taking policy to the
smart connected device

Archive for October, 2012

Network Operator and Cloud Provider Collaboration – It’s Great, but Needs End-to-End QoS Enforcement

Written by GoS on . Posted in Blog

An interesting news item caught our attention recently. It seems AT&T is going to collaborate with IBM for the delivery of cloud services. AT&T’s network expertise will be complemented by IBM’s expertise in cloud services. In their analysis, Yankee Group suggests that this points to growing maturity in the cloud market.

We agree – but it also points to further necessary steps towards cloud service delivery. We have long viewed the role of the network operator as being potentially crucial in cloud service delivery – they are able to control the connection to the customer and can help deliver the quality of service and performance guarantees that will be demanded by enterprise cloud customers, but which the cloud service provider cannot offer.

But it needs to go further – network operators cannot currently guarantee QoS in the access domain. This is particularly problematic in mobile networks, where RAN performance is a hotly debated topic. What’s needed is the ability to deliver and guarantee QoS from the devices that are used – tablets, dongle enabled laptops, mobiles and so on. This can only be achieved by extending the policy management framework and deploying PCEF functionality on devices.

By taking these steps, network operators can ensure that they offer clear and unique value to cloud service providers, forging the way for fruitful partnerships in the future. We expect there to be many more such announcements – but we also expect the issue of end-to-end QoS to become even more important as a consequence.

 

Angry Birds Raise up a Storm!

Written by GoS on . Posted in Blog

A recent blog by industry experts Core Analysis raised a very interesting topic. It’s well known that increased data consumption by mobile users has also increased the amount of signalling traffic within networks. This is believed to have had a profound impact on network load and, consequently, on the need to invest in RAN capacity.

The blog highlights one operator’s experience with one of the most popular smartphone games of recent years, Angry Birds. Apparently, Telecom Italia experienced significant difficulties with signalling congestion after the launch of Angry Birds on the Android OS. This was in contrast to the experience with users of the same product on iOS.

It turned out that the main issue and hence the cause of the signalling congestion arose from differences in how the applications are monetised. iOS versions are charged for, whereas Android versions are monetised through in-application adverts. The application automatically requests (and obtains) adverts and generates significant volumes of signalling traffic. The volumes had been sufficient to cause network outages.

Angry Birds is not the only such application to generate significant volumes of signalling noise. What’s more, even when users are not accessing games like Angry Birds, the applications can continue to make requests when they are supposedly dormant and in the background. The more such applications the user has open, the worse the effect. The solution proposed is to deploy routing solutions in the network.

Actually, it’s a much better idea to address the issue in the RAN itself, before the traffic hits the network and before it starts to create a problem. It is exactly these conditions and thinking that have driven a significant deployment for GoS Networks with a leading Tier 1 mobile network operator.

The operator in question has a significant installed base of Android users, which is growing rapidly. By deploying the GoS solution on mobile devices, they have been able to extend policy control into the RAN and directly to the user device. By doing so, they can moderate the frequency with which background applications generate signalling traffic in the RAN and before the resulting traffic reaches the core.

This has had the result of alleviating significant pressure on their network. The operator has calculated that they will save or defer up to $40 million in CAPEX costs over a three-year period simply by reducing this signalling overhead.

That’s a huge saving and one that is open to any operator experiencing similar traffic issues. By leveraging our solution in conjunction with their PCRF, operators can generate significant cost savings and dramatically increase user satisfaction. Why not find out how you can address this problem?