An artificial neural network-based distributed information-centric network service

Artificial neural networks (ANN) have been widely used in various areas. As a bottleneck, hardware specification affects the efficiency of an ANN. With the development of distributed computing, distributed ANNs show advantages in dealing with huge data. The network bandwidth is a new bottleneck restricting the performance of distributed ANNs. Information-Centric Networking (ICN) [1], as the Next Generation Network (NGN) solution, has shown merits regarding mobility, security, power consumption and network traffic. In this paper, we remodel the architecture of network service using ANNs. We proposed an ANN-Based Distributed Information-Centric Network Service (ANN based DICNS). The distributed nodes are connected like a neural network. When a client utilizes the DICNS, the data flow from the source to the consumer node like the signal traveling from an input layer to an output layer in a neural network. By using an ICN, our proposal can significantly reduce network consumption, and the named data can help the DICNS effectively manage and classify the data.