The 5G business case needs closer scrutiny
By Angel Dobardziev
3rd June 2018
For all its great achievements, the wireless industry at times took a ‘build it and they will come’ approach to new network investments. In the early 2000s operators invested billions on 3G spectrum licences hoping to offer a ‘multimedia experience on the move’ - only to write these off as it became clear this was not realistic at the time. Just a few years later WiMAX (IEEE’s 802.16 standard) was positioned as a possible 4G standard that would transform wireless broadband connectivity, especially in the emerging markets. Dozens of new entrants deployed WiMAX networks, most of which have since struggled (big wireless players chose LTE as their 4G standard).
The 5G hype cycle is near a fever pitch right now, as equipment vendors such as Ericsson, Nokia and Huawei step up efforts to secure mobile operator 5G trial and commercial launch commitments. Meanwhile, operators weigh the costs and benefits of taking the plunge to become the first 5G players to launch in their markets.
For sure there is plenty to like in the 5G standard. Broadband speeds of up to 1 Gigabit per second; latency as little as 1 millisecond; network partitioning techniques to offer different quality of service (QoS) options to variety of users; these are all significant improvements over 4G - should they be delivered beyond the testing labs and in the real world. Lots of clever network technologies underpin 5G, such as massive MIMO, distributed antenna systems (DAS), and network virtualisation among others. Indeed, some in the media industry hope that 5G will provide such a big boost of broadband capacity that it would accelerate a move from traditional TV to OTT video, a trend discussed here.
Amid the frenzy of activity, the industry is skirting around some fundamentals. For example, there is remarkably little focus on the fact that by far the biggest variable in the total cost of a wireless network deployment is the wireless spectrum band used. This means the business case for large-scale 5G deployments in higher frequency bands could be challenging, given the experience of WiMAX players with the same spectrum bands as that proposed for 5G.
Let’s explain these issues in as simple terms as possible. Across the wireless spectrum range there is an inverse relationship between wireless signal propagation (i.e. how far, and how well the signal travels, including through buildings and walls) and throughput (i.e. how much data can be packed into it). Lower frequency spectrum has stronger propagation but weaker throughput properties, while the opposite is true for higher frequencies. Prime spectrum in the 900Mhz-2.4 GHz band is very valuable as it provides a good balance of propagation and throughput, but outside of this bracket that balance starts to skew one way or another.
5G faces a challenge in the fact that prime wireless spectrum between 900Mhz – 2.4Ghz is already in use for 2G, 3G and 4G services, as well as for broadcasting, and government or military uses. 5G could be left with crumbs of ‘re-farmed’ spectrum within these bands that is no longer needed, or spectrum outside the prime bands. For example, in the UK, Ofcom has proposed spectrum in the 700MHz, 3.4GHz and even 26GHz bands for 5G.
Each of these 5G networks would have vast cost and performance differentials. 5G deployed in the 700 MHz spectrum would have high propagation properties, so it will need fewer antenna sites, as each cell site will have a big coverage radius – but it will have lower throughput (i.e. it will achieve lower data speeds). Conversely 5G at 3.4 GHz will have higher throughput (i.e. faster data speeds) but each cell coverage radius will be vastly smaller, and its signal will also struggle to go through building walls (i.e. lower propagation). To visualise these differences imagine an urban wireless network coverage map as a honeycomb pattern of hexagons as individual wireless cells: in the same city the 700 MHz 5G network picture will consist of just a handful of large hexagons, but the 3.4 GHz 5G network will have dozens, even hundreds of tiny ones.
The level of granularity of this honeycomb pattern of wireless cells matters hugely for the cost of the 5G network deployment and operation. Each wireless network cell, regardless of its range, requires capital and operating expenditure for its equipment including tower, antenna, electricity and backhaul as well as site rental. All things equal, an escalating number of required 5G cell sites starts to accelerate network costs so fast that they stretch the business case to a breaking point. Besides, the weak propagation of wireless spectrum above 3GHz through walls and buildings requires an additional cost of outdoor customer antennas.
Most WiMAX players, especially those deployed 3.5 GHz bands, have struggled with these inherent cost and performance limitations, even in the emerging markets with little fixed broadband infrastructure to compete against. Yes, 5G will be a more efficient wireless technology than WiMAX, but it still faces the same basic laws of physics and economics.
Besides costs, of course there is a revenue side in the 5G business case. Indeed, there is a view that new B2B use cases (e.g. IoT, connected/autonomous car services) may provide such a strong revenue boost to 5G operators that the inherent cost constraints of 5G networks in higher frequencies will be neutralised. This remains to be seen. Consumers historically accounted for approximately two thirds of mobile operator revenues through recent wireless technology iterations– and 4G has shown a reluctance to spend much more on mobile broadband service they see as just a little faster or better.
5G promises a lot, but faces a challenging business case for deployment in higher frequency bands. Wireless operators must skilfully construct their 5G business case by modelling 5G deployment costs in different spectrum bands in their local market context - before acquiring 5G spectrum; they must challenge vendors’ performance claims for network equipment in high frequency bands (i.e. 3Gzh and above); and they must carefully scrutinise 5G revenue expectations, especially lines of completely new B2B revenue streams. A ‘build it and they will come’ approach (again) is more than risky - in Santayana’s words, ‘those who cannot remember the past are condemned to repeat it’.
This article was first published in Total Telecom.