Contact us at 216.468.5200 or info@accelerationsystems.com

QoS/Adaptive Queue Management

Adaptive Queue Management/QoS

It takes more than just optimization techniques like TCP optimization, protocol manipulation, block level deduplication, in-stream compression, data caching to make your Internet experience better. Our technology not only accelerates and encrypts the Internet traffic passed through the RBA, we make the whole network capacity more efficient and useable.  All the optimization techniques in the world will not help your network performance and Internet speed if the data being sent is faster than the network on the other side of the gateway can handle.  If excessive traffic at the gateway is not properly managed, the entire network will experience negative effects in regards to network efficiency and quality of service.

Acceleration Systems has introduced an advanced traffic management algorithm that can solve this gateway capacity issue. We have figured out a traffic management method that senses changes in the Internet connection upstream capacity and then regulates the rate that the LAN-side applications transmit data to the Internet gateway device.  Pretty cool right?  We think so. We call this new technology Adaptive Queue Management (AQM).

So how does this new Adaptive Queue Management work?  Glad you asked.  It is actually quite simple when you break it down and look at it.  For a deeper dig into this technology, you can read our white paper about Adaptive Queue Management.

TCP Traffic Management

When two networks operating at different speeds merge together in order to move data back and forth, the flow of packets from the faster network must be managed in order to avoid packet queue, otherwise known as buffer bloat.  Buffer bloat will happen when the higher speed network continues to send data at a rate above the capacity of the other network and queues up at the gateway.  An unmanaged gateway causes the TCP to transmit to the gateway at rates which keep the buffer full or cause it to overflow.  This over flow causes a packet queue to increase and develop bottlenecks that impede the transfer of data to the slower connection.  Buffer bloat also negatively impacts the timely prioritization of packets by the queue manager.

 

Acceleration Systems AQM White Paper

 

Existing TCP mechanisms used to manage congestion rely on the ability to sense pocket drops to determine available bandwidth.  The problem with the reliance upon the mechanism to detect dropped packets assumes that packets are dropped in a timely manner.  This is not the case because TCP does not have a way to detect if data packets are stored in the buffer, therefore TCP does not slow down to match the second link until after the gateway buffer discards packets due to data overflow.

Adaptive Queue Management

We’ve determined that there had to be a better way of controlling the transmission speed rate, no longer just waiting for packets to drop. AQM keeps the buffer delay at or below 5 milliseconds by monitoring the time packets spent in the gateway buffer.  This technology allows for control and changes to the speed of data on a near real-time basis.  Adaptive Queue Management will remain dormant if the buffer is relatively empty or if the delay is below the 5 millisecond limit.

Advantages of Adaptive Queue Management:

  • Eliminates TCP overhead
  • Configuring parameters is a thing of the past
  • AQM responds directly to conditions in the buffer of the gateway router, not external conditions such as round-trip delay, connection rate, traffic load, and other factors that cannot be controlled or predicted at the local buffer.
  • Local queue delay is determined when packets leave the buffer alleviating further delay by waiting on acknowledgements from the next network segment.
  • AQM adjusts to changing link rates without negative impact on available network utilization.
  • Tests have shown that AQM produces link utilization consistently near 100% of the available bandwidth.

Traffic Prioritization

We understand the daunting challenges that network administrators are faced with when configuring Quality of Service (QoS) by establishing priorities and bandwidth allocations for various types of traffic on the network.  Conventional QoS assumes that connections run at a consistent speed, but most connections are contended, resulting in constant capacity changes. The amount of available bandwidth varies moment to moment as users grapple for a limited resource.

When a gateway buffer fills or overflows, time sensitive packets, like VoIP, can get delayed due to its position in the queue. Even if VoIP traffic is given a top priority in the QoS configuration, the queue manager may not prioritize VoIP packets quickly enough to maintain the expected quality of the call. Thankfully Acceleration Systems has developed a prioritization hierarchy for protocols and data types eliminating the need for custom configurations. This process eliminates long queues in the buffer at the gateway where Quality of Service is applied and avoids backups, improving the overall throughput. With AQM, the queue manager can easily prioritize time sensitive data.  Our priority stack has proven successful in over 95% of the environments in which it has been used. Of course, network administrators do still have the ability to adjust QoS settings in order to meet local requirements if needed.

With the assistance of Acceleration Systems’ Adaptive Queue Management controlling an efficient flow of data, the queue manager is able to prioritize the traffic in a timely manner.  This new combination of prioritization and flow control produces a more efficient use of the network capacity than previously achieved.  And the best part is that the whole network capacity is used in a more efficient manner without any extra work or effort from network administers.  Pretty sweet! We’re betting network administrators the world over will love us for this feature alone.