Some advances improve on what has gone before; others change everything going forward. Just as the smartphone wasn’t just a better mobile phone, 5G is not just an evolutionary advancement in mobile network speed. 5G will be a catalyst that creates new opportunities for communications service providers to deliver new applications that will change the way people live and work.
But in order to realize the promise of 5G, transformation of networks is critical to its delivery – and it’s happening now. Let’s look at the market shifts driving this network transformation revolution, and what technology and business innovations are required to deliver on the vision for 5G.
The 5G Landscape
5G will open up new spectrum at higher frequencies, increase bandwidth and transmission speed, reduce network latency, improve network reliability, and increase device density to better handle the onslaught of data brought on by increased mobility and the Internet of Things. The advent of 5G coincides with two other market shaping forces:
- The shift to programmable networks based on network functions virtualization (NFV) and software-defined networking (SDN) that deploy network functions on general-purpose compute servers.
- A new wave of applications that apply analytics and artificial intelligence to generate new business insights and societal value.
Fueling these forces is a projected 50 billion connected devices by 20201— devices like drones surveying pipelines and electrical distribution systems, sensors embedded in factories and transportation systems, driverless cars, and radio-frequency identification probes in retail inventory and distribution systems. 5G, built on a foundation of NFV and SDN, enables the agility and scale that will create an unprecedented opportunity for application innovation.
Ushering in these diverse new applications requires us to deliver network “slices” with capabilities tuned to the needs of each workload type, overlaid on a common infrastructure. An end-to-end solution, from the network and cloud to the device, must provide a few key features, including:
Ultra-reliability and low latency – for mission-critical applications such as vehicles communicating with each other and with traffic control systems for collision avoidance, remote surgery that can save lives in locations that lack specialized medical skills, and operation of industrial equipment in remote or dangerous locations. These and similar applications will require network latency of only a few milliseconds.
Massive machine-to-machine communications – for aggregating data from the billions of internet-connected devices in smart cities, smart factories, and smart homes that will place an enormous demand on network capacity. A single smart factory, for example, may generate as much as a million gigabytes of data per day.
Enhanced mobile broadband – to support faster speeds and greater capacity that enables immersive virtual and augmented reality experiences that can transport users to locales – ranging from mountain tops to ocean depths – all from the comfort of their living rooms.
What Needs to be Accomplished Today?
Providing a network that meets these new demands requires more than just running 5G wireless technology on today’s networks. It requires a complete transformation that adds intelligence and programmability throughout the new network.
The good news is that the process to a 5G ready network has already begun, and is giving operators a jump start on their transition to 5G, including:
- Dynamic use of network resources enabled by NFV/SDN.
- Enhanced service delivery at the network edge with multi-access edge computing (MEC).
- “Slicing” network resources by industry vertical or by horizontal segment use case.
By implementing virtualized network functions rather than deploying dedicated devices for specific needs like edge routing, firewalls and gateways, and by assigning servers and storage to sharable resource pools, communication service providers can compose more efficient networks in real time. Additionally, they can offer services to meet diverse application needs. This effort can also be accelerated to reduce costs by leveraging open standards and the open source community through initiatives such as ONAP (Open Network Platform Automation Platform), OPNFV (Open Platform for Network Function Virtualization), and DPDK (Data Plane Development Kit). Leaders in the industry are already reporting infrastructures that are more than 50 percent virtualized, and realizing the benefits of lower total cost of operations and faster service innovation and deployment.
In addition, communications service providers can use MEC on LTE immediately to deliver better user experiences to consumers – be they people or “things.” By moving the computing capability closer to the network edge, communications service providers can reduce latency and increase overall capacity for video and applications that are frequently used and delivered. By optimizing network slices for specific industries and use cases, enterprises can offer unique and customized services based on quality of service (QoS) traffic type, network congestion, user preferences, and security requirements, while fine tuning their network resources to meet their business requirements.
Networks must also become smarter. Adding artificial intelligence on a programmable network infrastructure will enable us to turn large amounts of data into valuable insights that spur new applications and services. Developers will be able to innovate new solutions that can help cure diseases, extend education, prevent crime, and enhance safety on our roads. Communication service providers can enhance their networks and build in flexibility and intelligence by implementing machine and deep learning, which through the collection and analysis of data, can help us spot trends that could be preemptive, protective, or recurring.
An example of more intelligence at the network edge today is the use of machine learning to determine how best to maximize signal strength from wireless base stations to connected devices. By collecting data on topography and traffic patterns, communications service providers are learning how to position their smart antennae at various times of day for the strongest connections. Just imagine how much more can be done with a finely tuned network that only uses the optimal resources and capacity required for each user and use case – no more, no less.
A new world awaits us. The time for network transformation is now. While 5G standards and technology are still evolving, the foundational capabilities offered by NFV and SDN are available today. And the work we’re doing as an industry today with MEC, network slicing, and the open source community will be buoyed by the expanded radio spectrum, connectivity, bandwidth, and lower latency of 5G as it rolls out.
Let the innovation begin!
1) IDC, IoT Market Forecast: Worldwide IoT Predictions, 2015
Sandra Rivera is Senior Vice President and General Manager of Intel Corporation’s Network Platforms Group.