The road to 5G is progressing rapidly, but prototyping is essential
Such networks by nature will be software defined so that they can be configured on the fly to cater for changing demand dynamics and to allocate resources quickly where needed. Separation of the control and data planes is essential to achieve this aim as well as to lower costs through virtualisation and to implement strict latency requirements for critical applications. Further, network intelligence needs to be pushed out to the edge as well and protocols for more efficient local routing implemented, for example, enabling direct routing between handsets.
Researchers are working on multiple fronts and at the moment are jostling to get their ideas accepted into the standards. For example, at the recent NI Week in Austin Texas, Samsung demonstrated a commercially viable Full Dimensional (FD) MIMO for spectrum bands currently in use in an attempt to be first to market with 5G capabilities. FD MIMO increases the number of antennas in a basestation enabling more users to be supported with high data rates and reliability. Samsung achieved this by implementing 3D beam forming, which enables beam energy to be directed to specific users in 3D space, and is not limited to the 2D plane as in currently the case in 4G.
James Kimery, Director of Product Marketing for RF at NI says, “FD MIMO is a first step to Massive MIMO and will probably be incorporated into LTE by 3GPP.”
James contends that 3GPP, which was created to deliver world harmonisation, is where the debate and most of the standardisation of 5G will take place. He expects that Massive MIMO will be a key part of the 5G standard with the promise of delivering a ten-fold boost to throughput.
To be 5G or 4G, that is the question
It is expected that as progress is made towards systems that are fundamentally different in architecture to 4G, that these elements will form the basis of the 5G standard. When bodies such as 3GPP eventually sign off on 5G, there will be many blurred lines between 4G implementations and early 5G proposals, in much the way 3G stole some of 4G’s thunder.
However, the key here is that a fundamental rethink on the architecture of the network to the way antennas are implemented as well as spectrum that is used needs to take place. To get 10x and 100x increases in throughput require radical re-engineering of the way wireless networks are built is required. To ensure latency and throughput improvements in the sub-6 GHz part of the spectrum Massive MIMO at the moment holds the key and represents an architectural shift in antenna design.
Anite, recently acquired by Keysight, has joined the 5G Innovation Centre at the University of Surrey (“5GIC”), the largest UK academic research centre dedicated to the development of the next generation of mobile and wireless communications.
Professor Rahim Tafazolli Director of 5GIC and Institute of Communication Systems had this to say, “5G will intelligently understand the demands of users in real time, dynamically allocating network resources depending on whether the connected device needed voice or data connectivity.”
Rahim adds, “Anite’s contribution to our 5G research programme will enable the development and testing of future technologies in a real environment and in an end to end manner.”
At the heart of the 5GIC, which is backed by both the UK government and leading global industry players, is a state-of-the-art test-bed for trialling emerging 5G ideas, proving concepts, validating standards and vendor inter-operability testing.
James Goodwin at Anite expects the air interface of 5G to be different to 4G and unlikely before 2020, although some elements will end up in 4G.
James adds, “5G will include mmWave but it is expected to appear in later variants of 5G. This is new technology and there are cost challenges in handsets to be addressed, but more importantly a unified spectrum plan is required. The WRC meeting in 2019 is the first opportunity to discuss mmWave frequencies on a global basis.”
Earlier this year, Anite announced that the Anite-led task group within the METIS project finalised the first 5G radio channel models, which will help to accelerate the development of 5G radio access technologies and future mobile industry standards.
The role of IP
A leading supplier of IP, Imagination Technologies has also joined the 5G Innovation Centre to collaborate in exploring, developing and defining underlying technologies that will power the next-generation 5G mobile communications network.
Richard Edgar, Director of Communications Technology at Imagination Technologies says, “We expect IP to dominate the wireless SoC market and consequently 5G. The progression at Imagination started in broadcast, then moved to short-range wireless (Wi-Fi and Bluetooth) and will be followed by IoT.”
According to Richard, 5G will be comprised of many standards. The 3GPP is primarily focused on cellular and this emphasis takes less account of Wi-Fi, White Space spectrum, and the emerging IoT.
