Throughput measurement and estimation model for concurrent communications of multiple raspberry pi access-points in wireless local-area network

Mousumi Saha, Rahardhita Widyatra Sudibyo, Nobuo Funabiki, Minoru Kuribayashi, Sumon Kumar Debnath, Wen Chung Kao

Research output: Contribution to journalArticle

Abstract

Recently, Raspberry Pi has become popular around the world as an inexpensive, small, and powerful computing device. We have studied its use as the software access-point (AP) for the IEEE 802.11n wireless local-area network (WLAN), where the channel bonding (CB) can increase the throughput by bonding two channels into one. However, the built-in wireless network interface card (NIC) adapter fails to support the CB. Thus, we have configured the CB using external NIC adapters, and investigated the throughput performance and estimation model for only one link between a PC host and a Raspberry Pi AP. In this paper, we present the throughput measurement for concurrent communications of multiple links with up to three APs under different adapters and channels with 11 and 13 partially overlapping channels (POCs) at 2.4GHz. Besides, we extend the throughput estimation model by introducing the throughput drop factor to consider interferences between the links. This factor can be derived from the CB use and the assigned channels for the links. The effectiveness of the proposal is verified by comparing the estimated throughput with the measured one.

Original languageEnglish
Pages (from-to)1009-1016
Number of pages8
JournalJournal of Communications
Volume14
Issue number11
DOIs
Publication statusPublished - Jan 1 2019

Keywords

  • Access point
  • Channel bonding
  • Concurrent communication
  • IEEE 802.11n
  • Raspberry Pi
  • Throughput estimation model

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Throughput measurement and estimation model for concurrent communications of multiple raspberry pi access-points in wireless local-area network'. Together they form a unique fingerprint.

  • Cite this