Licklider transmission protocol (LTP)-based DTN for cislunar communications

  • Authors:
  • Ruhai Wang;Scott C. Burleigh;Paavan Parikh;Che-Jen Lin;Bo Sun

  • Affiliations:
  • School of Electronics and Information Engineering, Soochow University, Suzhou, China and Phillip M. Drayer Department of Electrical Engineering, Lamar University, Beaumont, TX;Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA;TeraRecon, Inc., San Mateo, CA and Phillip M. Drayer Department of Electrical Engineering, Lamar University, Beaumont, TX;Department of Civil Engineering, Lamar University, Beaumont, TX;Department of Computer Science, Lamar University, Beaumont, TX

  • Venue:
  • IEEE/ACM Transactions on Networking (TON)
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Delay/disruption-tolerant networking (DTN) technology offers a new solution to highly stressed communications in space environments, especially those with long link delay and frequent link disruptions in deep-space missions. To date, little work has been done in evaluating the performance of the available "convergence layer" protocols of DTN, especially the Licklider Transmission Protocol (LTP), when they are applied to an interplanetary Internet (IPN). In this paper, we present an experimental evaluation of the Bundle Protocol (BP) running over various "convergence layer" protocols in a simulated cislunar communications environment characterized by varying degrees of signal propagation delay and data loss. We focus on the LTP convergence layer (LTPCL) adapter running on top of UDP/IP (i.e., BP/LTPCL/UDP/IP). The performance of BP/LTPCL/UDP/IP in realistic file transfers over a PC-based network test bed is compared to that of two other DTN protocol stack options, BP/TCPCL/TCP/IP and BP/UDPCL/UDP/IP. A statistical method of t-test is also used for analysis of the experimental results. The experiment results showthat LTPCL has a significant performance advantage over Transmission Control Protocol convergence layer (TCPCL) for link delays longer than 4000 ms regardless of the bit error rate (BER). For a very lossy channel with a BER of around 10-5, LTPCL has a significant goodput advantage over TCPCL at all the link delay levels studied, with an advantage of around 3000 B/s for delays longer than 1500 ms. LTPCL has a consistently significant goodput advantage over UDPCL, around 2500-3000 B/s, at all levels of link delays and BERs.