Evolution of Mobile Data Plans & the Critical Role of Network Intelligence
The sale of smartphones and tablets continues unabated with an ever increasing range of solutions. The push for more content, especially video, is a large contributor to this growth with a specific focus on HD solutions supported by larger screens. HD solutions are a key focus area especially amongst the Android based vendors, but many players are also focusing on new areas of specialization such as gaming.
The latest Cisco Visual Network Index (VNI) report once again predicts incredible growth in mobile video….
- Mobile data growing by 26x between 2010 and 2015
- Video overtakes peer-to-peer(P2P) as highest demand application in 2010
- Video is 40% of consumer related IP traffic
Clearly, data rates are going to continue to grow dramatically, with a wider range of applications creating network demand. The buildout of the LTE network will add significant capacity to help deal with the growth, but at 26x growth over the next five years, there will need to be more to the solution than simply bigger pipes.
Additionally, there are some specific observations and conclusions that can be drawn about the mobile market based on the VNI report and many other industry sources.
First, the mobile network operators were right to invest in the Policy and Charging Enforcement Function (PCEF) in order to protect the network from congestion and hogging – protection being the first objective for Deep Packet Inspection (DPI) in the network. However it's not just the P2P "hogs" that are going to be the problem. The aggregate consumption of the average smartphone user is also going to be a big concern going forward.
Second, it is evident that video is the killer application. However, video can come from many sources and, therefore, it is as important to understand where this video is coming from as it is to understand how it can be recognized, controlled and delivered in a more suitable format for the mobile network.
The means for accessing content is also changing dramatically for mobile. For the average smartphone user, access to services such as Facebook and YouTube is now via an application rather than through the Web browser. This changes how providers monitor usage but it also delivers new options regarding service levels. Rather than having better "browser" performance, you can now more easily tailor for improved app-specific service, in conjunction with the app being downloaded. For example, consumers could select between downloading a free app or a premium version with guaranteed bandwidth and video processing resulting in better video quality.
Today, virtually every mobile operator has a level of PCEF capability within its networks, and each typically has a contract that talks about data caps, excessive usage and fair share policies. However, at a practical level, many do not take full advantage of the capabilities and some do not even enforce these limits aside from exceptional circumstances or when network congestion occurs.
This brings us to the question of tiered pricing and deciding whether unlimited data plans are detrimental and no longer viable. As a quick digression, it's interesting to see that some iPhone data plans are now capped, rather than "all you can eat." However, when you dig into the details, it is in fact "downloads" that are capped whereas Internet browsing is not limited and no additional charge is incurred if you exceed the limits. So it begs the question, is video part of the reason for this policy?
All-you-can-eat data plans are manageable if operators have the right business models, reporting tools and customer communication models in place. What operators must ensure is that the customer is aware of what happens and what's being charged and that they don't feel they are being unfairly limited.
As an example, a user on the lowest tariff may be using far more than their fair share given the price they are paying, and might then be notified of their excessive use and invited to move up a tier. If they continue to use more than their allowance, the operator has the option to continue warning, or, at some stage, to take action by rate limiting or even automatically moving the user up to the next tier, with a corresponding increase in their monthly charge. This is the model some "free" fixed-line broadband vendors apply to their bundles where the internet access comes for free with an entertainment package.
The second aspect of charging is regarding tiered pricing. Originally, this was held up as a way to increase Average Revenue per User (ARPU) and start to close the gap between CAPEX investment and data growth. Today, clearly this is going to be a drop in the ocean given the rate of data growth, but it does have a place in an evolving data pricing model. The key questions are, how much is a user prepared to pay for a given improvement, can this "improvement" be consistently delivered and which applications could this realistically be applied to? Video and gaming are good examples where this might apply. Access to social networks, for example, may not be as suitable given the lower data rates, unless it comes with some other bonus (e.g. Wi-Fi hot spots or free access to the app when roaming). This is similar to aircraft based Wi-Fi promotions giving free access to Facebook.
The biggest challenge with tiered servicing is the variability on data rates in a mobile network. Currently, most systems cannot dynamically adjust data rates and are hampered by the static nature of the PCEF when facing individual loading of cells and radio conditions. The latest PCEF solutions can have incredible levels of granularity per user and per application, but the air interface continues to be a challenge and therefore the PCEF may often end up being implemented as somewhat of a brute force tool. Until the networks can identify individual cell loading in real time, both tiered and guaranteed services for a given application will carry risk.
In summary, PCEF and DPI for mobile is, and will continue to be, a critical technology. It has the possibility to provide highly-tailored solutions with high levels of personalization – and with this comes new business models and revenue generating possibilities. However, the solutions need to continue evolving to better support the specific use cases in the mobile network and especially deal with the variability of the air interface. All-you-can-eat data plans will clearly change. They open the possibilities to generate more revenue without causing customer objections, but the policies and approach must be carefully considered to avoid customer disappointment and resulting churn.
Karl Wale is director of Product Line Management at RadiSys.