6G Talk: Experimentation Platform, 5G Test Network | Olli Liinamaa
Hello and welcome to follow this presentation about the Experimentation Platform, 5G Test Network, which is used for 6G Flagship vertical experiments and as a test infrastructure for testing whatever prototypes the researchers come up with. My name is Olli Liinamaa. I'm working here as part of the 6G Flagship programme as a test network lead.
And my duty is really to align the infrastructure to serve then the needs and the requirements coming from these research programmes and individual researchers so that we would have really a working experimentation platform for all the good innovations that are popping up from this research. This presentation consists of three main parts. First, this 6G Flagship overall. And then the related test infrastructure, how this infra is serving and then the activities coming from the research projects. Second part is this 5GTN as an infrastructure for research projects, individual researchers, companies, students to play with.
And then the 3rd part is about the experiences from vertical experiments that we have done using this infrastructure so far. Here we see an overview about the Finnish 6G Flagship. Flagship is an 8-year- lasting programme funded by the Academy of Finland, the University [of Oulu] itself, and then the related separately funded projects. It is investigating new technologies, materials, business models, regulation, architectures. Assuming that 2030 we would have such a society which is data driven, sustainable, and using services enabled by near-instant unlimited wireless connectivity. The programme has four strategic research areas, as you can see here.
First one being the wireless connectivity solutions itself. So the frequencies, new algorithms, new protocols, hardcore wireless research. Second one being the devices and circuit technologies. So the materials when go into new higher frequencies, then it has a major impact also to the circuits, to the materials that we are using.
Third one being this distributed intelligence wireless computing. So the architectures, where do you have the processing? How do you utilize cloud technologies? And then how to solve the sustainability problems in an optimal manner so that the processing would be in the right location whether it is in the terminal itself where the data is sensed or provided or whether it is in the edge of the network or somewhere centralized in the core. And then the fourth area is the sustainable human-centric services and applications where we are doing all these experimentations on different use cases and verticals, as we say. So that putting this technology to several vertical areas.
Selected and strategic, important areas being health, industry, vehicular, and energy. The goals for this second phase is really that we are providing 6G technology enablers. Then this test network itself, currently 5GTN, would become 6GTN, 6G test network.
Then doing 6G vertical applications. And then, globally, to bring and enable and create and strengthen the 6G vision leadership: so what the 6G will be. And those are really the impacts that we are expecting from all that.
So, defining sustainable future society and not just wait someone else to do it. And, then as said, the progress on these strategic vertical areas. Then third one being increased academic and industrial collaboration globally. And then the 6G test network infrastructure as part of global network of test beds.
Here we have this strategic research area for sustainable human-centric services and applications in more detail. Here you can see that we have licensed and non-licensed wireless access alternatives. We are serving high speed and low speed applications. We have wide selections of different frequencies in use. We have outdoor coverages, indoor coverage. We have different architectures in use.
We have this edge processing capabilities close. So, providing then the processing power very close to user access. And then we have cloud-based core environment, and then the related transmission systems, with unlimited capacity in practice. And then the services, infrastructure for services, and interconnections, then to other test networks or public internet if that serves the purpose. Here we are really developing services and applications, different use cases, different verticals. So, forget the plain old smartphone and turn the other gadgets to mobile devices.
Here you see on the picture that there are cars, excavators, drones already visible. So, we want to cut the cables and integrate the modem to end devices and unleash really the potential of mobility of 5G or 6G then wider in this society. Create backend services with needed data analytics and algorithms that this is of course, then, part of the research projects. And then you need to do something with the data which is provided from the robot or from a measuring device. And we are also creating user interfaces or showing and using data in a meaningful way. Digital twins, augmented reality, virtual reality, embedded collected data into those, and show it.
And then also verifying business potential and solution feasibility. For the 6G, we have 3 development paths. First one being the obvious one. 3GPP is defining the global standards for mobile networks and they are progressing assuming then the 6G standards to be available from certain point onwards. And that's providing secured evolution from existing infrastructures then to future mobile network versions towards 6G. The middle part is then this open architectures part.
