An Efficient Content Popularity Classification using An Artificial Neural Network with Proactive Caching in Named Data Networking

Named Data Networking (NDN) is a new net-working paradigm aiming to overcome the constraints of IP-based networks. Each node or router in NDN can store data, eliminating the need for users to request it from the server. One of the essential components of NDN is the Content Store (CS), which maximizes caching performance by predicting content popularity using deep learning. This ensures that popular content is stored since numerous users request it. In addition to content popularity prediction, proactive caching is another method that can boost cache efficiency. Proactive caching involves storing content before users request it. This paper will apply the classification of popular content using an Artificial Neural Network (ANN) with proactive caching to the cache placement strategy. Furthermore, the method will be utilized to determine the impact of cache node locations. Based on simulation results applied to the edge-cache placement strategy, the proposed model achieves a 15% increase in cache hit ratio compared to the reactive edge-cache approach. Additionally, the average round trip time is reduced by 8% to 15%, and the cache miss rate is reduced by 15%. By employing the degree method for cache node placement, the cache hit ratio can be increased by 15% to 25%, and the average Round Trip Time (RTT) can be reduced by 15%. Moreover, the Cache Miss value is reduced by 10%. These findings demonstrate that the proposed method effectively enhances the cache hit ratio and reduces cache misses and average RTT.