Impact of the Content Store Scaling toward the LRU and FIFO Cache Replacements on NDN using Mini-NDN
Mochamad Soebagja Budiana, Muhammad Miqdad Nadra, Transmissia Ratu Hapsari, Ratna Mayasari, Nana Rachmana Syambas · 2021
The requisite access speed on a network is the main thing in the switchover of information. The IP-based network uses a client-server approach that provides a sufficient high delay because of the requested information placed on the server, which is considered less effective. One of the solutions to overcome this issue is to use the NDN architecture. The NDN architecture can responses the client request through the closest node without requesting the information from the server because the NDN node can save the data in the Content Store (CS). Another factor that can affect the service quality of the NDN network is the cache replacement strategy. Some of the common strategies used are Last Recently Used (LRU) and First In First Out (FIFO). We conduct a study on the effect of changing the CS size toward the LRU and FIFO cache replacements on the NDN network by using a Mini-NDN. This study will evaluate the impact of changing the CS sizes toward the cache replacements used. In this work, each customer periodically sends the interest packets to the producer with the same amount using a traffic generator. The results of the data processing obtained are increasing the CS size, the average roundtrip time will decrease. And also, the lower the CS size, LRU is more effective than FIFO.