There we have then perhaps low CapEx with moderate performance, then networks but, of course, then potentially requiring higher OpEx. And there we have interesting research questions like security, energy consumption, jitter, the performance stability overall. And here we see also then this O-RAN Alliance alternatives coming into the play. So this is one alternative. And then the third one is this disruptive path where we do not care about the legacy that whether we need to create an evolution from a certain platform to another, with just software updates, but instead, select a new path and then play with that and see whether it is feasible or not. Here in this infrastructure we are targeting to serve all of these avenues.
This test network, as such, has been here since 2015. Then, we started with the existing elements, existing technology that we had at that time. So, building first network assets, 4G LTE, outdoor and indoor coverage. Then moving on then to the first proofs of concept, devices, where we were introducing in this millimeter wave connectivity and even introducing that in the Olympics 2018. And then we opened the network for public use. So we really have our own SIM card, own network, so truly just plug in the SIM card then to your device and then you are able then to use this network as such.
And of course then there has been a lot of evolution since that 5G new radios were introduced. Now we are living the phase where we are having these 5G-advanced technologies under development. We are connecting this test network wider and more in the closely than integrated with other test beds in Europe as part of this ESFRI [European Strategy Forum on Research Infrastructures] programme where research infrastructures from different countries are collaborating more and then collecting then the data and assets from different locations and integrating and increasing then the collaboration between different academic institutes. Latest discovery, latest introduction here was this commercial millimeter-wave solution that was recently opened now during during past months.
And then the evolution continues and we really target to be 6G test network as soon as then the related technologies become available. We are in a luxury position here at the university campus and its Flagship to run really outdoor 5G with these commercial 3.5 GHz band. And that's due to regulation. In Finland, there are certain areas, they are called TTO areas, referring then to the technology research development, testing, and teaching activities.
In these areas, there are restrictions for using radio frequencies for mobile networks. And here on the map you see that our Linnanmaa University campus area is part of such area. So, we are together with the Finnish regulator, then able to run these 5G networks also outdoors in our neighborhood. And that's it for part one. And now we look this 5G desk network in more detail. We are having 5G on band n78, 3.5 GHz, and then n258 for millimeter waves.
So 24 GHz band. We are having both the non-standalone and standalone architectures in use. 4G on very many frequencies since we have had that already for quite some time.
IoT is served on Band 28. There we have hundreds of IoT sensors partially connected with narrowband IoT, and then of course partially with other technologies such as LoRa. Then we have these edge-processing servers. So, if there are such needs that researchers would like to have processing very close to the edge of the network or even inside the terminal.
Then we are doing integrations to different verticals. We have assets like energy consumption or production measurement environment in use and serving the researchers. Connected 6G cars, smart excavators, or drones as examples. And, of course, then the evolution towards 6G as part of the infrastructure evolution. So whenever new assets become available, then as soon as we are able then we are integrating those interesting things to be part of this infrastructure, serving researchers. 5G coverage is primarily here at the university campus area.
And here you can see the heat map, which is from last summer. So green is good, red is bad. Frequency license, 3.5 GHz, 60 megahertz band is in use at the campus. And this is one measurement which was done, as said. And there's 5G outdoor macro, which is serving this and how the measurements were done. So this is also one example on these kind of measurement devices and measurement capabilities that we have for radio parameters.
So that these kind of information can be used also as part of the research activities. And here we see an architectural overview of the network starting from the device, connected devices having ready this outdoor and indoor base station capabilities and then the related functions inside the network, including the transmissions and then the interconnections to other networks. Starting really from the left, you can see then the user devices, and we are definitely using not only the traditional, classic smartphones, but other type of gadgets or devices such as augmented reality glasses. I can just see here. So, showing data for the user on own screen. And this of course, then, opens an interesting question that, hey, what sort of processing should be here in this device itself? Or, or whether it should be there at the edge of the network or somewhere centralized.
