Intelligent approach to switch replacement planning for Internet service provider networks

Abstract

Abstract Network equipment management has a significant challenge in terms of operating costs and human resource management for Internet service providers (ISPs). Effective maintenance leads to a significant reduction in operating costs. It is necessary to identify the status of the networking device for effective maintenance. In this paper, we implement adaptive neuro-fuzzy inference system (ANFIS) to classify the switch for its upgrade, continuation, or replacement. ANFIS is an intelligent approach generally used in classification, estimation, prediction and forecasting. We develop a model to identify the switch status to take decision for its upgrade, replacement or continuation of the operation based on the device parameters obtained through simple network management protocol (SNMP) and knowledge-based system. The model is evaluated in terms of root mean square error (RMSE) and mean absolute error (MAE). ISPs can plan for the network device upgrade, continuation, or replacement planning based on the available budget constraints after the device status is identified. This helps to reduce the operational and capital expenditure of service providers for their sustainability.