Impact Analysis of Parameters in the Licklider Transmission Protocol

  • Kim, Gahyun; 
  • Song, Chihyun; 
  • Choi, Seongjin; 
  • Cho, Sungrae
Citations

WEB OF SCIENCE

0
Citations

SCOPUS

0

초록

The Licklider Transmission Protocol (LTP) is a core convergence-layer protocol for deep-space Delay/Disruption Tolerant Networks (DTNs), yet its performance is highly dependent on parameter settings such as segment size, session count, and transmission rate. In this study, we analyze the quantitative impact of these parameters on goodput, delivery latency, and protocol overhead using a simulation framework that models LTP's selective retransmission and timeout behaviors. Results show that each parameter has a distinct performance threshold, with optimal settings around 6-8 sessions, 4096-byte MTU, and transmission rates above 512 kbps. The empirical findings align with theoretical models, confirming key trade-offs between efficiency and reliability. Furthermore, we discuss the potential of integrating intelligent optimization meth-ods-such as multi-agent deep reinforcement learning and explainable DRL-to support adaptive and interpretable LTP control in future deep-space communication systems.

키워드

Deep Space; Delay/Disruption Tolerant Networks; DTN; Licklider Transmission Protocol; LTP; Reinforcement Learning
제목
Impact Analysis of Parameters in the Licklider Transmission Protocol
저자
Kim, Gahyun; Song, Chihyun; Choi, Seongjin; Cho, Sungrae
DOI
10.1109/ICOIN68469.2026.11480506
발행일
2026
유형
Proceedings Paper
저널명
International Conference on Information Networking
페이지
82 ~ 86