And how do you deliver that then to the user in a meaningful way. Plenty of IoT sensors installed. Some are like this very small, mobile sensors.
Some are stationary permanently installed into a roof or a ceiling measuring different things. And of course then, as I mentioned, in addition to smartphones, or tablets, or laptops, then there are 5G modems already in place, which can be used. Here's one example. Where we have this 5G test network SIM card included. And then this is providing then the mobile connectivity, and then you have to have something where this is then plucked in to really make the existing stationary device to be a mobile device.
If the end device does not have then the related or needed processing power, we may introduce here some small computer, like in this case, a Raspberry computer then as part of the integration. So, this is really what we do as part of our activities. The base stations, of course, then outdoor macros are outside there in the mast, but indoor coverage is built using indoor devices.
So these are really tangible base station devices that we have here at the campus area. And where we are providing then the connectivity. You see also in the picture then the other elements that we have: different core alternatives then switches, routers.
Then the possibility of routing the data locally to some edge-processing environment or some other solution, some other server stack here at the university campus. In the campus intranet itself, where we are then able to spot all the hops, how the data is transmitted, how many jumps does it have. Do we have unknown parameters in between? And then we have, of course, then interconnections to other networks or other test sites in Finland or in other countries. Here, you see also integration to our key partner Nokia, where we are doing a close collaboration, and some of these capabilities such as subscriber management and then subscriber databases are provided as a service from Nokia site.
For the core network, we have two alternatives. Traditionally, in 4G, there has been this LTE and EPC network-enhanced packet core network as a core element. And this was developed and standardized to be a smooth evolution utilizing then fully these existing 4G capabilities. So, the 5G base station is introduced in parallel, and then the users are able then to have the capacity from both and then benefiting really with best possible performance and best performance also from the evolution point of view. Reusing the existing assets. So that's one alternative and that's what the commercial networks are typically now about.
But then the native 5G network is there as an option as well. So-called standalone network where the session is started directly through 5G and it has no relation then to any legacy technologies. So, these two alternatives are there for researchers. Some solutions in more detail. Here's the, the latest opening this millimeter wave, 850 MHz band from 24 GHz frequency area.
This installation is here at the Tietotalo [at theUniversity of Oulu campus], second lobby. And then you just can see the details there as a picture. So, the n258 radio, millimeter wave radio ,shooting then downwards from the third floor. And there is anchor 4G -this is at the moment, non standalone alternative.
So, it requires then this 4G anchor frequency there. There are a couple of options for that. And then both options or both technologies are connected then to the base band and then the core element and then serving then the researchers.
Second solution that we are having here in more detail is this indoor 5G. And here we have these, as I already showed, these, kind of 5G and 4G base station elements there installed in the heart of the campus; in the Tellus innovation area, in FabLab area, and then there's related surrounding laboratories like the university polytechnics robotics lab is next to the Fab Lab. There we have also then these assets. These 4G / 5G elements are connected via these kind of hub, then to the basement and then do the core that we have here at the university. So there we have the indoor coverage: having frequencies B7 as 4G, and then the n78 as 5G in use. Second solution here or the next solution is this open source alternative.
We are doing 5G experimentation following also then this open air interface alliance progress. So, the architectures and assets provided there. And these are, of course, then highly customizable solutions for deploying these 5G networks, particularly for research purposes for our other academic use. So, these are really such sandbox where we can freely modify then the functionalities as much as we wish, and develop new, the previous ones where are more like this kind of production network where we want to have that as stable as possible so that the activities would not then harm the service from other users. But in this case, we have really the flexibility of introducing standard interfaces and reference architectures and developing network functions based on these open source assets. And here as an example, we have this open 5GC Core as an alternative where we have both these user plane functions and then the control plane functions embedded all into a single device.
And then we can modify those and develop those freely forward and then connect their commercial 5G base station element into it; or, a software-based radio elements if there are such projects, which are then more focusing on developing these RF units or the radio units. The latest one what we are currently working on is this OAIBOX, an open source 5G solution, everything in a box. So, that is also an alternative that we are providing then here as an open source playground and for research projects for this 6G Flagship.
