Society is in a data-intensive era of computing. Companies depend on big data platforms to reach conclusions in minutes that would take days without the help of technology. Artificial intelligence (AI) has also come into the mainstream. It gets utilized in industries ranging from health care to marketing. Consumers benefit from it too, such as when they invest in smart home technologies that give information, entertain and more.
As these technologies become ever more prominent, people often overlook the impacts such innovations have on computing networks.
Most IT leaders say their networks can’t handle big data
A recent survey from Accenture polled senior IT and business leaders, and only 43 percent felt their enterprise networks could handle the increased demands of big data and the Internet of Things (IoT). Moreover, they said business demands no longer kept pace with the capabilities of their IT departments.
They got into more specifics by citing issues such as old equipment, a lack of network skills and insufficient funding as among the problems faced. Respondents cited software-defined networking (SDN) as one of the ways around these networking difficulties.
SDN depends on software to centrally control and program the network according to changing network demands. Accenture’s research showed more than three-quarters of respondents had already deployed SDN or were in the process of doing it.
Insufficient data infrastructure hinders AI adoption
Company executives are quickly reaching an understanding that AI could unlock insights about their businesses that might otherwise go unnoticed. However, a 2018 study from MIT Technology Review Insights and Pure Storage found that infrastructure issues made it difficult for companies to adopt AI as intended.
The report brought up the increasing volume and velocity of data as being problematic, coupled with real-time processing needs that also tax infrastructure. Furthermore, 78 percent of respondents had difficulty using their data at scale, with 43 percent mentioned infrastructure as a potential barrier to AI adoption.
Solving those issues often means transitioning to a scalable, elastic infrastructure that can handle the specific demands of these data-intensive processes. Taking that approach requires substantial investments, but companies must recognize how infrastructure issues hold them back from seizing the full potential of emerging technologies.
Making improvements may require efforts to enhance certain aspects of the network infrastructure, too. One possibility is to use microducts, which are tiny conduits that hold microcables. Operators can deploy as many as 144 fibers when using them, which allows for future additional capacity and expansion.
Smart devices may slow down home networks
There’s seemingly no limit to the number or type of smart devices a person could have in their home. For example, they might have a smart doorbell, speaker, thermostat and even toys for kids.
When people decide to invest in smart home products, some become concerned about reduced home network speeds due to the growing number of devices they connect. The amount that a network slows down—or if it does—depends largely on how a person spends time in their smart home.
Video is especially bandwidth-intensive. If a person wants to stream high-definition or 4K videos and finds their home network already has an ongoing slowness problem, they may decide to speed things up.
The first step to take is for a person to perform a speed test and see whether their network speeds can handle the smart home activities they want to do. For example, a minimum of 50 megabits per second is best for a household that wants to simultaneously stream 4K videos and do any other internet activities.
Making improvements may involve investing in a new router. The latest models are better able to handle multiple connected devices while preserving more bandwidth. Alternatively, people who often encounter Wi-Fi “dead spots” in their homes may want to consider mesh networks. They create a web of internet connectivity around an entire home.
Edge computing investments projected to rise
Edge computing allows processing some data at the borders of a network rather than transmitting the information back to a centralized location, such as the cloud. It enables companies to expand their computing capacity through IoT devices and edge-based data centers. Doing things this way does not substantially increase the strain on the core network each time a company adds a new item to the edge network.
Edge computing also reduces latency, making it well-suited for machine learning tasks that must deliver near-instant results. For example, if an airport uses machine learning for passenger verification or a police force deploys it to check for an alleged criminal at a busy train station, delays could make a usually promising way to use the technology almost useless.
Statistics from Juniper Research for the period between 2019-2024 indicate that annual spending on edge computing is going up. The firm expects the yearly expenditures to reach $11.2 billion by 2024. The 2019 investments were $1.3 billion by comparison. If the analysis from Juniper Research pans out, the change will represent an annual growth rate of 52.9 percent.
Mobile operators must enhance infrastructure before 5G arrives
The 5G network is going through testing, and widespread rollouts are set to happen within the next year. The expectation is that 5G technology will be such an improvement over what’s currently available that it’ll support computing possibilities that are only within the realm of imagination now.
Some mobile networks hesitate to make massive upgrades yet, especially since they can’t definitively calculate how 5G would impact the bottom line. Network upgrades allow operators to delay some 5G improvements, but many nations are getting close to running out of network capacity.
That inevitability should happen for many network operators between 2020 and 2025. Then, they’ll need to build new small cell towers or macro sites to accommodate 5G. Although 5G isn’t here yet, network professionals must plan now, along with mobile operators, and upgrade their infrastructures to avoid getting left behind.
Proactiveness avoids problems
Some network professionals are already facing issues with insufficient infrastructures, while others will soon follow unless they make decisive changes.
Understanding the network requirements that new technologies have, and then researching the options, are two practical early steps to take.