A DETERMINISTIC LATENCY NAME RESOLUTION FRAMEWORK USING NETWORK PARTITIONING FOR 5G-ICN INTEGRATION

摘要

Achieving low latency is one of the goals of 5G networks and beyond. We consider Information-Centric Networking (ICN) as a candidate network architecture to realize 5G objectives. The trend of the future network is transformed from best-effort to deterministic transmission. In this paper, we introduce the concept of deterministic latency into the Name Resolution System (NRS) of ICN and propose a Deterministic Latency Name Resolution (DLNR) framework to provide name resolution service with deterministic low latency to satisfy the delay sensitive requirement of NRS. As one of the key techniques, firstly, a Latency-aware Hierarchical elastic area Partitioning (LHP) algorithm is designed to solve the problem of resolver placement with bounded and deterministic transmission latency conditions. Secondly, we design corresponding latency demand-aware name registration and resolution schemes so that constant forwarding hops can be realized. Simulation results show that deterministic low latency name resolution service is achieved by Enhanced NRS (ENRS), which is a system developed to realize the framework of DLNR. Furthermore, the average response latency of ENRS outperforms the Resolution Handler (RH) of DONA 25.2% at best. The query overhead of ENRS is stable due to the deterministic request-forwarding, and the single query overhead of ENRS is reduced by up to 30 times compared with KNN-DNRS.