Bandwidth Control, Rate Limiting, and Traffic Management: Everything You Need to Know

WISP Bandwidth Control, Rate Limiting, and Traffic Management Everything You Need to Know

by | June 13, 2019 | Networking

The Need for Bandwidth Control or Traffic Management

Whether it’s optimizing use of backhaul capacity or selling plan speeds based on customer needs, most ISP owners or operators need a bandwidth control solution. Some ISPs refer to wireless network bandwidth management as traffic shaping or rate-limiting, while others call it plan enforcement. Beyond the terminology, the importance of an effective bandwidth control solution cannot be understated.

Subscriber needs vary. As a result, most ISPs sell plans with speeds like 10 Mbps down and 1 Mbps up for small families. Or 25-50 Mbps or more for gamers and heavy video streamers or downloaders.

From an enforcement perspective, many owners and operators think that as long as they have a bandwidth control or management solution that limits the customer to their speed and plan, that’s enough.

In reality, however, there’s much more to ISP traffic management and bandwidth control solutions than simple enforcement. More importantly, how ISPs rate limit or enforce their subscriber plans has a large impact on the customer experience.

How ISPs Enforce Bandwidth Management Matters

Almost all ISPs enforce subscriber plans. The choice of a bandwidth control or ISP traffic shaping solution has a direct impact on subscriber Quality of Experience (QoE).

Poor plan enforcement typically leads to ‘my-internet-is-slow’ calls. Generally, when a customer calls in with this kind of complaint, they might be downloading the latest Windows update, using Netflix, or otherwise hitting the bandwidth limit of their plan or connection. This doesn’t have to be the case. This bad QoE is the result of using poor plan enforcement techniques.

Many ISPs deploy one of the following three bandwidth control methods. While each does a good job of basically limiting customers to their plan speeds, their impact on QoE is often not understood by ISPs.

1. Access Point (AP) Enforcement

Most AP hardware allows ISPs to set per-subscriber bandwidth limits directly on an access point. AP bandwidth shaping tools use simple queueing techniques. With simple queues, if more data comes in than what the queue can handle, packets are either dropped or buffered. This can lead to Bufferbloat problems or slow-internet complaints when multiple users in the home are online at the same time. In most cases, this leads to a very poor subscriber experience and complaints that require costly customer support or truck rolls.

Bottom line: AP-level shaping tools use simple queuing techniques which allow heavy/bulk applications like streaming video (for example, Netflix) to negatively impact the performance of smaller/more interactive applications (like gaming, VoIP, and DNS queries). So, if a household is streaming video and maxing out its plan speed, all other applications/users will suffer.

2. Routers

Some ISPs use scripts or rules set in their edge routers (like a Mikrotik) or customer-premises equipment (CPE). Due to hardware limitations, these platforms also use simple queuing techniques that let heavy applications negatively impact performance. Like AP-level enforcement, router-level traffic enforcement or shaping is incapable of subscriber-aware policies such as multi-site bandwidth limits.

Bottom line: Router or CPE-level shapers use simple queuing techniques that allow heavy applications like streaming video to negatively impact the performance of interactive applications.

3. Application-Aware (DPI-based) Traffic Management Tools

Some network managers choose solutions that inspect individual applications. These solutions use a complex set of rules to treat individual traffic flows differently and set individual priorities. For example, an ISP will set manual rules like capping Netflix traffic at ‘X’ Mbps, VoIP at ‘Y’ Mbps, and so on. DPI-based solutions require constant tweaking and upgrades to stay current with rapid application changes.

Also, setting arbitrary bandwidth limits is inherently complex and time-consuming. For example, it’s impossible to know the appropriate rate to shape Netflix, as it varies with the video encoding, content, and even the type of device that’s playing the content. Setting these values incorrectly can make the subscriber QoE worse than having no bandwidth control solution. Furthermore, with an increasing amount of encrypted traffic, DPI-based bandwidth control solutions have declining utility.

Bottom line: DPI-based application-aware solutions use simple queue techniques and are costly, complex, require constant tweaking, and have declining relevance given the rise of encryption.

Bandwidth Control: There’s a Better Way

When it comes to effectively managing bandwidth and enforcing subscriber plans, there’s good news for ISPs. Much progress has been made in the field of active queue management (AQM) in recent years. One innovation coming out of the IETF (Internet Engineering Task Force) is FQ-CoDel. This algorithm is a combined packet scheduler and AQM technique developed as part of the Bufferbloat-fighting community effort that’s also application-agnostic.

In the AQM approach shown on the right, flows are automatically categorized into bulk or interactive based on how much queue they build up. So, applications like streaming video or Netflix get categorized as bulk flows.

On the other hand, VoIP, gaming, or DNS queries that are latency-sensitive get categorized as interactive flows and are instantly prioritized.

As a result, even when a customer maxes out their plan with Netflix or other heavy downloads, the other applications don’t feel slow. The overall experience is much better.

WISP Bandwidth Control: Preseem FQ CoDel Diagram

QoE Optimization and Bandwidth Management

Preseem uses FQ-CoDel to offer an easy and effective traffic shaping and plan enforcement solution. This QoE-optimized bandwidth control solution helps ISPs:

  • Keep just the right amount of queue in the device to ensure high link utilization and low latency
  • Ensure per-flow fairness and prioritization (interactive flows get implicitly prioritized to the front, avoiding any impact from bulk applications)
  • Avoid complex policies, arbitrary rules, or encryption concerns

ISPs that deploy Preseem’s QoE optimized shaping (based on FQ-CoDel and AQM techniques) get:

For more information on Preseem’s QoE monitoring and optimization platform, click here to book a demo with us.

Subscribe to the Preseem Blog Newsletter

Stay in-the-know and get fresh content delivered to your inbox once a month.