Over the last year, industry research reports have examined the issues related to managing the growing volume of Internet traffic driven by advanced applications, particularly streaming applications like video.
Internet traffic rates worldwide are greatly expanding – between 50 and 60 percent annually – which would require service providers to maintain or increase their current levels of capital investment, while their “all you can eat” broadband service plans continue to provide them flat ARPU. In addition, growth of latency-sensitive traffic (especially video) raises questions whether a satisfactory level of end-user experience can be maintained.
To examine the issues faced by service providers as they look to deliver massive traffic volumes while also maintaining high service quality, we turned to Bill Mello, director of marketing for the Americas at Allot (News - Alert) Communications.
As Internet traffic continues to explode, what’s the top challenge for service providers?
The challenge is how best to deliver excellent service – and turn new applications into money – by optimizing the performance of existing network infrastructure. The easy – and wrong – answer is constantly invest in new bandwidth. Simply building bigger pipes isn’t efficient. Capacity would have to be designed to support peak traffic levels, leaving expensive bandwidth idle at off-peak times. A better solution is to dynamically allocate bandwidth when and where it’s needed.
For latency sensitive applications like online gaming, VoIP, video conferencing and IP video, a split-second disruption in the traffic flow can severely degrade service quality. And their growing popularity is increasing customer demand for reliable, high-quality broadband service.
What new demands do these applications bring to the network?
Take an application such as IP video, for example. According to industry researchers, one hour of high-definition video delivery requires 2.5 gigabytes of network traffic. Compare that to one hour of music delivery which only takes 0.1 gigabytes.
As video content owners distribute more traffic over IP, service providers must carefully manage bandwidth consumption through policies and customer quotas, and ensure high-quality performance, including speedy downloads.
How can deep packet inspection (DPI) address this challenge?
Deep packet inspection works by identifying traffic based on users and applications. It enables service providers to establish and maintain policies to prioritize specific types of traffic and users. For example, DPI can prioritize latency-sensitive applications over those that are not – such as Web surfing and e-mail.
What this means is as service providers run up against network congestion, DPI can alleviate the pressure of increased demand by enabling policies that for example, set maximum bandwidth quotas for subscribers during peak usage hours. Subscribers requiring additional bandwidth (for frequent video downloads, for instance) would pay more for their bandwidth consumption. The result is greater revenues and avoidance of unnecessary network build-outs for service providers, combined with higher service quality for subscribers.
Can you provide an example of how this works?
Let’s look at an ISP named Wisper High Speed Internet which leverages DPI to prioritize VoIP and gaming applications, while de-prioritizing traffic like email delivery and P2P, which can tolerate greater latency with minimal impact on the user experience. By utilizing DPI, Wisper keeps time-sensitive traffic flowing through the network first, while not blocking or disallowing access to any services subscribers may wish to experience.
How else can DPI help service providers improve user experience?
Service providers can combine DPI with other network services so application and subscriber information for each traffic flow can be identified by a single DPI process and then dispatched to relevant devices for appropriate actions.
This approach minimizes redundant packet analyses, resulting in improved overall network performance, reduced complexity and cost of new service deployment, and better subscriber experience. It also provides greater revenue opportunities for service providers who can offer a range of value-added services on their own networks.
Some examples include:
-- Video caching – By combining DPI and caching, service providers can efficiently identify video traffic on networks and dispatch it to a caching service which stores frequently downloaded videos locally.
-- VoIP optimization – DPI working in conjunction with VoIP optimization solutions enable service providers to combine consistently high quality VoIP delivery with accurate billing.
-- Security – Together, integrated DPI and security solutions help service providers identify, isolate and remedy subscribers infected by spam-generating botnets, and prevent network-level Distributed Denial of Service (DDos) attacks.
-- URL filtering – New DPI engines can identify and block attempts to access illegal Web sites, including inadvertent attempts on the part of unsuspecting users.
TMCnet publishes expert commentary on various telecommunications, IT, call center, CRM and other technology-related topics. Are you an expert in one of these fields, and interested in having your perspective published on a site that gets several million unique visitors each month? Get in touch.
Edited by Michelle Robart