Network Function Virtualization enabled Cache Placement

Abstract

The Internet of Things (IoT) has ushered in a new era of connectivity in which all facets of existence are technologically interconnected. However, this also presents new network management challenges, including resource utilization, security and management. Big concerns like security, computational capacity, bandwidth, latency and energy consumption are a part of enterprise network which have become pervasive issues. The cache placement problem entails determining the most efficient arrangement of cache resources in a network to minimize the cache miss rate and maximize the cache hit rate. This is especially important in IoT networks, where the amount of data generated by connected devices can be enormous, and where the ability to access this data rapidly is essential. Using Software-Defined Networking (SDN) and Network Function Virtualization (NFV) is one strategy for mitigating these obstacles. SDN enables dynamic control of the network's data traffic, directing it to the appropriate resources, including cache resources. NFV, on the other hand, enables the virtualization of network functions, including cache functions, which can be altered dynamically to meet the network's changing requirements. By combining SDN and NFV, it is possible to construct an IoT network that is more efficient, more secure, and more flexible, and thus better able to handle the IoT's challenges. Our work entails the development of an intelligent SDN-NFV-based network topology that provides efficient cache allocation for small to medium enterprise networks. Our work describes the implementation of an SDN-NFV-based cache placement solution employing the Open Network Operating System (ONOS) SDN controller, OpenFlow switches i.e., Zodiac FX and Squid Proxy Server. The VNFs consist of cache servers that can be dynamically modified to satisfy the network's fluctuating requirements.