Frequently asked questions.

Bequant detects network congestion and manages the network capacity while improving the subscriber quality of experience. Keeping QoE constant, the network can accommodate more subscribers. Anibal Enríquez, CEO of Communicate, has reported 10% more users per Access Point.

Keeping the number of subscribers constant, the subscriber QoE is improved by reducing the network latency and losses. Cajun Networks has reported 83% reduction in latency and retransmissions going down from a high of 6,6% to an average of 0.32%.

Bequant traffic AI congestion mitigation (ACM) mitigates the congestion, active queue management (AQM) enforces rate limits in an optimal way and application-based traffic policies (DPI enabled) moderate the demand of heavy applications during peak hours.

More information can be found on Cajun Case study and Enhanced Subscriber QoE.

You can prevent routers from being overloaded by offloading subscriber plan enforcement to Bequant traffic management system. At the core of Bequant architecture sits an advanced queue management (AQM) built to scale. Bequant runs on off-the-shelf Intel or AMD servers and supports very high throughput, up to 400 Gbps in one single server.

More information can be found on BQN Platform.

Routers run an optimal packet fast path for core routing functionality and a less optimal path for ACL processing or QoS queues. In many models, just activating those features increases the CPU load and memory usage and reduces the throughput capacity. The impact becomes bigger as more entries are added.

In Bequant, our optimal path involves traffic classification, queues and policy enforcement because traffic management is our core functionality. Bequant incorporates technology such as DPDK to obtain optimal performance.

Bequant runs on off-the-shelf Intel or AMD servers, which typically have faster CPU and memory than router hardware. Thanks to superior hardware and implementation, Bequant supports very high throughput, up to 400 Gbps in one single server.

More information can be found on BQN Platform.

Bequant can enforce per-subscriber traffic policies with very little work by the network engineering team. Bequant can get policies or plans automatically from RADIUS signaling, interacting with the billing system (20 vendors so far and growing) or getting them via Bequant REST API.

Bequant AQM automatically detects and prioritizes interactive flows such as teleconferences or gaming, and its AI congestion mitigation (ACM) is self-learning and does not require manual fine tuning. The manual policies can be much simpler and stable, freeing engineering time.

Bequant improvement of subscriber QoS reduces support calls, further alleviating network engineering workload. Chadd Giles, CTO of Resound Networks, reports a 35% drop in support calls.

Yes, Bequant can enforce per-subscriber traffic policies interacting with the billing system. Bequant is currently integrated with 20 vendors so far and growing. The billing vendors include:

• Aradial
• Azotel
• Gaiia
• Gestfy
• HubSoft
• ISPCube
• ISP Gestión
• ISPSolution
• IXC
• Mikrowisp
• Phicus
• Powercode
• Sonar
• Splynx
• UISP
• Vision by Fibersmith
• Visp.net
• WISPControl
• Wisphub
• Wispro

More information can be found on BQN integrated Billing Systems.

Application-aware traffic shaping is a network management technique that identifies the type of application generating traffic (gaming, VoIP, streaming, large downloads, etc.) and then dynamically prioritizes the flows based on that identification. Bequant implements application-aware traffic shaping in two ways:

• Deep-packet-inspection (DPI): identifies the traffic to apply configured policies. One example is the identification of video streaming during peak hours to moderate its demand.
• Prioritization of interactive flows: Bequant automatically identifies interactive flows by their behavior and prioritizes them in the AQM. This requires no explicit policy configuration.

In most cases, additional network capacity will not deliver any gains, because the speed is mostly limited by the access network and the Wi-Fi last hop. Bequant traffic optimization features (TCPO, ACM, AQM, etc.) will reduce congestion in those network elements and improve the quality of experience in a way that an increase in network capacity will not deliver.

An example of this is Cajun Networks, who has reported 83% reduction in latency and retransmissions going down from a high of 6,6% to an average of 0.32% just deploying Bequant, without any network capacity expansion.

You can request a Bequant free trial on BQN Free Trial Application Form.

Managing subscriber policies in a centralized node scales better as traffic grows because a product like Bequant has bigger capacity, it is more economical to deploy and easier to maintain.

Regarding bigger capacity, Bequant uses Intel/AMD server with faster CPU and memory than router hardware and its internal architecture is optimized for policy management. An added benefit is that Bequant sees all traffic and can apply network-wide fair use policies using very simple rules.

Regarding deployment, one single powerful hardware can be much less expensive to acquire and roll out than many smaller ones, including the interaction with a RADIUS or billing system to get the policies from.

Finally, a distributed solution is cumbersome to maintain, because each element will have its own queue configuration to keep coherent with the rest.

You need to apply fair usage policies based on actual network bandwidth limitations. This will not degrade but improve the QoE for most users.

