Interest in policy control in the past year has grown dramatically, to the point where it has become one of the hottest subjects for communications service providers (CSPs). Driving this is the fact that many CSPs are experiencing a growth in data traffic that is far outstripping the growth in their revenue, and affecting the performance of their network. This is a trend that is clearly unsustainable, so CSPs are looking for ways to redress the situation. They not only want to make better use of available resources, but also offer more personalized services to their customers, with the hope of generating more revenue.
Better use of resources
Policy control is seen as the means to achieve these goals. It has a multitude of definitions, objectives and solutions, but broadly speaking, policy control is about managing the access to services by subscribers.
Historically, policy control can be thought of as having two very separate origins. In the fixed environment, policy control was about network resources allocation. An early example of policy control in action is the use of class of services (CoS) in MPLS networks to differentiate the delivery of enterprise services. In the mobile space, policy control was about charging - for example, taking action when a prepaid customer's credit runs out.
Now, with the advent of the mobile broadband era, policy control is stepping out of those silos and coming of age. Most importantly, policy control is shifting from a network-centric and defensive approach to one that puts the customer experience first.
To achieve this, policy control has become far more dynamic, taking a multitude of factors into account - in real time. These factors include not only the type of service but also the current network conditions, the user's profile (e.g. business or consumer, gold or standard, high or low spender), the type of device being used to access the service and even the location of the user.
A good illustration of a flexible application that ensures a good customer experience is bandwidth management.
The initial problem can still be seen as a classic network-centric one: CSPs want to ensure that the bandwidth available to users does not become squeezed as a result of excessive use by a minority of subscribers who do not contribute proportionately to revenues, such as heavy peer-to-peer download users. When that situation occurs, the majority of users experience a reduced quality of experience (QoE), which may lead to churn. This problem is most acute in mobile networks, where bandwidth is clearly limited.
State-of-the-art bandwidth management solutions allow CSPs to monitor usage in real time and, when congestions occur, to dynamically adjust the access for specific services and specific users (at a cell level for mobile operators) to free up capacity. Such a solution is not just about defending the network - it's about providing the optimum broadband experience for the majority of users.