TCP Optimization & Acceleration
Bequant TCP Optimization
TCP is the most widely used protocol in the Internet, but, like other transport protocols, when networks are not perfect, it tends to transmit at a lower speed than what the network allows. The Bequant TCP Optimization functionality, makes sure that TCP traffic flowing through it reaches optimal speeds, resulting in faster downloads and uploads.
.svg)
Transparent
TCP Proxy
The Bequant BQN works as an in-line transparent TCP proxy (as defined in RFC 3135), to perform TCP Optimization, making data appear closer to TCP clients, which results in a faster delivery. This kind of TCP proxies have been used for many years in satellite networks and in large wireless cellular networks.
Faster TCP Stack
The Bequant TCP stack detects congestion based on the measured delivery rate, as described in our patent portfolio. Our improved congestion detection algorithm allows our TCP to be more aggressive when there is no congestion and to back off in real congestion, being able to carry out this discrimination in challenging environments: with shallow buffers, wireless links, packet losses or large network delays.
The following graph shows the reduction in the page loading time of several popular web sites and applications, measured in a large mobile cellular network over one week with thousands of measurements.
Results in Different Environments
Our TCP Optimization has been in operation for years in over 250 ISP and enterprise networks around the world. The following graph shows the average speed of all Netflix downloads in very different networks, from Wireless ISPs to fibre and cable networks. In each case, over a one-week period, we measured the speed of Netflix downloads accelerated by our application and Netflix downloads that were not being accelerated, and compared the results.
Deployment
The TCP Optimization, Bandwidth Management and Network Visibility functionalities run, in an integrated way, on the Bequant BQN software platform, which can be deployed on off-the-shelf Intel- or AMD-based servers and on virtual machines (KVM or VMware vSphere hypervisors).
