Panel: “Challenges for further trialling and beyond 5G research” at the IEEE 5G World Forum

Anastasius Gavras from Eurescom GmbH and project 5G-VINNI moderated a panel discussion among Prof. Riku Jäntti – Aalto University, Prof. Sławomir Kukliński – Orange, Dr. Ville Niemelä – University of Oulu and Dr. Maria Guta – ESA. The topic of the panel was “Challenges for further trialling and beyond 5G research”. The panel was attended by about 40 participants that actively contributed to the discussion.

The panellists agreed that a number of challenges are still ahead of us in the context of 5G trialling. For beyond 5G research the challenges are still to be identified, but considering terahertz carrier frequencies, we may need to rethink what testbeds need to address. For example materials play an important role in this frequency range.

Concerning current 5G trials we must define well the purpose of the trials (conformance, interoperability, new features…) in order to identify the right environment. Even if more than 150 5G-related trials are on-going in Europe we still do not have a commercial implementation of MANO or end-to-end slicing. These are all things that need to be tested and trialled. In the satellite/5G convergence area more trials are needed and starting small/humble and increasing complexity might be a possible way to control costs.

Concerning the evolution of testbeds we should exploit the programmability at all layers. Most parts of the testbeds are software and easily upgradeable. Software-based techniques can also support better openness and accessibility to the testbeds.

Other future work includes the evolution of central parts such as MANO and OpenRAN/antennas as well as full convergence down to the radio level for terrestrial and non-terrestrial networks. We should be more carefully designing the testbeds (block design in a sandbox) and recall techniques like formal specifications. We should make an effort to increase awareness that good documentation is essential for open source. Finally open source in the radio part is promising but very few people actually contribute. How can we change that?

Overall there was agreement that testbeds and trialling are an essential part of the innovation cycle and that in the foreseeable future we cannot progress without testbeds. Cost is an important element that will always provide the boundary of what testbeds capable of.