More experiments with Delay and Disruption Tolerant Networking over AX.25 Networks

John Ronan, Darren Long, Kristian Walsh

Research output: Contribution to conferencePaperpeer-review

Abstract

Previously the authors gave an overview of a Delay Tolerant Network Convergence Layer implemen- tation that operates over Connected Mode AX.25, and detailed the results of some performance tests using it to transfer data over Amateur Radio channels. These results were compared to both a native AX.25 and a TCP/IP- over-AX.25 implementation. The investigation of TCP/IP was undertaken because, while it is generally accepted that TCP is unsuited to wireless links, it has become the dominant protocol in real-world applications, with over 50% of internet traffic now accounted for by TCP over port 80 [1]. As some issues were highlighted in experiments leading to the authors’ prior publication, these have been worked on and have been largely resolved. It was also found that our model for an ideal AX.25 communications channel had deficiencies, so a correction is offered. Additionally, we also tested our Convergence Layer alongside a TCP/IP over AX.25 implementation on both a 1200 and 9600 baud point-to-point link and give comparative results between our Convergence Layer implementation and TCP/IP. Real-world behaviour of the data link still diverges from the model, but the authors provide some possible reasons for this.
Original languageEnglish
Publication statusPublished - 2011
EventTAPR and ARRL 30th Digital Communications Conference 2011 - Baltimore, Maryland, USA, Baltimore, Maryland, USA
Duration: 01 Jan 2011 → …
http://www.tapr.org/pub_dcc30.html

Conference

ConferenceTAPR and ARRL 30th Digital Communications Conference 2011
CityBaltimore, Maryland, USA
Period01/01/2011 → …
Internet address

Keywords

  • Disruption tolerant networking
  • Internet-working
  • Packet radio networks
  • TCPIP
  • Transport Protocols

Fingerprint

Dive into the research topics of 'More experiments with Delay and Disruption Tolerant Networking over AX.25 Networks'. Together they form a unique fingerprint.

Cite this