Rein in heavy usage and implement network fairness to ensure the best QoE for the largest number of subscribers, while also deferring congestion CAPEX and building new revenue streams
If you’re a NOC/SOC manager or engineer, you’ve got a 24/7 responsibility in managing, scaling, optimizing, and securing networks. Your work not only ensures networks perform well, but also the quality of experience (QoE) for the applications through which services are increasingly consumed.
As we wrote in our ‘Gaming QoE’ blog, users don’t really care about the “G” underlying their favorite apps and services, as long as the subscriber experience is good:
- Did the videoconferencing and collaboration apps help me accomplish what I set out to do?
- Was the audio and video quality good while I binged my favorite show?
- Did my characters immediately respond and perform the intended actions in my favorite game?
In wireless, 4G/5G, cable, fixed, and satellite, a positive user experience is inextricably tied to growing network congestion and unfair usage on networks. That is even more true as apps become more complex, mashed up, multiplexed, and demanding on network resources. The subsequent explosion of traffic means a less-than-equitable result, with heavy usage from a small contingent of users disproportionately affecting network quality and application QoE for the greater whole of customers.
It’s a perversion of the well-known Pareto principle, better known as the “80:20 rule,” with as little as 5% of subscribers consuming as much as 35-40% of network resources, and about 8% of total subscribers generating almost one-quarter of total traffic!
The culprits are usually P2P services like BitTorrent, streaming services like Netflix, YouTube, and Hulu, and games like Destiny, and Counterstrike. Their impact is particularly detrimental to the performance boost people need for smarter, more sophisticated applications, and fast-evolving metaverse applications. And they also complicate resource planning (driving up average resource usage for everyone), service profitability, service running cost, ROI, and brand perception (which in competitive markets is everything).
Traditional methods of conquering heavy-usage problems rarely work: profiling and capping heavy users; capping applications; quota tiering; unlimited plans; and throwing more money into RAN investments. All are tedious and time consuming, delivering suboptimal results, and leave you with underutilized spectrum and cell capacity.
That’s why old-school heavy-usage management has to give way to modern, multidimensional approaches that bring about “network fairness,” equalizing the use of network resources so that overall application QoE improves while managing congestion in the network.
The goal is better performance for the largest number of customers, while also deferring the CAPEX and OPEX associated with congestion management and network optimization.
Five Steps for Managing Heavy Usage on Next-Gen Networks
A phased, multidimensional approach can be broken into five steps:
- Examine closely subscriber usage and application bandwidth behaviors to distinguish users on a spectrum of “light,” “average,” “power,” and “heavy.” Decipher which users require no management, fair-share management, or “target” management.
- Distinguish between different types of heavy users – i.e., P2P seeder/leechers that use P2P traffic at full speed for upload, download or both consistently; pseudo service providers reselling services to multiple customers without consent; hosting servers with high upload traffic for HTTP, HTTPS and other TCP ports; Malware zombies. Determine a target GB/user threshold (per week or per month), and manage those who exceed the targets.
- Implement granular policies that measure and control usage (hourly, daily) for excessive users, as well as control bandwidth-intensive applications to ensure network fairness for each user (by preventing congestion, enforcing SLAs).
- Incentivize heavy users to move to different tiered plans and personalized plans that include highly valued zero-rated applications. Upsell, cross-sell, and use quota top-ups.
- Develop plans and pricing that are continuously tailored to customers’ needs and usage patterns (and their correlating impact on the network).
Sandvine Heavy Usage Solution
Subscriber Service Analysis of subscriber and subscriber groups, individual services and service categories, as well as devices.
Heavy User Analysis to identify different types of heavy users, their behaviors, preferred apps, and impact on overall QoE of other subscribers.
Heavy User Management for more effective decision making and policy validation through monitoring, control and use case-specific visibility.
Video Streaming Management through more control of video resolutions (by application, device, ToD, and location), service differentiation, and increased ARPU (e.g., HD Hour/Day Pass).
Automated Mobile Congestion Management improves application quality of experience while managing congestion in the network.
Intent-Based Congestion Management with analytics and closed-loop automation handles congestion based on QoE targets.
|This phased approach requires intelligent access-network analysis and constant monitoring of heavy usage and subscriber QoE. With the classification of different subscriber groups and usage behaviors, you are empowered to implement smart policies and to evolve toward closed-loop automation and Intent-Based Congestion Management.
To execute on that approach, we have packaged key use cases and capabilities into a heavy usage solution (see sidebar, Sandvine Heavy Usage Solution).
By combining our heuristics-driven Network Optimization with Application and Network Intelligence, Sandvine customers get deep analysis of application usage trends – by subscriber –viewed through the lens of different plans, devices, and classifications.
When you leverage analytics and closed-loop automation, it’s possible for application-specific policies to automate network management, which eliminates the need for continuous manual parameter fine-tuning.