ICNRL: An Initiative Framework Towards Information Centric Network Representation

The exponentially growing demand for computational resources prevents the Information Centric Networking (ICN) being deployed in practice due to the high dimensional sparse data computation. However, we argue that Network Representation Learning (NRL) can help to solve the problem by transforming the raw network information data into low-dimensional dense adjacency matrix representation. In this paper, we propose ICNRL, a novel task-based NRL scheme for ICN. Based on the adjacency matrix generated by NRL, ICNRL can calculate the index threshold value to support the networking decision making of content-store (CS), pending information table (PIT), and forwarding information base (FIB), and therefore improves the management and processing capabilities of ICN.