Wireless loss detection over fairly shared heterogeneous long fat networks

Alan Briones, Adrià Mallorquí, Agustín Zaballos, Ramon Martin de Pozuelo

Research output: Indexed journal article Articlepeer-review

3 Citations (Scopus)


The quality of inter-network communication is often detrimentally affected by the large deployment of heterogeneous networks, including Long Fat Networks, as a result of wireless media introduction. Legacy transport protocols assume an independent wired connection to the network. When a loss occurs, the protocol considers it as a congestion loss, decreasing its throughput in order to reduce the network congestion without evaluating a possible channel failure. Distinct wireless transport protocols and their reference metrics are analyzed in order to design a mechanism that improves the Aggressive and Adaptative Transport Protocol (AATP) performance over Heterogeneous Long Fat Networks (HLFNs). In this paper, we present the Enhanced-AATP, which intro-duces the designed Loss Threshold Decision maker mechanism for the detection of different types of losses in the AATP operation. The degree to which the protocol can maintain throughput levels during channel losses or decrease production while congestion losses occur depends on the evolution of the smooth Jitter Ratio metric value. Moreover, the defined Weighted Fairness index enables the modification of protocol behavior and hence the prioritized fair use of the node’s resources. Different experiments are simulated over a network simulator to demonstrate the operation and performance improvement of the Enhanced-AATP. To conclude, the Enhanced-AATP performance is compared with other modern protocols.

Original languageEnglish
Article number987
JournalElectronics (Switzerland)
Issue number9
Publication statusPublished - 1 May 2021


  • Bottleneck
  • Fairness
  • Heterogeneous long fat networks
  • Loss episode
  • Transport protocol
  • Wireless


Dive into the research topics of 'Wireless loss detection over fairly shared heterogeneous long fat networks'. Together they form a unique fingerprint.

Cite this