Publications 2021

Conference articles (4)

Jani Puttonen, Lauri Sormunen, Henrik Martikainen, Sami Rantanen, Janne Kurjenniemi, “A System Simulator for 5G Non-Terrestrial Network Evaluations“, IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM), Non-Terrestrial Networks in 6G Wireless workshop, June 7-11, 2021.

3rd Generation Partnership Project (3GPP) is working on the specifications related to the 5G satellite component, i.e., 5G Non-Terrestrial Networks (NTN). 5G NTN specifications can be used to build different spaceborne or airborne 5G systems.

This article presents a 5G NTN extension to an open-source Network Simulator 3 (ns-3) and its 5G extension (5G LENA). The objective is to use the resulting System Level Simulator (SLS) in 3GPP standardization to evaluate different system concepts and parameterizations.

M. Majamaa, J. Kurjenniemi, L. Sormunen, J. Puttonen, “Usage of GPGPU computing in LEO satellite network simulations”, Ka and Broadband Communications Conference, Arlington, USA, September 27-30, 2021.

Low Earth Orbit (LEO) satellite constellations have been under intense research and development activities in recent years. These satellite networks aim to tackle the latency problems associated with the Geostationary Earth Orbit (GEO) satellites while still providing wide area coverage in case of large constellations.

Satellite movement, non-uniform, and evolving user demand as well as coordination with other systems and regulatory constraints, sets new problems for large LEO constellations. Satellite payload flexibility and reconfigurability will be key features of on-going and future constellation projects.

Development of these new systems also set new requirements for simulation tools used to develop Radio Resource Management (RRM) algorithms and there will be a need to run large scale simulations with a high number of satellites and end users. These kinds of simulations will be complex and time- consuming, thus new solutions are needed to speed-up the simulations. With the aid of a graphics processing unit (GPU) in general-purpose computing one can execute programs exploiting GPU’s massive parallelism to reduce simulation time.

The objective of this paper is to study the usage of General-purpose computing on graphics processing units (GPGPU) and code parallelization in central processing unit (CPU) in LEO satellite network simulations and evaluate the gains in simulator run-time performance.

R. Järvinen, J. Alhava, S. Sourulahti, S. Haka, J. Puttonen, J. Kurjenniemi, G. Acar, “A System Simulator for Broadband Global Area Network (BGAN)“, Ka and Broadband Communications Conference, Arlington, USA, September 27-30, 2021.

Broadband Global Area Network (BGAN) is a mobile satellite communications service by Inmarsat [1], which is providing reliable, cost-effective global broadband data and voice using compact, lightweight portable terminals. BGAN provides almost global coverage with three geostationary satellites for maritime, enterprise, aviation, and governmental customers.

BGAN uses a proprietary Inmarsat Air Interface-2 (IAI-2) protocol stack, which is also the basis of the technical specification of the SL (Satellite Link) family of the Satellite Universal Mobile Telecommunications System (S-UMTS) standard, which was proposed by Inmarsat and published by ETSI as TS 102 744 [2-5]. Unlike IAI-2, S-UMTS Family SL is an open standard.

In this paper, a system level simulator that models the S-UMTS Bearer connection and Bearer control layers is presented. In particular, the focus is on the return link scheduling utilizing adaptive modulation and coding scheme (MCS) to balance reliability and efficient bandwidth usage. A custom algorithm was designed for the return link radio resource management (RRM) of the S-UMTS simulator.

A. Masri, J. Kurjenniemi, J.Puttonen, T. Varjus, M. Majamaa, “Capacity Level Simulator for NGSO Satellite Constellations“, Ka and Broadband Communications Conference, Arlington, USA, September 27-30, 2021.

Mega sized next-generation Low Earth Orbit (LEO) satellite constellations are being deployed by Starlink by SpaceX, Project Kuiper by Amazon, OneWeb, Lightspeed by Telesat, among others. Those state-of-the-art LEO constellations will be orbiting the Earth at different altitudes, speeds, and inclinations. Moreover, some of them will target providing a 5G like services known as the 5G Non-Terrestrial Network (5G NTN) [1], while others will be using the proprietary air interfaces in the satellite link, and thus providing only e.g., backhaul connectivity to terrestrial 5G connections.

In both cases, the targeted system of high satellite velocity with doppler effects, propagation delay, Inter-Satellite Links (ISLs), power constraints at both sides of the link, mobility management, global coverage, high data rates, not to mention all, will impose higher degrees of complexity at different levels of the system stack compared to current Geostationary (GSO) systems [2] [3]. On top of that, future satellite payloads are capable of dynamic determination of the beam (coverage), frequency, time (beam hopping) and power.

Consequently, one critical aspect that needs special attention is Radio Resource Management (RRM) (resource allocation). In which, RRM algorithm shall meet the non-uniform and time-evolving user demand by means of payload flexibility, reconfigurability, and other systems and regulatory constraints. Consequently, it is important to have a dedicated RRM tool that could be utilized to assess the evaluation and the comparison between different RRM algorithms at the capacity level for mega-constellations.

In this paper, we design a capacity-level modular simulator for evaluating and comparing different RRM algorithms as part of the Constellation – Dynamic Resource Allocation Management (C-DReAM) project under the fund from the European Space Agency (ESA).