Last week I had the privilege to discuss the latest wireless industry trends with colleagues at the RAN and Backhaul conference. We discussed 5G in addition to host of other topics ranging from architecture to virtualization and much in between: spectrum, Het Nets, IoT, etc. As I contemplate the proceedings, I came to wonder about the relevance of 5G!
Today, the focus is on developing consensus on what 5G is and will be: what are the applications and which use cases will it serve? Which technologies and architectures will be best to meet the objectives? Which spectrum it will operate in? What type of physical layer it will employ? These are but a few questions the ecosystem is debating. But what everyone agrees on is that 5G will support IoT connectivity for a great number ‘things’ in addition to providing better than ever personal broadband connectivity services.
But while all this is happening, the industry is moving on with its own solutions. LTE has a long and rich roadmap to improve the performance for mobile broadband services in addition to a roadmap for IoT applications based on LTE-M. On the other hand, low-power wide area (LPWA) technologies which are specifically designed for IoT connectivity are being deployed and those based on narrowband waveform are in process of standardization at GERAN. So how and where will 5G fit with all of this?For a perspective, consider the migration of cellular technologies. 2G introduced digital technologies to gain voice capacity over 1G analog technologies. 3G added a data packet switched network to the circuit-switched voice network. 4G made the entire network based on IP. All the while, performance and user experience improved. The application for all previous migrations – personal communications – was relatively well defined which consequently led to a prominent role for technology in the evolution. But now, personal broadband over IP networks is here and data services are hugely successful. The main issue for carriers becomes how to monetize data and generate revenues which takes precedence over how to deliver more data at lower cost. From this perspective, technology is not the critical factor for 5G, but rather the applications, use cases and business models that will be key to 5G success or failure. All the while 5G technology is being defined and until the time it becomes commercially deployed (say in 10 years?) new use cases, applications, and business models are emerging and evolving. From this perspective, how will 5G matter? Will it be another technology in search of a problem to solve? For a while, it seems this is like 2000 all over again. Then, 3G came on the market, but the device form factor was not right until iPhone showed up in 2007 to pull 3G through from failure. 4G is a clear winner by improving on the services that 3G is struggling to provide in large due to its own eventual success! But as we move into 5G, how will operators receive the technology without prospects to improve revenues through new applications? The heavy capex required to roll out 5G is unlikely to happen quickly in light of stagnating revenues. Furthermore, there is the risk of IoT and other services not coming onto the 5G network as other technologies supporting these applications are evolving and can supplant anticipated 5G revenues further delaying deployments.
The in the absence of a clear winning use cases and applications why would operators make the move to 5G similar to what they did in LTE? This is where the issue lies in 5G: it’s in the value proposition and business case, not the technology. If technology is a mean to an end, as it should be, the way we think about 5G will lead to a radically different implementation making 5G truly relevant.