Performance evaluation of caching techniques for video on demand workload in named data network

The rapid growing use of the Internet in the contemporary context is mainly for content distribution. This is derived primarily due to the emergence of Information-Centric Networking (ICN) in the wider domains of academia and industry. Named Data Network (NDN) is one of ICN architectures. In addition, the NDN has been emphasized as the video traffic architecture that ensures smooth communication between the request and receiver of online video. The concise research problem of the current study is the issue of congestion in Video on Demand (VoD) workload caused by frequent storing of signed content object in the local repositories, which leads to buffering problems and data packet loss. The study will assess the NDN cache techniques to select the preferable cache replacement technique suitable for dealing with the congestion issues, and evaluate its performance. To do that, the current study adopts a research process based on the Design Research Methodology (DRM) and VoD approach in order to explain the main activities that produced an increase in the expected findings at the end of the activities or research. Datasets, as well as Internet2 network topology and the statistics of video views were gathered from the PPTV platform. Actually, a total of 221 servers is connected to the network from the same access points as in the real deployment of PPTV. In addition, an NS3 analysis the performance metrics of caching replacement technique (LRU, LFU, and FIFO) for VoD in Named Data Network (NDN) in terms of cache hit ratio, throughput, and server load results in reasonable outcomes that appears to serve as a potential replacement with the current implementation of the Internet2 topology, where nodes are distributed randomly. Based on the results, LFU technique gives the preferable result for congestion from among the presented techniques. Finally, the research finds that the performance metrics of cache hit ratio, throughput, and server load for the LFU that produces the lowest congestion rate which is sufficient. Therefore, the researcher concluded that the efficiency of the different replacement techniques needs to be well investigated in order to provide the insights necessary to implement these techniques in certain context. However, this result enriches the current understanding of replacement techniques in handling different cache sizes. After having addressed the different replacement techniques and examined their performances, the performance characteristics along with their expected performance were also found to stimulate a cache model for providing a relatively fast running time of across a broad range of embedded applications