Audio transcoding in the network is most often performed by Session Border Controllers. Traditional physical SBCs invariably make use of specialized hardware designed to minimize the cost, space occupancy and power consumption of the transcoding function. At the heart of that specialized hardware we find Digital Signal Processors (DSPs).
This white paper discusses the transition of the audio transcoding function from proprietary hardware-based SBCs into the NFV cloud environment, and sets out to answer the following questions:
How effective is today’s industry-standard hardware at performing audio transcoding?
To what extent do suitable acceleration technologies (e.g. DSPs) improve the effectiveness of audio transcoding in an NFV environment?
What are the practical aspects of deploying transcoding accelerators in an NFV environment?
How is industry-standard hardware evolving in respect of its ability to support audio transcoding, and how does that impact on decisions around the deployment of acceleration technologies for audio transcoding?
The aim of the paper is to give planners the tools they need to make informed decisions about how best to support audio transcoding in an NFV environment today and in the near future.