And, of course, then for the network, we have also tools for performance measurements for measuring KPIs, various KPIs. And there are two main tools for this purpose. One being the Keysight NEMO.
Both, the NEMO Handy and NEMO Outdoor solutions, are in use. And then the second key tool is called Qosium from a local company called Kaitotek. These tools are used for measuring radio interface, radio parameters, particularly with this NEMO tool. And then the network QoE / QoS are then more in the network side, provided by Qosium tool. In addition, of course, then to traditional open tools like Reef Shark or any oscilloscope, or so. But these two solutions that we have are then for, really, as key KPI tools that we have.
Both of them, of course, then have various data storing options and visualization options. On the left-hand side, you see the heat map that was already shown from this NEMO tool. And in this example we have then the received power from base station visualized in different colors. And then of course these are adjustable that how the trigger values should be defined, and what is really the parameter that we are measuring from this radio interface. And then for the network side, this Qosium tool is used for measuring the performance between tool agents or probes that are installed in the network.
It may be in a terminal itself, in the user device. Or, it may be in some network elements in the chain. And this tool is measuring really then these network characteristic, network performance: how many packets were received, how many packets were lost, how much bandwidth was received, et cetera. And then these delay jitter parameters on one way. So, not a round trip, but really one way so that the data is really valid and relevant. That was end of part two.
Okay. Let's then have some true evidence of this platform that we are really using in different projects. So, some experiences about the utilization or the different verticals that we have done so far with this test network. I already mentioned that we have hundreds of sensors integrated into this university campus area.
Here we see a screenshot out of those. There are sensors here at the campus area and we are really opening and showing and allowing then this real time data available in those campus net pages, smartcampus.oulu.fi/manage. And then we have also published a history data for full year data from these sensors. And then openly that is also available if you have analytics, which would require long-term data or big data, so that's also free to use and the address being shown on the material. But let's have now a look real time what the sensors are showing at the moment. Okay.
Here we see the smartcampus.oulu.fi webpage and if you open the map, we see the visualization of this university campus area and all these bubbles are sensors. Some seem to be non-operational, but most are already operational. And if you zoom in and pick any of these sensors, then we are able to see the data out of it. Let's take, for example, from one of those big lecture rooms, L3, one of the sensors.
This seems to be having voice data also included. So, we'll open that. Here we see the values, key values at the moment and how it has been.
Here's even a picture about the sensor and if you open then the Grafana, here we see real time values for various parameters. For example, light here, sound peak, motion sensors, sound, average temperature, et cetera, CO2. And here we see, for example, today, 23rd of March that there seems to be that there's a lecture started because the lights were turned on and that is then included in this data. So these kind of data, real time is available, as I said.
And then the long-term history data for a full year is published on our web pages. The next example is from our sustainability research activities and then related assets that we have. We have done this kind of experimentation already in many projects and this is one of the key assets that we are developing also further. On the university rooftop, we have 18 solar panels, which are connected for research purpose.
Of course there are plenty more solar panels. But these 18 ones are with detailed censoring data. So, there are environmental sensors connected to those panels itself. So the energy production is one of the key themes that we have here. You can see also in the picture that they are different angles even included. Then there's energy forecasting data, which we are able to get from Finnish Meteorological Institute.
So we have the forecast that how much energy are we going to have tomorrow or day after tomorrow, and developing algorithms then based on that data that whether this real energy production from this solar panels are matching really the forecast and then developing algorithms to improve that. And of course then there's this connectivity point of view. Of course we are a mobile test network, so there's 5G next to these panels, shown also in the picture, is connected to the same measurement environment and we are able then to see the consumed energy versus produced energy, and then the relation into that.
So that we are able then to evaluate the sustainability of that solution against the production, local production. Here we see that in more detail. So we have these radio modules and then the related base band and then the switch elements. They are connected to this measurement system. And that consumption is there on the red. There seems to be on that particular day pretty average consumption, and then some sort of peak, small peak before noon, and then continuing a stable consumption.
