Optimizing Existing Networks
Vocality is about throughput and getting more devices and more services on IP. Users want more data, more video, and more audio. Yet increasing bandwidth to meet these requirements is not without challenges, and often not possible. This paper walks you through the demand for more communications and how Vocality’s optimizing technologies provide for more throughput and higher quality.
When you need to communicate with people or devices at the forward edge of the network, there is often a price to pay when it comes to efficiency and quality of the communications – especially when transmissions rely on satellite or other air-based methods. Cases of this can be seen when you do not have enough bandwidth to make that additional call, you have to yell or repeat yourself when you shouldn’t have to, or you have given up on video conferencing (VTCs) altogether on encrypted networks.
Well the problem seems clear at first look: you need either more bandwidth or a new network. However, that’s not necessarily true. What you need is a way to optimize your existing network to handle the increased load generated by video conferencing, VoIP connections, encrypted communication, and other data applications. If you can optimize the traffic, without changing the devices the users connect to, you would be able to handle the demands of a growing worldwide enterprise.
For example, it is easy to see above how on a 1Mb link you are limited to twelve concurrent calls using a G.711 64K codec. And that’s assuming no one else is trying to use any of the bandwidth for anything else. When safety is on the line, that thirteenth call can be just as important as the other twelve, but the network cannot support the call.
The logistics around inline optimizations has to be easy as well.
- You never know where users will end up, so remote administration is an obvious requirement.
- Power is sometimes very hard to come by, so it has to run at very low amperage.
- Size matters for mobile deployments. Something that can fit in a pocket allows for quick and low cost deployment by reducing logistics management and overhead.
- It must not impede already encrypted or otherwise optimized traffic.
- Finally, in addition to being in-line, it must have a fail-to-wire capability so that the customer can always revert to what they previously had, even if it was not enough.
With input from customers, Vocality created BASICS Optimizer. This paper doesn’t talk to the product specifically as much as the optimization techniques that are implemented in its use, namely Header Compression, Packet Aggregation, Codec Forcing, and Intelligent Dejittering.
It is recommended the reader click on the picture below to see a two minute video on YouTube that walks through what BASICS Optimizer does in a humorous, story-telling way.
After watching the video, read the explanation behind the technology below.
When you remove the extra payload caused by unnecessary IP headers, you regain bandwidth for other services. With G.729 8kbps call payloads, that can amount to over a 40% recovery of bandwidth available for other services.
A 60% bandwidth savings may be realized when deployed with encryption. This is because the bandwidth that would be used to encrypt the compressed headers is no longer needed.
With the efficiencies gained by IP header compression, the packets themselves comprise of less information. That allows the merging up to 26 concurrent calls in the same packet previously handled a single call. This aggregation greatly reduces the load of in-line encryptors and modems by reducing the processed packets by 96%.
The fewer packets they process, the fewer packets they drop. And the fewer packets they drop, the clearer the voice and video.
In larger networks where VoIP phones have been deployed, it is not feasible to have a technician go to everyone’s desk and configure each phone to remove its default codec, generally G.711 64kbps. However, G.711 will use eight times the bandwidth of a G.729 8kbps. The technique must be employed where the optimizing network forces G.711 off the negotiating codec library. In this scenario, an 8x saving is achieved through optimization of IP headers through Packet Aggregation and Header Compression, but also with better utilization of the payload.
There is a proprietary technique used that inserts small timestamps between the packets prior to aggregation. These 4-byte stamps are coupled to compressed packets that represent the relative rate in which each packet was captured. This allows the hub and spoke network to release the packets on the other end of the network in the right order, and at the same rate they were captured. Now voice and video is as clear and clean as it would be if it were travelling over a terrestrial connection.
Intelligent De-Jittering Specific to Video:
The Intelligent de-jittering technique improves video transmission as well. Generally, video is compressed; however, IP packets are still susceptible to arriving out of order. When that happens, these UDP packets are dropped, and the viewer experiences “green screen”. Green screen is the way the video encoder presents frames when data has been lost. Vocality’s intelligent de-jittering optimization techniques reduce instances of green screening and improves associated audio that may accompany videos, such as with VTCs.