ANFIS based Classification Model for Network Device Migration towards SoDIP6 Networks
Babu Ram Dawadi, Danda B. Rawat, Shashidhar R. Joshi, Pietro Manzoni · 2021
Financially sustainable solution is the need to migrate legacy IPv4 network infrastructure into latest networking paradigms viz. Software-defined networking (SDN) and the Internet protocol version 6 (IPv6) addressing. For Internet service provider (ISP) networks, IP routers are either to be upgraded or replaced with new device to transform the network into Software-defined IPv6 (SoDIP6) networks. In this paper, we classify the network routers require its hardware/firmware either to be upgraded or replaced using adaptive neuro fuzzy inference system (ANFIS), a well-known intelligent approach used in classification, prediction, and estimation. Device specification details with end-of-life and end-of-support information are stored in the knowledge base (KB) system. Device performance parameters, for example average CPU usage, throughput, and memory capacity are being extracted in real time using simple network management protocol (SNMP) and mapped with information obtained from the KB for input to ANFIS. The experimental results show that proposed approach is more accurate and optimal that assists service providers for smooth transitioning to SoDIP6 networks with optimum cost of migration.