4 minute read 31 May 2018
Man holding smartphone

How IoT connectivity technologies might not be evolving at pace


EY Global

Multidisciplinary professional services organization

4 minute read 31 May 2018

IoT technology is evolving fast, but there are still concerns, among some, about adopting it.

The adoption of IoT technologies has, for some organizations, been hindered by their current limitations. But the technology is rapidly evolving, creating ever-lower barriers to entry.

Power consumption

The primary objective of evolving wireless technologies from 2G to 4G was link capacity, driven by access to the internet for mobile devices. But for many IoT applications, link capacity is not so important: if telemetry data amounts to only a few bytes per hour, link capacity can be tiny.

The more important consideration for radio modems is low energy consumption, because telemetric sensors often cannot be connected to the grid.

This defines a new class of solution in wide area networking, called low-power wireless access (LPWA). There are few technologies classified as LPWA, such as 3GPP NB-IoT, Weightless, Sigfox and IEEE. They differ in the delivery business models, costs and technical parameters. Each will suit specific IoT applications.

Computing power

One factor that often accompanies LPWA connectivity is the constrained computing power of IoT devices. There are significant efforts to make devices autonomous, with built-in machine learning and AI processing units, which probably indicates that augmented processing by a central unit is currently more trusted by developers.

Yet when a device has almost no data, it does not need high computing power — and limiting computing resources cuts costs.

Low-cost, battery-powered devices with simple computational power and connectivity will enable massive IoT deployment.

Low-cost, battery-powered devices with simple computational power and connectivity will enable massive IoT deployment for telemetry and environment sensing.

Link latency

Another important factor that has limited some technology is link latency, which is not ideal for industrial IoT control loops.

3GPP 5G has posed link latency challenges in wide area wireless networks. Ultra-reliable low latency communications (URLLC) have 1millisecond latency, which would suit applications classified as critical machine type communication (MTC), as well as many industrial applications, traffic safety and control and remote manufacturing.

But while URLCC has proved itself suitable for local area networking (for example, in a store), it has yet to be deployed on an economically feasible, ubiquitous, continuous, wide area URLLC 5G network.

Real time or low latency is part of the required solution for many applications. So for application level guarantees, more complicated (and still relatively unpopular) protocols must be used for data distribution.

Radio spectrum resources

For massive deployment of IoT and mobile broadband, greater radio spectrum is needed; this is a scarce resource that can significantly impact IoT economics.

There are two interesting developments here: utilization of high-frequency bands and a shared spectrum in lower bands.

For massive deployment of IoT and mobile broadband, greater radio spectrum is needed; this is a scarce resource that can significantly impact IoT economics.

Millimeter waves in 60GHz are practically unoccupied, because of small coverage ranges in this band; but recent developments in massive multiple input and multiple output (MIMO) and beamforming can make it attractive.

Another approach is sharing a spectrum that is licensed but unused. The spectrum-sharing concept, known in the US as CBRS (citizens broadband radio service), provides technically constrained regulatory access to the 3.5GHz band, which is allocated and used by license holders. CBRS opens the door for so-called private-LTE deployments with primary applications for IoT.

A similar type of access is under consideration by regulators all over the world, although US CBRS success would be all the endorsement needed to unlock these opportunities.

Areas where greater evolution is needed

  • Security
    Connectivity security is a widespread concern. In the case of LPWA and low computing power devices, even transport layer security requires novel approaches and solutions that cannot be imported from IT.
    In the IoT space, the variety of devices running proprietary firmware will require something different from the systemic IT approach to security; no user can be expected to become a security expert for all their IoT devices.
  • Position tracking
    Another aspect of IoT evolution is device position tracking. The technology available now has centimeter-accuracy, which is appropriate for asset tracking and safety applications.
    For robotics control or machine guidance, however, this is not accurate enough, so it is worth watching research efforts for evolution in this domain.


While some areas of IoT technology are evolving faster than others, it is worth remembering that devices exchange data. Many protocols have been used on a massive scale for many years already, such as MQTT for telemetry.

As long as the data is exchanged by devices in a closed environment, it will be understood. For open environments, though, protocols and models are still to be decided.

There will no doubt be developments in all these areas in the very near future, which will change the game again.


Protocols and models have not yet been established for scenarios in which devices exchange IoT data in open environments. 

About this article


EY Global

Multidisciplinary professional services organization