Subscribers are increasingly unhappy paying flat rates for unlimited mobile usage. Why?
First, “unlimited” is a more effective marketing pitch than “actual” usage. How many people are using more than 3 GB of network bandwidth per month? Some studies show that the average 3G subscriber only consumes around 800 MB per month. Even with 4G, studies predict average consumption of around 1 GB to 2 GB. Therefore, subscribers keep paying for more than they actually use.
Second, the ARPU is strongly variable by country. Up to $150 in some countries, even if quality of experience (QoE) does not always meet expectations.
If You Drive 50 Miles…
There is a high interest now in pay-per-use models similar to the automotive experience: “You drive 50 miles, you pay for gas for 50 miles.” In 4G networks, the mobile service providers (MSPs) will not be able to replicate the unlimited usage for a flat rate, mainly due to network congestion issues. So by proposing tiered pricing by subscriber type, MSPs could win on two fronts: higher customer satisfaction (lowering churn rate) and ability to upsell.
But beyond the mobile plans and service catalogs, operationally, it is technically challenging for mobile networks to accommodate these new requirements. If a large enough community of end users pays for premium usage with high data bandwidth, they have to get what they paid for.
Thanks to new virtualized architectures based on SDN and NFV, it will be possible to enable new, dynamic charging schemes. It will be possible to offer tiered pricing, application charging, and other freemium/premium plans which will be beneficial for MSP revenues. Networks could become service-aware and integrate application intelligence at strategic points. Service awareness could be at the core of the network, where the policy control and charging (3GPP PCC) could integrate DPI modules inside (virtual) networking equipment such as the PCEF, PDN gateway, or TDF.
Once the end users are able to buy their services in line with their expectations, the overall (next-generation) OSS can trigger the management and orchestration system to provision the correct policies and rules in each networking node. The (virtual) DPI node can monitor applications relevant for freemium, premium and customized plans.
By leveraging the information from the DPI function, the virtual PDN gateway will be able to either gate, steer, or prioritize specific portions of the IP traffic. The provisioning of virtual DPI and virtual PDN gateway nodes could be done using 3GPP standards or, if their forwarding planes are decoupled from their control planes, an OpenFlow-like protocol.
A DPI module could also be positioned in the (virtualized) infrastructure layer inside the hypervisor, which is a strategic point of consolidation between multiple (virtual) networking nodes providing both context information and security isolation. Such a DPI module can be integrated with a virtual switch to enable, for example, quality-of-service (QoS) per user in the flows/packets.
In conclusion, pressure from mobile subscribers, combined with virtualized network architectures could lead to a new mobility paradigm: pay for what you actually use!