One challenge to the retrieval process of Named Data Network (NDN) is to take advantage of largely distributed copies in in-network caches. Potential based routing (PBR) uses a potential field defined in the network to facilitate routing data packet towards a desired location is an effective solution. But due to the characteristics of “Volatilize and Dynamic Update” on caches, there exist problems of high request failure rate and transmission delay in real applications. This paper proposes a novel distributed PBR scheme with prioritized connection probability of the nodes calculated by the popularity and similarity. It mapped the nodes into a hyperbolic plane according to the sequence of the caching time, and utilized radial coordinates and angular coordinates to represent the popularity and similarity, which will be used for the function of Fermi-Dirac distribution to calculate connection probability. The connection probability will be prioritized as the weight of the nodes for construction of potential fields. The simulation results show that our proposed scheme can not only significantly improves cache hit and utilization ratio, but also effectively controls network delay and overhead.
Read full abstract