At the Edge of Something Big: How Edge Based Analytics is Preparing the Path for 5G and IoT.
At certain periods of time we see technology-driven inflection points, and most would say we are in the midst of one today. But as in the world of fashion – technology can also be a place where everything old is new again, and these points of inflection, while certainly game changing – can also create a sense of deja vu.
Back to the Edge
In the early days of computing, in the age of main frames, we relied upon a centralized model. Then as popularity grew and demand increased, we decided this was too slow and confining, so we moved to a client-server based model that utilized PCs, where storage and computing was done locally at the edge, which is physically close to the user. But once again, demand continued to increase and we realized we could gain economies of scale and better collaborate if we worked from a shared cloud, which brought about the need for NFV and SDN. The ‘cloud’ has now brought us back full circle, to a more centralized model. But just like the never-ending fashion cycle of wide ties and skinny ties, it’s once again time for a shift back to the future, thanks to the Internet of Things and 5G. We see 5G powering the IoT by connecting massive amounts of devices and sensors, while enabling super low latency applications and services and significantly higher data rates.
The IoT will enable billions – maybe even trillions – of newly connected things. For networks to manage all these new connections, we’re quickly realizing that the centralized cloud model – even at 4G and 5G speeds – won’t work in the long run, or at least not optimally. This is because having sensors or other devices gather and send massive amounts of data back to a central cloud for analysis and processing, no matter what the speeds, won’t be fast enough for many of the new applications and services in our very near future.
Service providers are working hard to prepare their networks and business operations so that they are ready to offer some of the new services this will enable; changing their legacy connectivity provider mindset to focus on new digital offerings across various verticals. For instance, specific services related to connected cars, augmented reality, eHealth services and many more we can’t even conceive of yet, will be enabled thanks to the extreme low latency and high bandwidth capabilities of 5G. This CSP business model shift is happening for good reason. Revenue from traditional services is declining. SMS is just one example, where $54 billion in revenue was lost in 2016 to OTT applications like WhatsApp.
The Network of Tomorrow
Complexity will continue to define the network of tomorrow. The new reality for the foreseeable future is a dynamic, hybrid network comprised of physical, virtual and traditional infrastructures. This new complexity will leave operators struggling to support this evolving environment, and the evolution will surely continue for many years to come. We predict that other connectivity channels will eventually enter into the mix. Wireless networks will no longer just be managing wireless data. They will eventually need to become access agnostic, able to incorporate data from other protocols as well. Wireline, Bluetooth, Zigbee, NFC, Wi-Fi, and others will all be gathering and sharing data.
With huge amounts of data coming from all of these devices, this means many tasks will be better managed by separate clouds located at the edge; once again, creating a more distributed model, but one that is still inter-connected. We won’t be abandoning the central cloud model entirely. This will still be relevant for core network functions, working as a central ‘brain,’ processing huge amounts of data for further analysis. This can then be augmented through machine learning and artificial intelligence, and then fed out to the end devices, for services that don’t demand super low latency response rates.
This new network model will require real-time connectivity and orchestration and a set of decision based analytics tools that think of it as one network, although a highly complex one. Tomorrow’s networks will become a cloud of clouds. Incredibly complex, hybrid environments that are automated, agile and constantly learning, growing and re-shaping to deliver instant, reliable connectivity to just about everything, everywhere.
How TEOCO is helping
At TEOCO, we are focused on building the networks of tomorrow. Today we work with over 300 service providers around the globe to support their growth initiatives in areas like NFV, Network Automation, Edge-based Analytics, 5G and IoT. As founding members of the 5G Innovation Center at the University of Surrey, we have our RAN and Service Assurance solutions deployed in a 5G test environment. Our automated network design tool lets operators perform early modelling of 5G networks, and our edge-based analytics tools are in operation today at some of the world’s largest service providers, doing things like improving and ensuring quality for VoIP services, and providing insight into how subscribers are using and interacting with OTT applications, so CSPs can better compete.
We constantly strive to stay ahead of the technology curve, building solutions that will help our customers be more successful. Deciding which tie to wear – skinny or wide? Well, we’ll leave the fashion decisions up to you.