Skip to Main Content

Strategic Business Insights (SBI) logo

Internet of Things October 2020 Viewpoints

Technology Analyst: David Strachan-Olson

Race for the Edge

Why is this topic significant?

Modern cloud services, such as AWS, have revolutionized the way companies deploy web-based applications for both external and internal use. In recent months, cloud-service providers have begun to offer new types of edge-computing infrastructure to support new low-latency applications.

Description

In December 2019, Amazon Web Services (AWS) introduced Local Zones—a new type of AWS infrastructure deployment that places compute, storage, and cloud services in metropolitan areas to provide single-digit millisecond latencies to end users. At launch, Los Angeles, California, was the only Local Zone, but AWS is working to bring the infrastructure to more metropolitan areas. AWS Wavelength brings AWS services into cellular carriers' infrastructure for running low-latency applications on 5G networks. AWS recently partnered with Verizon Wireless to deploy Wavelength technology in California's San Francisco Bay Area and Boston, Massachusetts. AWS is also expanding availability of its AWS Outpost service, in which AWS deploys and supports AWS hardware and infrastructure at a client's data center or colocation center.

AWS's competitors are also increasing their edge-computing offerings. Microsoft's Azure Edge Zones provide Azure infrastructure and services to urban areas, with initial deployments in New York, New York; Los Angeles, California; and Miami, Florida. Microsoft is working with AT&T to bring Azure Edge Zones into cellular carriers' edge networks. Google recently announced the launch of a new project, Global Mobile Edge Cloud, which expands Google's cloud-computing reach with edge-computing nodes at 130 locations worldwide. Cloudflare, one of the leading content-delivery-network (CDN) providers, continues to expand its Cloudflare Workers platform to bring edge-computing capability to its already extensive CDN.

Implications

The biggest advantage of edge-computing services is the significant improvement in latency between the client and the cloud service or application. Latencies of a few milliseconds could enable many new applications and features, including real-time AI inference, IoT data analysis, connected vehicles, remote rendering, and cloud gaming. Low-latency access to computational power will become even more important as cellular companies continue to deploy 5G networks. The benefits of low-latency cellular communications are meaningless if data must still make multiple network hops to reach a distant server.

Impacts/Disruptions

Although many initial applications for edge computing target low-latency services for enterprise and consumer users, edge-computing resources will also support IoT systems in unique ways. AI inferences that operate at the network edge instead of in the cloud could reduce latency by tens or hundreds of milliseconds, which could mean the difference between an AI triggering a machine to shut down ahead of a component failing and a catastrophic failure. Even in applications where data need to migrate to a data center, edge computing can provide a valuable capability to preprocess data and reduce network traffic to the main data center.

One of the biggest obstacles to implementing edge computing is the cost of securing and deploying infrastructure across a wide range of geographical locations. For every major urban area, cloud providers need to acquire server space and will likely need to work with major cellular carriers. Deploying edge-computing resources will likely be more expensive than deploying central cloud-computing infrastructure, and companies will likely need years to deploy sufficient computing resources to all major metro areas. Users of cloud-computing services might be hesitant to invest in implementing low-latency apps, services, and features until they can deploy such services across many regions.

Scale of Impact

  • Low
  • Medium
  • High
The scale of impact for this topic is: High

Time of Impact

  • Now
  • 5 Years
  • 10 Years
  • 15 Years
The time of impact for this topic is: Now to 5 Years

Opportunities in the following industry areas:

Cloud computing, mobile-communications services, network-equipment manufacturing, managed services, corporate IT, smart manufacturing, logistics, IoT networks

Relevant to the following Explorer Technology Areas:

Understanding 5G Terminology

Why is this topic significant?

Stakeholders are well aware that 5G could affect many industries, but cutting through marketing material and buzzwords to understand the true capabilities of 5G is difficult. A major complicating factor is that 5G is not simply one thing; it represents a wide variety of technologies that improve and expand the capabilities of cellular networks.

Wireless Bands

Cellular carriers' access to new segments of wireless spectrum is one of the most important aspects of 5G. For reference, most 4G LTE (Long-Term Evolution) networks operate in frequencies between 700 megahertz (MHz) and 2,500 MHz. Low-band 5G uses bands similar to those LTE networks use but with improvements in network communications. Low-band 5G provides wide-area coverage with a modest improvement in data throughput in comparison with LTE Advanced Pro (the most advanced pre-5G cellular standard). 5G New Radio (5G-NR) refers to both high-band and midband 5G, which are new spectrum additions to cellular networks. High-band 5G technology, or millimeter-wave (mmWave) by common reference, can have very high data throughput (multiple gigabits per second), but these signals have a very short range and minimal object penetration. Spectrum for mmWave is generally allocated between 20 gigahertz (GHz) and 100 GHz. Midband 5G (sub-6 GHz) occupies a middle ground between the two other types of 5G and offers a balance between signal range and data throughput. For initial 5G deployments, many cellular carriers are focusing on low-band and midband 5G networks because of the high costs of deploying the large number of small cells necessary to support mmWave technology.

5G Technologies

The 3rd Generation Partnership Project (3GPP), a consortium of standards organizations, is the main body developing and defining the protocols and capabilities for 5G communications. Regular releases from the 3GPP incorporate technical specifications and documentation that cellular carriers, hardware suppliers, and others use to develop and implement 5G standards.

Deploying enhanced mobile broadband (eMBB) is the first major goal of early 5G networks. eMBB will improve the performance of existing cellular networks for consumer and enterprise smartphones, hot spots, laptops, and similar devices. Improvements in network architecture and the way devices communicate as well as the availability of new spectrum will serve to decrease latency, increase data throughput, and support increased device density. eMBB will also support the most data-intensive IoT devices.

For low-power Internet of Things (IoT) devices with modest data connectivity needs, carriers will continue to offer enhanced machine type communication (eMTC) and narrow-band IoT (NB-IoT). As part of the 3GPP's Release 16, eMTC and NB-IoT will become incorporated into 5G core networks and support communications in 5G-NR bands. However, 3GPP is defining a middle communications solution, 5G NR-Light, between eMBB and NB-IoT/eMTC for low-end mobile devices and IoT devices with moderate communications needs.

In industrial settings, synchronized sharing of 5G spectrum can enable improvements to quality of service and reliability. Two important capabilities relevant to 5G in industrial applications include coordinated multi-point (CoMP) and time-sensitive networking (TSN). CoMP involves using multiple cells to provide redundant communication pathways to enable ultrareliable low-latency communications in industrial settings. 5G TSN will enable 5G networks to work with other systems that need time synchronization, such as Ethernet, and to implement features to prioritize communications traffic.

Scale of Impact

  • Low
  • Medium
  • High
The scale of impact for this topic is: High

Time of Impact

  • Now
  • 5 Years
  • 10 Years
  • 15 Years
The time of impact for this topic is: Now to 10 Years

Opportunities in the following industry areas:

Mobile-communications services, network-equipment manufacturing, mobile-device manufacturing, IoT networks, smart manufacturing, logistics, smart cities

Relevant to the following Explorer Technology Areas: