Thursday, January 12, 2017

Concerns about About Need for 5G Bandwidth are Substantaily Correct; Also Irrelevant

As often happens, important new technologies and products have a “hype cycle,” where high expectations then encounter fatigue, and only later is the value obvious. So it is not surprising that some are skeptical about 5G.

To be sure, there also is a certain amount of what amounts to posturing, as always happens. Firms and countries that fear they will not be able to move as fast have good reasons to downplay the importance of early 5G deployments. Other firms, in a position to do so, obviously will tout their leadership.

As a matter of industrial policy, rightly or wrongly, policymakers also see potential “leadership” in various industries is seen as contingent on early leadership in 5G deployment.

There are some obvious challenges, especially related to the capital cost of creating a much-denser network, with more-sophisticated cell sites, to support the strategic millimeter-wave spectrum that 5G will bring, in huge amounts. Aside from the new investments in platform, 5G is going to require smaller cells than has been possible for 2G, 3G and 4G. The reason is physics: millimeter wave signals do not travel as far, and do not penetrate obstacles such as glass and concrete.

One recent study by Rethink Research, for example, suggests that the obstacles range from the typical (sites) and site approvals to backhaul. But there are lot of issues. And that refers only to the physical challenge of creating the 5G access network. Ultimately more challenging is the business model for 5G.

Specifically, will 5G actually lead to creation of huge new businesses, and revenue streams, that not only justify building 5G, but also are substantial enough to offset legacy revenue source declines. In a nutshell, the key issue is that revenue from sales of services to human beings will be under pressure. The big hope is that new services to support enterprise applications will be huge new drivers of revenue and growth.

One can argue that human beings, as consumers do not “need” 5G bandwidth. That is substantially correct. But bandwidth is not the issue. Aside from latency performance, and the changes in core networks (virtualization), the paramount issue is whether 5G creates a platform for new machine-to-machine services.

source: Rethink Research (chart shows perceived issues on a base of 75 respondents)

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...