Richard adds, “At the moment there are around 29 candidate spectrum bands being considered for 5G, some old ones, others new, spanning 600 MHz to 60 GHz.”
Why spend all this energy on a modulation scheme to span 600 MHz to 60 GHz when solutions already exist as every vendor has a complete Wi-Fi stack. 60 GHz can be added to the stack as an amendment and currently 60 GHz chips are small and the antennas required are also small.
Carriers are embracing Wi-Fi as a critical addition to their networks, and though it does not have the command and control of cellular (QoS), it is getting much better. Richard expects Wi-FI to remain a useful component of 5G and expects other standards to also become part of the 5G scenario such as 15.4.
“5G will be a fully comprehensive suite of wireless technology with intelligence built into the handset. The network will be very different in design and expect IP to be the way to go for small devices.
IP will be key in providing the communications expertise to the multitude of companies that are expected to embrace the IoT, home automation, Industrial IoT and so on. IoT and 5G are also set to revolutionise medicine. The key here is the billions of sensors involved and the need to communicate. Most companies will not have RF or microwave communications expertise and IP is the best way of buying this in.
Richard comments, “Companies can either buy a wireless SoC or integrate IP into their own silicon — believe that the second choice will dominate as volume costs are very important.”
5G and the IoT
The IoT poses a specific challenge to 5G. IoT is typically low bandwidth, uses small data packets, has low communications overheads and is very low power as most devices are battery powered or use energy harvesting. Further, the battery needs to last the specified life of the product. However, some elements of IoT require very low latency. It is combining all these needs in to 5G that is challenging. Incorporating multiple standards into 5G seems to be a logical progression.
IoT is independent of 5G but to capitalise on the economics of widespread networks, it will need 5G.
At NI Week, Stanford University Andrea Goldsmith remarked that the next-generation of networks need to support an exponential rate in data growth and a surge in diverse wireless devices. 5G research is not just about more data but these networks need to be more reliable, predictable and energy efficient, which is required by the emerging IoT.
This is key to enabling diverse applications such as remote surgery, autonomous vehicles, and health/wellness monitoring. The key aspect here is diversity of needs and an explosion of sensors communicating data.
Prototyping is key
James Kimery at NI says the company has a leading role to play in 5G with its ability to enable researchers and organisations to rapidly prototype wireless technology with its LabView-defined PXI platforms.
Nokia at NI Week demonstrated a 73 GHz system that delivered a peak rate of 10 Gbit/s. The use of mmWaves poses a whole set of challenges including penetration loss, diffraction loss, beam steering, and the efficiency and cost of RFICs. Started two years ago the Nokia project is resolving many of these issues such as beam forming and recently achieved a 200m range at 200m.
According to James Kimery thousands of research papers are written on wireless but there is a need to move beyond simulation and go to prototyping. NI is alleviating this problem and enabling researchers to implement prototyping and have their voices heard. Standardisation bodies require a prototype otherwise the idea lacks credibility — it is difficult to invest in an idea without a prototype. Prototyping provides a glimpse of the road to commercialisation.
All the companies involved, of which only a few have been covered in this article are looking at different aspects of 5G. While we can agree on what 5G will do in terms of throughput, latency, power efficiency, reliability and so on, what is more opaque is how such a network is realised. Researchers looking to define 5G and companies looking to leverage these standards will need to do cover a lot of testing and prototyping of complex and emerging technologies. While 5G might encompass many standards already in existence, the network will need the intelligence to select the best protocol for the use case and implement in the most efficient manner possible. New standards will come into play to take advantage of spectrum that is unlicensed or in the mmWave bands, but it suffices to say that at this level of complexity prototyping and network test beds will be essential in getting to the point where 5G can be implemented. As is the case with complexity, it is best dealt with reuse, whether on the test side or implementation side in the form of IP.
To conclude, there is significant progress being made on 5G the point where standards start to come into play is near. Most of the next decade will be dominated by 4G and improvements to 4G, but the seeds for 5G have been sewn.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
