Layer 7 Application Filtering and Traffic Shaping
Layer 7 Traffic Shaping delivers one of the highest returns on investment of any optimization technique we implement. The concept is relatively simple:
- Identify your traffic
- Prioritize or shape the traffic so as to improve the subscriber experience.
In addition to the benefit of improving the experience of the user, the first step which is the least expensive part of the whole process is invaluable, and that is the enhanced network visibility so you really know what is going on with your users. This invaluable tool gives us the analytics that will enable us to make better decisions on how to allocate, and grow network resources.
Enhance Network Visibility
Before we can shape the traffic, we have to know what it is and thus we start with a visibility solution that gives us an accurate picture of real-time network usage providing valuable insight into usage trends. This increased visibility adds proactive troubleshooting capabilities, traffic and trend analysis. Real-time and long term reporting tools provide multidimensional views of application traffic, subscriber traffic, value-added service traffic and HTTP traffic.
Improve the user experience
With 43% of application traffic on a network now comprising of video, this represents a significant opportunity to differentiate your network, and improve your subscribers experience. Hence we use the visibility to shape or prioritize the traffic that matters the most.
We shape the traffic by applying QOS rules to the application layer. So for example referring to the pie chart above, we would want to give our Voice the highest priority, then optimize for the video streaming, next the http traffic and then finally the file sharing. While this does not change the bandwidth at all it gives priority to the things that our clients judge us by and can "feel" and reduces churn since in this particular case, the Video QOE that we have introduced becomes a "visible" quality advantage and provides a new benchmark to the overall network quality.
Understand your customers, and the different users on your network
Based on the insights that we gain from our analytics, we will discover that there are various types of users, and we can categorize them in to five groups. For now we will focus on the diferentiation of two particular segments, what we call the "info seeker" and the "digital mover and shaker".
The Info Seeker represents 32% of our subscriber base but only 12% of our overall traffic. Their usage is characterized by browsing, and that is 75% of their traffic. The Digital Mover and Shaker on the other hand represents 34% of our subscriber base but accounts for 58% of our overall traffic. The Digital Mover and Shaker is very active socially and uploads and shares information content that interests them. They stream videos, instant message, watch social video and browse. Most importantly they are significant not just because of the heavy traffic that they consume but also mainly because the digital mover and shaker's usage on the network spawns even more usage across the network.
Based on the different types of usage, next we will develop plans that cater to both of them and fairly allocates and charges them for the resources they consume on the network. Otherwise, we could get saddled with all Digital Movers and Shakers while our competition's subscriber base is all info seekers, and as a proportion, their users use much less bandwidth meaning that our competitors have a lower cost of operation.
Differentiate Service Offering to increase ARPU
The next logical step is to then tier our services and offer different plans catering to the different users. By implementing these users into logical groups on our appliance, we can then apply QOS and traffic shaping rules as policy and control elements that provide the real differentiation for the tiered services and motivate the customers to upgrade to premium services.
Cost effective to implement, scale investment with proven ROI
The first aspect of implementation on this type of platform is typically very inexpensive since it is just the appliance and all we are going to use it for is gathering information on our traffic. The larger portion of the expense comes later with some additional software and a significant effort of engineering time that you will undertake to fine tune your policies.
What we have described here is just the tip of the iceberg with what we can do with this type of technology, and if there is any justification for it, the largest network operators, including the largest and fastest satellite network, in the world are all using this, so you should be too. If for nothing else other than the analytics, by implementing just the appliance initially, for a very low cost typically as much as 1 MHz of space segment, network operators gain a visibility into their network that will enable them to make more strategic decisions with regards to optimization of their resources.