A new survey from Rethink Technology Research, commissioned by Amdocs, finds that 94 percent of service providers anticipate 2000 to 5000 percent increases in mobile data network capacity will be required in the next five years to meet market demand. The survey also finds that, in order to expand capacity, 59 percent of providers plan to deploy 1000 percent more small cells in 2017 than they did last year. Further, 88 percent of respondents plan to offer WiFi as part of their mobile services by 2016 and plan to have WiFi interfaces integrated into mobile cell sites increasingly in the coming years. All of this extra gear in the network likely means increased demand for edge-based intelligence so that the network is “smart enough to understand what you’ve got;…the service you want; the capacity on the box at any given time; and your tariff plan as well,” says David Chambers, solutions manager- mobile, for Amdocs.
This isn’t a simple matter of overprovisioning to stay ahead of demand, as we saw in the early 2000s with massive, raw expansion of landline fiber networks. In this case, the network needs to be smart enough to apply the right resources to each service, customer, and situation in real-time. This kind of capability has been discussed, if not anticipated, for a number of years in the OSS/BSS sector. Says Chambers: “Some of the things people have been forecasting for some time, and not necessarily believing, may become true in [the next 5 years].”
Chambers explains that “we’ll have a box with all three (3G, LTE, and WiFi) interfaces with authentication done so that you don’t have to [reauthenticate] every time you change radio interfaces.” With that array of options, the network needs to understand the best way to treat each session, or type of traffic within a session, as well as each customer and what level of quality they receive based on their subscription or SLA. To make this heterogeneous, hybrid network model work optimally, “we’ll expect to see a bit more intelligence” pushed to the edge of the network as a “significant architectural emphasis in the future,” says Chambers. Today, the mediation, intelligence, or policy systems that make those kinds of decisions, albeit in much simpler ways, tend to reside “in a central box at the back end,” explains Chambers.
Smarter, Gentler Networks
If this type of intelligence is applied at the edge of the network, it also begins to empower many more sophisticated ways to interact with customers. Chambers notes that the somewhat rudimentary models we see today will be able to evolve to become more refined, or simply less draconian. He cites the example where Smartphone users tend to be more profitable than “data dongle users” because they often don’t consume their entire data allowance, whereas data dongle users consume their allotments “until they run out.” He says one European operator’s approach to preventing network abuse is simply to put a “daily 500 megabyte limit on dongle users regardless of what plan they’re on.” That sort of network control can be “better refined” as the industry moves forward, which in turn means “better use of policy management.”
This kind of scenario points to the convergence of capabilities we see in development today. For example, the idea of the “bandwidth boost” – as demonstrated in TM Forum’s Real-time Charging for High Volume Mobile Data catalyst – can be applied here to turn a draconian data limit into an opportunity for customer outreach. Instead of shutting down the user, an operator could reach out with a real time message that says, “you’ve reached your limit; we’ll need to throttle you back a bit; but you can also boost your plan temporarily for an additional $10.” That puts the user in control while creating an add-on revenue opportunity for the operator.
Similarly, increased intelligence at the network edge can provide operators with advantages over OTT competitors. As the industry tries to move away from selling infrastructure-based products only, like a 2 GB plan, and toward application-based services, like premium digital video, intelligence plays a role. An operator can offer unlimited video, served via WiFi whenever possible, as a premium service that doesn’t draw on the customer’s data limit. A third party OTT service, however, would draw on that limit, giving the operator the chance to offer essentially the same video service, but with a significant value add based on its ownership of the network and the mobile subscription.
Chambers takes this scenario a step further. He suggests that “in the future, we might see a platform that…gives consumers choices passed through from third parties,” as an extension of the “right offer, right time” capabilities real-time analytics and policy management can enable at the network edge. Companies like Amazon, who need and want to become increasingly mobile, are likely to require these capabilities in order to continue to enhance their personalized recommendations within the mobile environment.
In short, if operators really want to move away from being “dumb pipes” then they’ll need make sure that as they add new mobile data pipes they don’t fail to install their brains.