Exploding demand for high-bandwidth content such as video streaming and mobile apps is threatening to saturate networks and negatively impact performance. What’s more, the next generation of gaming, virtual-reality and low-latency 5G mobile applications will exacerbate the problem, making it even harder to deliver the flawless QoS that today’s users expect.
Because of these trends, forward-thinking content providers are beginning to move the network edge closer to where customers are. In this network model, connectivity, processing power and storage for high-bandwidth and latency sensitive apps is handled before traffic hits your core network, helping you deliver truly great user experiences and freeing connections for less time-sensitive apps and use cases.
What, exactly, is the new network edge?
It’s likely that the network edge will look slightly different for every provider, depending on your specific priorities and business models.
For some, a small to mid-sized data-centre close to an end-user population will make sense, providing an adequate balance of service performance and deployment and operating efficiency. For those who specialize in premium, low-latency services, a more distributed approach will work better, possibly using a larger number of edge “stacks” located at distributed sites very close to subscriber populations.
Building business cases for edge deployments
It’s early days for edge networking deployments and the industry is still in the process of developing concrete business cases that accurately predict ROI. However, mirroring Netflix’s successful “build as you grow” approach, it’s likely that providers will deploy edge infrastructure close to large population centers first, where demand for certain services promises a rapid return on their investments. Once business cases can be established based on past success stories and deployments in similar towns and cities, it may be possible to adopt a more ambitious “build it and they will come” approach to edge rollouts.
Potential roadblocks for edge networking strategies
Putting network equipment closer to customers is great in theory, but there are some practical considerations to consider. Of these, lack of physical real estate close to subscriber populations is a significant obstacle to overcome. After all, the edge networking model requires tens, hundreds or even possibly thousands of new sites to be deployed and with only limited real estate available, this could significantly delay content providers’ edge strategies.
The other possible limitation for edge projects is the requirement for high-bandwidth network connections at every proposed edge site. Additional requirements include adequate power and cooling provision, along with physical site security.
Overcoming the challenges in partnership
It’s likely that building a successful edge-networking strategy will require support from multiple technology partners with different resources and skills sets. Partners may potentially include providers who already have sites close to subscriber populations; network suppliers that offer edge technology leadership; data centre companies with knowledge of AWS, Google (News - Alert) and other key data centre apps; and systems integrators with the skills to deploy and test multi-vendor networks and manage complex projects from end-to-end.
Creating new edge opportunities with street-hardened equipment
In the near future, we expect to see organized consortia of edge deployment partners who are able to deliver turnkey deployment projects for content provider customers with a single point of contact and responsibility. This model provides major commercial opportunities for content providers, network vendors, service providers, SIs, consultants, and others, as well as minimizing project risks.
In addition to partnering with those who already have sites close to customers, content providers may include building their own micro data centers in street cabinets. This approach can reduce pressure on existing data centre space, while also cutting red tape for planning, and for increasing power allocations.
Edited by Maurice Nagle