Analyzing information propagation in a transoceanic aircraft delay tolerant network

Abstract

In this paper we analyze information propagation inside a large scale delay tolerant network (DTN) of transoceanic aircraft. We characterized this network at LeN 2013, a scenario constructed from more than 2,500 traces of trans-Atlantic flights in which communications rely on the sporadic contacts between airplanes. We observe the traffic of this network and analyze how two epidemic based routing protocols behave in this scenario. We use simulations to compare their performance through metrics such as the network traffic per packet, end-to-end-delay and delivery ratio. The analysis of data propagation inside this DTN aims to evaluate the feasibility of using this network as a reliable way to deliver on-flight generated data (either from passengers or company) to the ground. This approach seeks to provide applications with a cheaper alternative to other communication systems. I. INT RODUCTIONData communication in aeronautical scenarios is very lim ited [12]. Airplanes periodically send to ground stations their identification, position, and company related data Gust a few tens of bytes per second). Any other control information is sent over traditional voice-based radio links. These limitations in data communication become more evident when looking for the causes of aerial accidents, as in the case of the ill-fated MH370's disappearance [16].Regarding passenger data communication, some companies have deployed, for domestic flights, solutions based on ex ternal antennae directed to ground [2]. As to transoceanic flights, very few companies choose to use expensive satellite links [6] while, in most of the cases, passengers must remain disconnected for hours. On the other hand, most of the research papers on aeronautic transoceanic communications propose the use of communication systems with large coverage (from 300 to 600 km) and thus consider airplanes connected to form an Aeronautical Ad-hoc Network (AANET) [13], [15], [12].Looking for a cheaper solution than those based on radio links with high coverage or satellite links, we proposed in LCN 2013 [14] a new approach based on aircraft equipped with IEEES02.11 bIg conununication links of only 20 km. This limited range leaded to a network with a reduced node degree, fully in range with the characteristics of Delay Tolerant Networks (DTNs) [7], [3]. To build our proposal, instead of using synthetic data, we departed from traces of 2,550 transAtlantic flights to reach a characterization of a large-scale real-world aeronautical DTN. We analyzed the distributions of active nodes, node degree, number of clusters, any contact time (ACT), and inter-any-contact time (IACT) of that network.The objective of this paper is to analyze how data propagates inside our proposed transoceanic aircraft delay tolerant net work, in order to evaluate the feasibility of using this network as a reliable way to deliver on-flight generated data (either from passengers or company) to the ground.Our contributions are as follows:• We provide an evaluation of two different routing proto cols for t...

Methodological evaluation of architectural alternatives for an aeronautical delay tolerant network

“…In this section we recall our scenario from [8,11] and proceed to describe the methodology used to define the network model and its characteristics. This model has been implemented using the ns-3 simulator [28] and currently represents the different features of the network in great detail, from link layer to application layer, including a custom mobility model and the routing protocols.…”

“…In LCN 2014 [11] we analyzed how data propagate inside our previous transoceanic aircraft delay tolerant network, in order to evaluate the feasibility of using this network as a reliable way to deliver on-flight generated data (either from passengers or company) to the ground. First, we provided an evaluation of two different routing protocols for this scenario: A basic routing protocol using an epidemic dissemination approach, and a version of the protocol which uses a decision function based on an Expected Landing Time (ELT) metric to reduce the data overhead.…”

“…Our final conclusion in [11] was that the architecture of our network had to be improved in order to increase the percentage of nodes capable of performing successful air-to-ground communications. To this end, we planned as a future work to devise a more suitable network architecture by combining nodes with just opportunistic communications with nodes equipped with default aircraft satellite communications.…”

“…The main objective of this paper, as an extension of [11], is to evaluate the different ways to create this new architecture in order to find the combination that improves the network performance while minimizing its deployment costs. We conduct this evaluation through the analysis of the graph representation of the network.…”

“…Section 2 presents the background and previous work on Aeronautical Adhoc Networks, on performance evaluation of DTN over realistic scenarios, and on the use of graph analysis on such networks. In Sections 3 and 4 we provide a comprehensive description of our previous papers [8,11]. In Section 5 we present the new architecture, the two types of nodes which compose it, and the characteristics of the new scenario.…”