheme in cooperation with establishing a path in which the required in-network functions are well chained. The proposed data transport manipulates data retrieval and function calls in the network by exploiting the features of interest message types that provide efficient data transport in ICN-based in-network computing. Using an actual implementation, we demonstrated that the proposed scheme reduces the interest traffic overhead and latency for completing a task compared to a conventional scheme for ICN-based in-network computing.