By Jim Poole
Mobile networks have a problem and opportunity rolled into one.
On the one hand they are well positioned to profit from the explosive growth of mobile data: Everything Everywhere’s Peter O’Donnell suggested at Carrier Ethernet World Congress that “1/3 of UK adults use their smartphone as their primary means to access the internet.”
On the other hand, the gap between revenue and traffic is widening: that’s to say that while both are rising on a chart, traffic is rising much faster than revenue.
The infrastructure problem faced by mobile network operators is that they have invested heavily in their mobile networks, vastly reducing latencies and increasing throughput in the last decade.
For example in the days of GPRS, an acceptable round trip latency was considered to be 800ms and under. Now as the industry moves towards LTE, that figure has been cut to below 100ms. What this has brought to light is the need for quality in the underlying transport network.
In the GPRS days of 800ms round trips, the transport and backhaul of data was a minimal part of the customer experience. As content demands ever-more bandwidth and the latest wireless networks are up to speed, a poor underlying infrastructure will translate to unacceptable performance for the customer, or in O’Donnell’s words “the transport network must not be allowed to throttle the expensive radio networks.”
To encapsulate the problem: to deliver customer satisfaction, mobile networks need to invest yet more in their infrastructure, despite still smarting from the heavy investment in wavelength rights and 3G, and at a time when revenues are under pressure.
“Backhaul needs to be scaled in advance of customer demand,” says O’Donnell, “If we wait to be driven by customer demand, it’s too late.” He explained that the ideal backhaul infrastructure solution was one with a relatively small topology, that still offers proximity to the web’s major destination and content delivery networks.
Clearly, this man was at a carrier Ethernet conference for a good reason, because a swift move to Ethernet will offer mobile network operators the opportunity to implement the kind of performance and scalability they will need to retain their customers. Surprisingly many mobile networks still rely in part on TDM-based backhaul.
Still if Ethernet is the medium, how can we ensure the infrastructure’s topology remains as small and efficient as possible while still being resilient enough to guarantee a good service for mobile customers? By routing into a neutral data center with an established ecosystem of Ethernet Carriers, web giants and Content Delivery Networks.
Whether a mobile operator simply colocates in a content hub and purchases their own direct connections to the content providers they need, or if it accesses an ecosystem of content and network providers through an internet or Ethernet exchange model – the neutral data center is clearly an option that mobile networks should be considering very carefully as they strategically plan their next phase of transport networks to support the rapid growth in mobile data.