5G channel models
Current radio channel models will not work well in 5G. Why and what is being done?
We continuously strive to optimize mobile communications networks, in order to get the maximum performance out of them. Accurate radio channel knowledge is crucial to do this well. One example: antenna transmission with narrow beam-forming is a technique to improve network performance. When optimizing that, it is very important that the modelling of the radio channel directional characteristics is realistic and not over-simplified.
Extensive research is currently on-going all over the world to understand the specific radio propagation characteristics that will be important for the next generation mobile communications systems – 5G. The radio channel models that are currently in wide use, for example WINNER and ITU IMT Advanced, have been developed to match the needs of 3G and 4G mobile networks. From several radio propagation aspects, they are not suitable for 5G.
One of the most obvious aspects is the radically higher carrier frequencies, in the millimeter wave (mmW) range, that are considered for 5G. Another is the use of substantially larger antenna arrays – for example massive MIMO – and narrow beam forming, which is facilitated by the short mmW wavelengths. These antenna techniques require better channel model directional resolution.
Direct device-to-device communication, an important area for 5G, presents one challenging scenario from a radio propagation perspective: dual link end mobility. Both ends of the link may be moving, which is in contrast to standard cellular communications where the base station end of the link is fixed. Another scenario is radio communication through crowds, where a large number of people may move around and temporarily block the radio link.
Part of our 5G research at Ericsson is done together with industry and academia partners in the European 7th framework project METIS. METIS has set up a dedicated channel measurement and modelling group to work with the 5G channel modelling challenge. The results of this group are published in the public deliverable “Initial channel models based on measurements.” Here, we describe challenges and requirements for the propagation modelling of important 5G scenarios. We also explain in detail a set of proposed new channel modelling approaches.
One result we can highlight is the ray-tracing inspired approach which is employed for providing realistic modelling of all the radio wave echoes which are caused by scattering objects like buildings, vehicles and people. This approach assures that the needed high resolution in radio wave propagation directions is provided.
Last week at the Wireless World Research Forum, the group Chair Tommi Jämsä of Anite presented some of the key results.
The deliverable was first released in April of this year. The work is continuing and the final METIS 5G channel model will be provided by end of February, 2015.
Jonas Medbo, Ericsson Research
Partners in the METIS channel measurement and modelling group: Anite (Chair), Ericsson, NTT DoCoMo, DoCoMo Euro-labs, Fraunhofer HHI, Nokia, Aalto University, University of Oulu, Elektrobit.