Friday, 22 September, 2017, I gave a talk at the IEEE 5G Summit held at Kogakuin University in Tokyo. A well organized event at the center of Tokyo attended by over 200 people. Top figures and players in the academia and industry gave their views at the Summit.

My talk was about 5G security reality check where I covered 3GPP standardization status, solutions that are already in specifications and several other aspects necessary to make 5G secure. I will share details on this separately.

All presentations were very practical showing what is expected and realistically possible. A very good overview was given by opening speakers and Japanese government representative where Japan is expecting B2B2X businesses to grow with 5G.

It was repeated over and over that multi-business model is the way to go in 5G, i.e. bringing various industries together. Multi-business model in mobile world is new and will result in expanding the market and potentially making it all inclusive but that also brings several challenges which have to be overcome. Challenges include trust among verticals that have never or not worked as close before, understanding of each others business and it could also lead to cross industry mergers.

Various views were presented regarding radio aspects as well, this is of particular importance as we move towards higher GHz band. As we know, higher frequency would normally equates to shorter range and higher GHz (millimeter (mm) waves) have different radio characteristics since the wavelength becomes similar in size to air molecules. Utilization of carrier aggregation - licensed and unlicensed - as well as dual connectivity were discussed. Pros and cons of beam-forming and array antennas were brought forward. There are challenges but that only makes the job interesting for engineers.

Complexities related to slicing was another issue touched during the summit. Market / service / user needs are leading towards flexible networks. If we go towards flexibility at its best then RAN should also be considered. All this mean that very complex management and orchestration will be required. It will also lead to serious security considerations so that slices are well isolated. Resolving the management, operations and security concerns are key to reaping the business benefits that slicing (or vertical networks) bring.

Another point is Artificial Intelligence (AI) / Machine Learning (ML). There is a big hype that AI can solve all human issues but the fact is that logical explanation of any action by AI, especially when wrong, is also required. AI and its applicability was also extensively discussed. There were different perspectives but one approach presented was to use AI for understanding the situation including learning how to "go beyond" what is known and then applying that to our usual system / algorithms. There is no universal AI method and thus usage based AI will be required. One can apply AI for orchestration as well as for beam-forming.

Last but not least there was discussion regarding the cost of deployment and return of investment. Past generations of mobile systems have shown that cost of deployment has been decreasing while data usage has increased with every new generation. One could expect the same for 5G. Still, to benefit from 5G, different verticals will have to join hand together - mobile communications will thus truly become integral part of human life.