Olympian offloading and creating a (profitable) “Games Lane”

Outside of London, the rest of the world will also be spectators, however through different means than Olympics past. Global mobile data traffic increased by 2.3 percent from 2010 to 2011, the fourth straight year in which data consumption has more than doubled, with video streaming being responsible for over half of 2011’s traffic. This explosion of data usage is a clear indicator that for the first time in Olympic history, millions of enthusiasts across the globe will be jamming networks from their portable devices with the expectation of being able to stream real-time data and video of their favorite events.

Unfortunately for Mobile Network Operators (MNOs), the Average Revenue Per User (ARPU) for data traffic does not correlate with the increased demand, prompting many in the communications industry to adopt an economics term, “scissor effect.” Further, with the plethora of opportunities for users to stream video over the next couple of weeks, MNOs can expect to be faced with a network overload comparable to those already being seen on the streets of London, necessitating a network “Games Lane” that can maintain connectivity and Quality of Service (QoS) while creating a platform for revenue-generating policy.

Developing the data “Games Lane”

Connectivity, QoS, and policy control, like most things today, are dictated by available technologies. From a traditional network standpoint, the way in which a network operator would handle increased bottlenecks is through the addition of elements to the core network, and correspondingly Drew Sproul of Adax commented that “in the first stages of policy control, policies were really designed to protect the network from bit torrents and peer-to-peer music sharing and downloads that were going on, and also to give people some notification to avoid bill shock when they were going over their plan.” Of course, this solution results in exponential growth of the network infrastructure and the corresponding maintenance and operating costs. More importantly, it does not provide a way for network operators to implement effective policies for managing data traffic and differentiating services.

The goal is to “get the junk off the network,” Sproul continued, “if you have the core wireless network dedicated to the functions that it needs to be in the middle of, then it can handle traffic better than if it has to be the conduit for traffic that is unmonitored, maybe even unmetered, to the Internet.” Ironically, the technology helping many MNOs with this bottleneck/monetization/policy creation conundrum is Wi-Fi, the same ugly duckling that was long dismissed by mobile operators due to negative revenue and security implications.  Wi-Fi has proven itself an essential cog in Mobile Data Offload (MDO), a solution that is being increasingly deployed by MNOs in traffic management and bottleneck relief. This is enabled by “a new generation of gateways that will authenticate you as a Verizon customer sitting at a Starbucks with Wi-Fi access to the network, and the gateways will take radius-based authentications, and then have a communication back to the LTE HSS (the home subscriber) to validate you as a Verizon customer,” said Sproul. “The corollary to monetizing the data pipe is prioritizing traffic that belongs on the macro network and offloading traffic that doesn’t.” This prioritization can be implemented through either packet intelligence on the network side or in devices themselves on the client end, in either a hardware or software solution.

For the judges

In policy today, network operators are hopeful that policy control servers are able to implement Policy Charging and Rules Functions (PCRFs) that can manage offline charging algorithms, as well as offer real-time, dynamic services that are billed at a premium. This type of policy requires information such as IP flow, device, service sought, QoS, and so on, which on the network side is based on packet intelligence. Until recently, however, “Ethernet ports and Ethernet drivers have just spewed packets into a system willy-nilly without order, so order has to be imposed on them,” according to Sproul.

To impose that order and create an efficient “Games Lane,” Deep Packet Inspection (DPI) device positioned at the access point to the core network that imposes policy control functions on packets being transferred and analyzes traffic patterns to decide what traffic can or cannot go to the Internet. Solutions such as the Adax PacketRunner (APR) ATCA carrier blade and PktAMC, which are powered by Cavium OCTEON processors, are two such high-performance components that act as a network doorman, if you will, being the first point of contact with Layer 2/3 packets and then routing them based on operator policy. “There are the analytic engines that sit above us, so we are in the service of the application to manage the IP flows, to manage the packets, to implement the rules that we are told to implement in terms of packet processing,” said Sproul of the role Adax plays in the network intelligence infrastructure.

For the athletes

On the client end, software solutions have been developed that move from the edge of the network to the core, utilizing a Mobile Network Director (MND) to create an “intelligent client” that can interpret service provider policy algorithms to redirect traffic before it reaches the network. These MNDs manage smartphones, notebooks, tablets, and the like autonomously in heterogeneous networks to provide devices access to the best network resources available, be they from Wi-Fi, the macro network, or otherwise, transitioning from connection to connection dynamically and in real-time. Carla Fitzgerald of Smith Micro Software explained that MNDs such as the [email protected], or blog back!