For example, if a segment of your network has a backhaul link that gets saturated during peak-hours, you can use Bequant traffic management to set a limit for the most bandwidth-hungry applications (e.g. video streaming), and a global rate limitation with scheduling that guarantees a minimum bandwidth for all users. As a result, the QoE of the customers served through that link will improve.

In the end, when well implemented, fair-usage policies allow you to share a limited bandwidth between many users while improving their quality of experience.

A flow monitor is not an active element, it does not act on the traffic, it gets flow information from the traffic devices and makes that information available for analysis. A traffic optimization platform such as Bequant, is deployed online and improves the traffic performance leveraging on the metrics it gets, so the metrics become actionable straight away, even when nobody in network engineering is analyzing the data.

To identify the applications with the biggest bandwidth demand, a traffic equipment with deep packet inspection (DPI) is needed. Bequant has DPI capabilities and it can identify the most popular applications of a specific subscriber, a group of subscribers or the network as a whole. Traffic management rules can be defined to manage those applications and prevent their congestion effects during peak times.

Find an example on how an ISP solved streaming overload in La Harpe Telephone Case study.

It helps by providing key insights into how the subscriber is using the Internet connection. For example, using a DPI enabled product such as Bequant, a technical support team can see what applications are being used at the home network and realize that the reported problem is related to that usage, for example to a big download in combination with streaming sessions by several members of the family.

The network metrics that can signal degradation of the quality of experience better are congestion, losses and latency. If any of these three or all of them combined increase and remain high for some time, there is a high probability that the customer is suffering from a degraded service and will soon complain, or worse, decide to leave without prior notice. This is the reason a product such as Bequant is important; to mitigate QoE issues and to give an ISP the network visibility it needs to proactively solve network issues before they cause churn.

See more information in this Case Study on churn reduction.

By using a product with network visibility such as Bequant, the ISP support team have access to key QoE indicators per subscriber and per network sections that will allow them to focus on impending network issues and address them proactively. Bequant has an AI Congestion Mitigation feature (ACM) that is constantly looking for congestion symptoms per individual subscriber. ACM automatically acts to reduce congestion, but it also reports the congestion found to the network visibility feature, so support teams can rate subscribers based on their experienced congestion. Bequant has tools to facilitate data analysis, such as an automatic ranking of the worst percentile of congestion-suffering subscribers.

Bequant also enables traffic captures and a live view of a subscriber connection at any given time. This real-time view of the network at different levels is what Victor de la Nuez, CEO of WiFi Canarias, highlights about Bequant.

Traffic optimization and network performance data can be combined in a product such as Bequant, that optimizes traffic at the same time it provides metrics of the resulting network performance.

The provided metrics leverage on the traffic optimization part, because Bequant TCP optimization provides accurate latency and losses values, obtained in flows under load, which is what end users experience. This accuracy is key for reliable congestion detection.

TCP Optimization also measures its own acceleration effect by leaving a small fraction of the traffic as a control group to compare the improvements against.

In the latency case, Bequant provides three separate latency metrics to give as much visibility as possible: an average of minimum values (a lower bound of latency values), the median and the percentile-80 (value not exceeded by 80% of the measured values).

The network metrics that can signal degradation of the quality of experience better are congestion, losses and latency. If any of these three or all of them combined increase and remain high for some time, there is a high probability that the customer is suffering from a degraded service. Latency affects the responsiveness of an application, critical in real time applications like teleconferences and gaming. Losses reduce the application available goodput (actual transfer rate of application content), decreasing the effective speed. Finally, congestion brings a degraded experience, with latencies and losses increasing while the available goodput collapses.

This is the reason a product such as Bequant is important; to mitigate QoE issues and to give an ISP the network visibility it needs to proactively solve network issues.

By using a traffic optimization and visibility product that has enough capacity to process all the network traffic from one single location. If geographically distributed locations are required, Bequant servers can be configured in a cluster where any location is accessible from any of the other sites. Of course, collocated servers can also be added to the cluster to collaborate seamlessly with the rest.

AI-driven network analytics speed up the identification of network problems by allowing the support team to express the goals of the issue search rather than following the steps needed to find them, which is more time consuming and requires knowledge of the product capabilities. An AI assistant is a quick way to obtain results for the most common cases, leaving direct interaction with the visibility data for the most advanced cases.

First, the network has deployed a traffic optimization element such as Bequant, because the first level of proactiveness is improving network performance without any human intervention, just using a closed control loop that is continuously processing the traffic, identifying issues and solving them as they occur.

The second level of proactiveness is for that traffic optimization product to provide network visibility that is accurate and detailed enough to detect QoE issues as soon as they happen, even before the affected customers complain.

The third level is importing the traffic optimization metrics into the network management system for analysis and notifications. Next generation management systems also integrate BSS and OSS system information to have a more holistic view of the subscriber satisfaction: the experienced network quality obtained from traffic optimization along with the history of interactions with customer support.