Luxembourg National Research Fund

When the first satellite was launched in the 1950s, earth orbit was a lonely place. Since then, more than 11,000 satellites have been launched into space and over 3,000 are still in operation. Estimates suggest an exponential increase in satellites in the next years, creating a challenge for the effective allocation of the needed bandwidth and power. Researchers are developing algorithms to more effectively allocate the resources where and when they are needed.

Satellite technology provides access to weather information, far-away TV channels and GPS, but also plays an important role in gathering information about climate change, as well as providing emergency medical assistance in remote areas. A range of limited resources – including bandwidth and power – are relied upon to make this happen.

“Unlike traditional satellites, newer satellites are equipped with digital payloads, so they can be flexibly configured to allocate satellite resources according to heterogeneous demands needed by individuals, the maritime, and aeronautical industries.”

“These reconfigurable digital payloads open a door for a substantially higher resource utilisation efficiency and demand satisfaction,” explains computer scientist Tedros Salih Abdu, whose PhD revolves around algorithms for satellites.

More efficient allocation of power and bandwidth in line with individual satellite needs

Research has shown that the way resources are allocated uniformly in satellite technology in practical scenarios for diverse traffic demands is not efficient: Some users may have high traffic demands, some low, meaning a one-size-fits-all approach is not the way forward. The question is, considering the benefits of digital payloads, what is the most efficient way to manage satellite resources to accommodate these different demands requirements?

“Solving the resource allocation problem requires formulating mathematical optimisation, which needs to take into account the utility function of resource efficiency and demand-matching, all considering aspects such as system power and bandwidth, quality of service, as well as user-demand constraints,” Tedros explains.

It is a challenge for researchers, as it is complex to come up with solutions that are both optimal and involve low computational time, due to what Tedros describes as non-linearity functions and the integer variables that are involved in this optimisation problem.

“A key aspect of my research is identifying and developing computation-less resource optimisation algorithms, that can effectively allocate satellite resources according to diverse demands, while at the same time maximizing the utilisation of the overall satellite resource.”

“In addition, my research focuses on designing algorithms that improve the system’s spectral efficiency, while gaining optimal resource utilisation through joint precoding and resource optimisation. In this way, satellite resources are properly managed, and satellite operators can meet the needs of customers,” Tedros explains about his research.

As part of the research, Tedros and the team he is part of have examined algorithms for the so-called ‘GEO’ scenario and are now focussing on constellations in lower orbits. The research is a collaboration wit SES – a Luxembourgish satellite and terrestrial telecommunications network provider.

Tedros Salih Abdu is a 3^rd year AFR PhD candidate at the University of Luxembourg, in the group of Dr Eva Lagunas. Eva herself was featured in Spotlight on Young Researchers in 2018, read her feature here.

RELATED FUNDING