Title: Performance analysis of WMNs by WMN-GA simulation system for two WMN architectures and different TCP congestion-avoidance algorithms and client distributions

Authors: Keita Matsuo; Shinji Sakamoto; Tetsuya Oda; Admir Barolli; Makoto Ikeda; Leonard Barolli

Addresses: Department of Information and Communication Engineering, Fukuoka Institute of Technology (FIT), 3-30-1 Wajiro-Higashi, Higashi-Ku, Fukuoka 811-0295, Japan ' Graduate School of Engineering, Fukuoka Institute of Technology (FIT), 3-30-1 Wajiro-Higashi, Higashi-Ku, Fukuoka 811-0295, Japan ' Department of Information and Computer Engineering, Okayama University of Science (OUS), 1-1 Ridaicho, Kita-Ku, Okayama 700-0005, Japan ' Department of Information Technology, Aleksander Moisiu University of Durres, L.1, Rruga e Currilave, Durres, Albania ' Department of Information and Communication Engineering, Fukuoka Institute of Technology (FIT), 3-30-1 Wajiro-Higashi, Higashi-Ku, Fukuoka 811-0295, Japan ' Department of Information and Communication Engineering, Fukuoka Institute of Technology (FIT), 3-30-1 Wajiro-Higashi, Higashi-Ku, Fukuoka 811-0295, Japan

Abstract: In this paper, we evaluate the performance of two wireless mesh networks (WMNs) architectures considering packet delivery ratio (PDR), throughput, delay and fairness index metrics. For simulations, we used ns-3, distributed coordination function (DCF) and optimised link state routing (OLSR). We compare the performance of transmission control protocol (TCP) Tahoe, Reno and NewReno for normal and uniform distributions of mesh clients by sending multiple constant bit rate (CBR) flows in the network. The simulation results show that the PDR for both distributions and architectures is almost the same, but the PDR of I/B WMN for uniform distribution is a little bit higher than normal distribution. For both WMN architectures, the throughput of normal distribution is better than uniform distribution. The delay of Tahoe is a little bit lower compared with other algorithms for normal distribution and Hybrid WMN, but in case of uniform distribution, the NewReno performs better than other algorithms for both architectures. The fairness index of normal distribution is higher than uniform distribution.

Keywords: wireless mesh networks; genetic algorithm; node placement; WMN architectures; TCP congestion-avoidance algorithms.

DOI: 10.1504/IJCNDS.2018.091068

International Journal of Communication Networks and Distributed Systems, 2018 Vol.20 No.3, pp.335 - 351

Received: 14 Sep 2017
Accepted: 18 Sep 2017
Published online: 09 Apr 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article