And then the production from 18 panels is there on the curve. It goes above this consumption there during the daytime. And there we are producing more energy than we are consuming. And then later when the sun goes down, then we are again, consumers, not producers.
But here similarly this is real time data. That sample is from last August. But let's see how things are today and now. Here on this webpage, we see the energy consumption from yesterday, 22nd of March 23. And similarly then the 5G system there is stably consuming about 900 watts.
And then the solar panels are then providing a bit more during the daytime and then less, of course, then during the dark time. And today, 23rd afternoon, let's see if we have a data for today here. Yes. This is really updating real time. So that it seems to be sun shining now at the moment because there's a peak on energy production.
And then the consumption remains stable. So these sort of assets, and then user interfaces we have already now, and developing further as part of the activities. Of course, there's plenty of room for more research to be done and developing this infrastructure further. Like bringing in different energy sources like windmills and then storing energy to batteries, and then using that when there is no production. And then of course then adjusting, then the network behavior according to the energy forecast or the available energy.
Then dynamically so that take these kind of steps to the sustainability and that way also then to address really seriously the United Nations sustainability goals, which is part of the Flagship initiatives. Next example is smart excavator. So, a digger or this kind of heavy infra machine connected via 5G using this test network assets. So, the 5G modem was integrated to this excavator and then with different antenna structures, we were researching, investigating those and then controlling this excavator remotely. And then following, then the autonomous operations that the excavator was doing.
And then of course there was plenty of sensoring integrated to these excavators like lidars and cameras and other sensors. And then the data, all that data was transmitted over this test network then to the backend service where this data was then further processed, and then the machine was controlled on remote connections. So this sort of, assets we have and have tested already this and developing, of course, then this path as industry or vehicular connectivity then further in this programme. And as part of the test network assets.
Next example is then drones. Drones are very handy in carrying 5G terminals or user equipment as cargo. And, of course then, we are doing integration as well so that the drone itself would be then a flying 5G user equipment. And there can be this kind of, for example, these tools that we introduced earlier. There's NEMO tool or Qosium tool can be then integrated to a phone, or a small computer. And then flown with the drone there and then that sort of assets we have developed and then providing then the new ways of doing field testing and radio performance measurements or network performance measurements.
In this was one project activity where these assets were pushed forward and there was also then this kind of measurement management assets that were developed. And then trialing management assets that we were able to develop during the project. So there are web portals that you can use for defining the trial capabilities, trial requirements, and then pushing that to the infrastructure itself and then reserving that asset for yourself. And then storing the data coming out from the measurement. So drones, very handy and usable tool and part of our research initiatives. And then test cars.
We have two 6G test cars in use. And we have already done 5G integration also then to these vehicles. So, 5G modem as I showed then really connected then via suitable adapters then to the car CAN bus.
And then after that, then when the car was driving, then we were able then to see through 5G network that what are the parameters coming out from the car itself. So, a big step towards autonomous or remotely driving cars and really the future of mobility partially defined here in this Flagship programme. Here we see a collection of other, and partially the same result, examples that we have done as experiments utilizing this platform, not all directly then using this production network, but very much with the same experiences.
Just go to YouTube and check for 6G Flagship channel for demonstration videos, and there are examples on this millimeter wave communication, visible light communication, lighter-than-air drones, 3D scanners, sensing on high frequencies on telehealth frequencies, e-health applications, and then some location awareness. So this kind of sensing and then accurate positioning technologies based on the radio interfaces. So plenty of these kind of research examples that we have published or also. And so just click in and have a look on more detail details.
So that was it. So, 5GTN is providing these experimentation possibilities, serving all paths towards the 6G: evolutionary path, open architecture's path, and then this rollercoaster path, this disruptive pathway where we don't care about any limitations or any legacy technologies. So, just plug into the 5G test network and become the user of this mobile network. Welcome to experience your innovations, to be part of that infrastructure that we're developing. Thank you for watching.