A significant number of ESA studies require considerable effort on the packet-level Satellite Network Simulator (SNS) development for performance evaluations and optimizations of new network protocols/mechanisms.
Often SNSs are either developed in-house or using commercial libraries which introduces several drawbacks.
Simulators developed in-house are not compatible with each other nor can they be expanded due to their proprietary and inflexible nature. Use of commercial libraries either preclude or necessitate additional costs.
This project aims to develop highly flexible and modular SNS on top of an open-source simulator to be distributed to industrial and academic organizations for use in future ESA activities.
There are two high-level objectives of this project. The primary objective is to develop a packet-level Satellite Network Simulator (SNS) on top of a well-known open-source network simulator. The secondary objective is to devise and implement to SNS RRM algorithms that can exploit and are better suited for a variety of technical capabilities that are introduced in the recent DVB-RCS2 return link specification.
SNS shall exhibit the necessary modularity and flexibility so that it can be used to simulate different interactive transparent GEO satellite networks with minimal source code modifications. At minimum, the simulator shall be able to model a reference system scenario which includes an interactive broadband satellite network with a multi-beam geostationary satellite with a transparent payload and it adopts ETSI DVB-S2 and DVB-RCS2 specifications as the communication standards on the forward and the return links.
RRM algorithms shall be designed in parallel to the early architectural definitions of the simulator, and shall be integrated in the final release of the simulator. It is expected that the flexible and modular architecture of the simulator will reduce the cumulative code development efforts in subsequent ESA activities.
The key issues that are addressed in the project are concerned with defininga suitable open-source platform for packet-level satellite network simulator, refining reference system and models, developing flexible and modular satellite network support for selected open-source simulator and develop, and integrating DVB-RCS2 RRM algorithms to the simulator.
The current packet-level Satellite Network Simulator (SNS) developed for ESA performance evaluations and optimizations of new network protocols/mechanisms are in most cases incompatible with each other and they cannot be expanded due to their proprietary and inflexible nature.
In addition, the use of commercial libraries either precludes or necessitates additional costs. Developing SNS on top of existing open-source simulator – such as NS2, NS3, and OMNeT++ will mitigate described drawbacks and provide variety of benefits due to those simulator platforms being already very popular in academia and industry.
The main strength of such simulators is that they are built on software architectures that are well-understood and tested by many developers from various organizations. This significantly facilitates the joint development of highly modular and flexible network simulations and thus when expanded with packet-level SNS modules it is also expected to significantly reduce simulation software development efforts in future ESA activities.
Simulator architecture will capture necessary functionality to be able to simulate different interactive transparent GEO satellite networks with minimal source code modifications.
At minimum, the simulator shall be able to model a reference system scenario which includes an interactive broadband satellite network with a multi-beam geostationary satellite with a transparent payload and it adopts ETSI DVB-S2 and DVB-RCS2 specifications as the communication standards on the forward and the return links.
This will entail main modules for User Terminal (UT), Satellite, Gateway (GW) + Network Control Center (NCC) and Terrestrial nodes (End users) and interaction between those modules (see attached architecture overview picture, NCC not present in the picture as it is shared module with all GW nodes).
In general, most resource consuming implementation activities are related to DVB-S2/RCS2 MAC and PHY layers (inside Satellite netdevice in architecture overview) and the physical channel models (Satellite channel in architecture overview).
MAC and PHY layer modelling will include incorporation of new RRM algorithms that can exploit and are suited for a variety of technical capabilities that are introduced in the recent DVB-RCS2 return link specification. Other parts, such as the GW, UT, and Satellite nodes, are assumed to require quite minimal changes when compared to existing simulator nodes.
The project is divided into 8 main tasks.
Task 1 will review open-source network simulators and simulator functional requirements.
Task 2 will focus on reference system definitions and traffic models.
Task 3 will define simulator architecture and design definitions followed by a simulator prototype release in Task 4.
Task 5 will focus RRM algorithms design.
Task 6 will contain 1st actual release of the simulator and followed by simulator development final release with RRM algorithms for DVB-RCS2 (Task 7).
Finally, in Task 8 user manual and recommendations for future evolutions are finalized.
Estimated final delivery is T0+19 months.
The project kick-off took place on 1 October 2012, and at the moment Tasks 1-4 have been completed.
The current work focuses on finalizing RRM algorithms in Task 5 and constructing simulator release-1 on top of the prototype release created in Task 4.
Completed Tasks include the following achievements. A detailed review of different open source platforms has been made and reported. After the analysis, NS3 was selected as a platform for ESA SNS. In addition to selection of platform for SNS, reference system, traffic models, simulator architecture and design definitions have been devised. Finally a prototype release of the new SNS has been completed, modelling end-to-end packet transmission via all satellite system nodes.