Traffic Routing in a Satellite Backbone Network for Global Air Traffic Management

Motivation and Background

The Perspective: Advanced satellite-based aeronautical passenger communications (APC), as soon to be provided from WirelessCabin, inherently bears the potential to boost some improvements in air traffic control (ATC) commu-nications, just because of the sheer appeal of an available infrastructure, equipment, and capacity. The take-up of usage for ATC purposes will start from less safety-critical applications such as crew communications, general airline operational control (AOC), and bi-directional data transfer of supporting air traffic management (ATM) information between cockpit and control centers. While such evolution is just going to happen, effectively shaping the future in this area in some way, it is worthwhile to also formulate a vision and strategic goal, including an overall long-term networking concept and related ideal target system to provide a global satellite ATM component in a dedicated and optimized way.

Combining Satellite Communications and Navigation: In the long term, there can be significant benefit from combin-ing advanced satellite communication concepts and networks with a satellite constellation designed for global navi-gation, such as Galileo, in a truly integrated overall system design. Such constellations inherently provide excellent multiple visibility (i.e, "hot" redundancy) and can therefore meet highest availability requirements. This may even contribute to acceptance of such a system solution to provide safety-of-life critical services in the future.

Traffic Routing is one key challenge in such an architecture. The characteristics of the heterogeneous topology will directly influence the performances of routing methods under consideration and therefore, will determine the choice of the candidate routing algorithm(s).

The envisaged topological nodes comprised in the system are:

• fixed terrestrial nodes (ATCs, airport towers,…)
• fixed space segment nodes (geostrationary satellites (GEOs))
• dynamic and deterministic space segment nodes (medium/low earth orbiting (MEO/LEO) satellites)
• dynamic and (almost) predictable air segment nodes (air vehicles)

Task Description

As one distinctive backbone component of a global routing architecture, a meshed satellite sub-network based on MEO satellites will be defined and assessed in its performance; while the connections between moving satellites are established by means of intersatellite links (ISLs) which are variable in pointing and distance, a sophisticated MPLS-based overlay network may ensure completely reliable and predictable routing performance over this dynamic space-based backbone topology. This approach is also well suited to support multicast routing approaches which appear to become of major importance in future global ATM networking.

The diploma thesis work aims at

1. setting up a concept and related modeling of a global multiservice ATM/ATC satellite network,
2. devising a reference traffic routing approach and algorithms to operate in the satellite ISL backbone, and
3. concept, design and implementation of system simulations. In this task, focus will be on an overall clean and modular approach and its top-down verification, including some test simulations, rather than extensive programming of less important detail aspects.

The concrete steps to be performed include the following (not preventing adaptations and flexible reaction to lessons learnt while performing the work):

• Study the relevant general literature (ATM/ATC basics, satellite systems/constellations, intersatellite links, traffic modeling, routing basics, MPLS).
• Study in detail the concepts, models and code available as input and/or framework for this thesis.
• Galileo-like MEO constellation and design of a suitable ISL network (C simulator)
• various user traffic models related to passenger and ATM/ATC communications (Matlab etc.)
• global commercial flights database (ASCII->Matlab)
• mapping tools of aircrafts to satellite spot beams (Matlab)
• reference ISL backbone routing concept and algorithms (C simulator; ns2 simulator)
• The relevant parts of the C simulator need to be ported from Sun/Solaris to Linux before ; this should how-ever be unproblematic since all C code is strictly ANSI.
• Develop an own structured and comprehensive problem formulation for the thesis.
• Setting up an overall ns2-based discrete-event system simulator, with all relevant C code integrated, and in-terface specification and implementation with respect to the Matlab modules.
• Numerical studies, based on simulations, aiming at (i) a reliable estimation of both peak/average values and dynamics of traffic load on various ISLs and (ii) quantitative estimation of the impact of typical multicast traffic patterns.

The work will be done in close cooperation and mainly at the premises of TriaGnoSys GmbH, contact person Dr. Markus Werner.