Affiliations: [a] Department of Computer Applications, Anna University Regional Campus Madurai, Madurai, India
| [b] School of Computing, Kalasalingam Academy of Research and Education, Krishnankoil, India
| [c] Institute of Informatics and Computing Energy, Universiti Tenaga Nasional, Malaysia
| [d] Department of Information Systems, STMIK Pringsewu, Lampung, Indonesia
Correspondence:
[*]
Corresponding author. Shankar Kathiresan, School of Computing, Kalasalingam Academy of Research and Education, Krishnankoil, India. E-mail: [email protected].
Abstract: Named Data Networking (NDN) is a recent Internet model, takes the advantages and resolves the drawbacks of traditional TCP/IP architecture to satisfy the increasing demands on communication. Every NDN router has a pending interest table (PI) to store all the interest packets (I-pkts) waiting for the arrival of data packets (D-pkts). In case of the arrival of enormous I-pkts, it might be hard for an NDN router to store all I-pkts in the limited PIT space. So, a novel PIT management scheme becomes essential for effective PIT utilization. This paper presents a new PIT sharing algorithm based on the hybridization of bat optimization (BO) algorithm and fuzzy logic (BFPIT), to accommodate number of I-pkts in a sharing NDN (SN) node. The BFPIT algorithm intends to identify the optimal SN node sharing node to store I-pkts of a requestor NDN (RN) node. The proposed BFPIT algorithm operates in two stages: BO algorithm for preliminary SN (PSN) node selection and fuzzy logic for final SN (FSN) node selection. For validation, the proposed method is implemented, and the simulation results are compared with the k-nearest neighbor (k-NN) algorithm. The experimental results revealed that the BFPIT algorithm significantly increases the CHR and minimizes the average content delivery time.
Keywords: Network, NDN, fuzzy logic, PIT, bat optimization
