Gillott: 3 wrong assumptions (and one right one) about 5G

If there is one thing the wireless industry loves, it is a good debate about the next generation of technology before the current generation has been completely deployed.  With each successive ‘G’ in the evolution of the wireless network, the industry gets more aggressive in promoting the planned capabilities of the next generation.  This is especially true with 5G which, despite the claims of some operators, will not truly be deployed until the end of this decade.

In the meantime, the wireless industry is still deploying LTE and upgrading that technology with new releases.  In fact, the reality is that the majority of the world’s mobile connections today are 3G - 4G LTE connections will not make up the majority of connections until early next decade.  And looking ahead to when 5G networks are deployed, 5G connections are forecast to never be more than 10 percent of the global base for the first six years of next decade.

So why are we, as an industry, getting all excited about 5G networks now?  And with all the competing claims, what actually constitutes a 5G network?  To answer these questions, lets address some of the major rumours and common discussion points about 5G.

1. 5G is a major step in technology, like the move from 3G to 4G - Wrong

At an analyst conference last week in Dallas, AT&T, T-Mobile, Sprint, Telefonica and a host of other mobile operators from Latin America and the Caribbean discussed the move to 5G and what it would require.  In these presentations, and in subsequent discussions with vendors such as Nokia and Ericsson, it became clear that ‘5G’ is defined as 3GPP Release 15.  Release 14 is considered part of the LTE family - Release 15 is 5G.  In effect, the industry has drawn a technical line in the sand and declared that Rel 15 is the new generation and Rel 14 is not.

Interestingly, Release 14 LTE includes many technical concepts that will be needed for Release 15.  And Release 14 builds on Release 13.  Mobile network technologies are now so complex (and efficient compared to previous generations) that it is very difficult to simply rip out an existing network and deploy a new version.  Instead, LTE networks really do evolve, with each release building on the previous version and implementing new solutions.

The move from 3G to LTE was far more significant in terms of technology evolution, since LTE introduced the use of OFDM in the air interface together with major changes to the network core.  The 3G air interface was based on CDMA (remember there were multiple 3G standards but all used variations of CDMA).  But 5G still uses OFDM (some enhancements will be made in later releases) in the air interface and is therefore not a ‘rip and replace’ upgrade from 4G.

So, 5G is Release 15 of the 3GPP, which itself builds on Release 14 of 3GPP (LTE Advanced Pro).  Ergo, 5G can be thought of as an evolution of LTE.

2. 5G networks are just for mmWave frequencies - Wrong

Many people have talked about how 5G networks will use spectrum in bands greater than 6 GHz - this is true.  But this does not mean that 5G will only be deployed in higher bands (it will be seen in sub-6 GHz bands as well) or that all networks using higher bands will be 5G (they will not).  

Remember that the old LMDS bands (28 GHz and 31 GHz) are again being considered for fixed wireless services, but this time labelled as ‘5G’.  What has changed between LMDS and the new proposals are advances in antenna and radio technology - these advances are making the discussion about the use of the 28 GHz bands possible.  And not just for fixed wireless applications - trials are planned using the 28 GHz band for mobile applications as well.

But many companies, including Cambridge Broadband Networks and Siklu, have been making equipment that uses mmWave bands for point-multipoint networks for years.  These networks are not 5G but simply make effective use of the spectrum bands available. 

So, mmWave does not automatically mean 5G and 5G will not only use mmWave bands. 

3. 5G will be driven by the needs of the IoT industry - Wrong

IoT (Internet of Things, in case you have been living under a rock) networks may seem new and certainly the hype cycle is at its peak, but from a wireless perspective, these solutions have been around for a while - we just used to call them M2M (Machine to Machine).  The major difference between M2M and IoT is that IoT layers big data analytics capabilities onto the device connectivity.  Obviously, M2M was around long before 5G was discussed.  So why is 5G being so closely tied to IoT by some in the industry?

The answer seems to be nothing more complicated than timing.  IoT needs to connect billions of devices to a low cost network that provides an economical data transfer.  Low Power WAN networks (such as Ingenu and Sigfox) achieve this (both are building their networks today) whereas most will agree that a 4G LTE connection is overkill for many IoT applications that just need to send a few bytes of data.  Certainly LTE connections can be used for IoT applications that require higher bandwidth (such as video, for example).

One thing 5G does need is a compelling business case - the so-called ‘killer application’.  So what better than 5G IoT?  The reality is that Narrowband LTE (NB LTE) is actually specified in 3GPP Release 13 (which was finalized in June 2016) and is likely to be deployed in major networks in 2017.  As the name suggests NB LTE is a LP WAN technology and is designed specifically for low power IoT applications.  But of course, Release 13 is classed as LTE Advanced Pro, not 5G.

Certainly, Release 14 and Release 15 will build on the narrowband capabilities but it is clear that we do not have to wait for 5G to implement IoT and that IoT networks are not by definition 5G. 

4. 5G needs MIMO, small cells and edge Computing architectures - True

The wireless industry has some very aggressive goals for 5G network performance, in terms of both bandwidth and latency.  The reality is that these goals cannot be met without changing the fundamental architecture of mobile networks as we know them.  For example, lower latency can be achieved by moving the content or application processor closer to the radio (edge computing) and MIMO antennas can be used to increase the bandwidth available to the end user device.  And small cells, of course, are part of the solution set to densify networks and hence increase capacity.

iGR believes that 5G cannot be implemented without small cells, MIMO antennas and edge computing architectures if 5G is to meet its performance goals.  But small cells, MIMO and edge architectures do not have to wait for 5G - in fact, these solutions are being deployed today in various forms.

Summary

Putting this all together, a clearer picture of 5G emerges:

- 5G is technically defined by 3GPP Release 15 and subsequent releases

- 5G will use a wide range of spectrum bands, including mmWave and sub-6 GHz

- 5G will support a wide range of business models and applications, including IoT, fixed wireless, mobile broadband and others yet-to-be-defined

- 5G is an evolution of the LTE technologies and will require MIMO antennas, small cells and edge computing architectures.

Of course, defining 5G using a technical definition does not stop the marketers from using ‘5G’ for other capabilities or prior to the implementation of Release 15, etc.  As we joke at iGR, by the time 5G is technically deployed, the branding will be using 7G :)

Iain Gillott, the founder and president of iGR, is an acknowledged wireless and mobile industry authority and an accomplished presenter. Mr. Gillott has been involved in the wireless industry, as both a vendor and analyst, for more than 22 years. iGR was founded in 2000 in order to provide in-depth market analysis and data focused exclusively on the wireless and mobile industry. In recent years, research has expanded to cover broadband telecom services to the home, as homes and businesses have become more connected.