Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal SMARTICS Journal
Munaldi
Universitas Ahmad Dahlan
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Evaluasi Kinerja Software-Defined Networking Pada Topologi Fat-Tree Untuk Efisiensi Bandwidth Munaldi; Tole Sutikno; Rusydi Umar
SMARTICS Journal Vol 12 No 2 (2026): SMARTICS Journal (Oktober 2026)
Publisher : Universitas PGRI Kanjuruhan Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21067/smartics.v12i2.14594

Abstract

The increasing demand for digital services requires network architectures that are flexible, programmable, and efficient in utilizing available bandwidth. This study evaluates the performance of Software-Defined Networking (SDN) and a conventional network on an identical Fat-tree k = 4 topology. A quantitative experiment was conducted in Mininet using Open vSwitch, OpenDaylight, OpenFlow 1.3, iPerf3, and ping. Traffic loads were configured at 25%, 50%, and 75% of a 100 Mbps bottleneck-link capacity, with ten repetitions per scenario. Performance was assessed using throughput, round-trip latency, packet loss, and bandwidth efficiency. SDN achieved an average throughput of 81.8 Mbps compared with 60.8 Mbps for the conventional network, reduced average latency from 21.8 ms to 17.1 ms, and reduced packet loss from 2.3% to 0.8%. Bandwidth efficiency increased from 60.8% to 81.8%, equivalent to a relative improvement of 34.5%. A paired-sample t-test across the three load-level means showed a significant throughput difference (p = 0.018), whereas latency (p = 0.088) and packet loss (p = 0.105) were not significant at α = 0.05. The findings indicate that centralized flow control and multipath utilization improve performance under increasing traffic loads, although validation on physical testbeds and larger paired samples remains necessary.
Co-Authors Rusydi Umar Tole Sutikno