Which Media Transcoding Strategy Is Right for Your Network?
Total Page:16
File Type:pdf, Size:1020Kb
White Paper Which Media Transcoding Strategy Is Right for Your Network? (Hint: It might not be the strategy you’re using today) Which Media Transcoding Strategy Is Right for Your Network? White Paper (Hint: It might not be the strategy you’re using today) Executive Summary The rapid and rising demand for multimedia mobile devices such as smartphones and tablets has changed the landscape of telecommunications networks. As Communications Service Providers (CSPs) engage in network transformation efforts around Long Term Evolution (LTE) and IP Multimedia Subsystem (IMS) specifications to meet this growing demand, they are also facing new media transcoding challenges brought on by the multiplicity of mobile and legacy devices and the variety of media (in which video will play an increasingly vital role). How CSPs respond to the media transcoding challenge will have a profound impact on their business, as nearly half of all voice/video sessions will require some type of transcoding. This whitepaper examines requirements for media transcoding and processing in modern networks, as well as “best practices” and strategies for handling media in the future. In particular, the focus will be on technologies such as virtualization and Cloud- based service delivery, and the roles they play in providing a more scalable, flexible media transcoding platform. 2 Which Media Transcoding Strategy Is Right for Your Network? White Paper (Hint: It might not be the strategy you’re using today) Table of Contents The Market Challenge: More, More, More Media ................................................4 Media Processing Requirements In A Modern Network............................................5 Voice Quality Enhancements ...........................................................5 Support for the Latest Codecs . .5 Support for Video Codecs .............................................................5 Scalability .........................................................................6 Support for New Standards ............................................................6 Best Practices For Media Transcoding ........................................................6 Modern Strategies For Media Transcoding .....................................................6 Media Transcoding at the Network Edge ..................................................7 Centralized Media Transcoding .........................................................8 Cloud-Based Media Transcoding ........................................................8 The Dialogic Family Of Media Transcoding Solutions..............................................8 Dialogic® I-Gate® 4000 PRO Media Gateway ...............................................8 Dialogic® PowerMedia™ XMS...........................................................9 Summary..............................................................................9 About Dialogic .....................................................................9 3 Which Media Transcoding Strategy Is Right for Your Network? White Paper (Hint: It might not be the strategy you’re using today) The Market Challenge: More, More, More Media Each year it seems smartphones and tablets get smarter, bigger and more powerful, which in turns drives consumer appetite for richer media communications. As mobile devices evolve, the result is more media traffic (particularly video) and more bandwidth constraints for telecommunications networks. The ability of CSPs to meet the demand for better, faster delivery of voice, video and new apps over mobile, fixed or wireless access networks – some with less than adequate bandwidth - will depend heavily on the capability of next-generation media processing platforms and their support for the wide variety of encoding methodologies. There is a significant number of different voice and video encoding methodologies or codecs used in networks and devices. Codecs have been developed over the years to suit a variety of purposes; some have been designed to reduce bandwidth consumption (e.g., narrowband codecs), while others have been designed to deliver superior voice quality (e.g., wideband codecs). (See Figure 1 below for a sampling of narrowband and wideband codecs.) fullband stereo Opus fullband MP3 AAC Vorbis super-wideband G.719 G.722.1C wideband royalty-free, open-source AMR-WB free license, not open-source licensing fees, not open-source Quality G.722 Speex narrowband AMR-NB iLBC G.711 bitrate (kb/s) Figure 1 - A sample of narrowband and wideband codecs used today (Jean-Marc Valin – www.opus-codec.org) As mobile communications evolve and new technologies such as WebRTC emerge, new and innovative codecs are also developed to support them. The proliferation of codecs places greater demand on CSPs to support the interworking of both new and legacy codecs in their networks. The process of converting one encoding methodology to another is called transcoding, and media gateways and media servers can use software-based or Digital Signal Processing (DSP) approaches to perform this task. Both software- and DSP-based transcoding approaches have their benefits and drawbacks: • Hardware-based transcoding provides fast processing and high scalability, but it can take months or even years to accommodate and test new codecs, depending on the underlying equipment vendor’s roadmap and resources. • A software-based approach may require more CPU resources, but new codecs and thus new services can be spun-up fast without needing to upgrade hardware. Also, the processing power of modern commercial-off-the-shelf (COTS) servers is constantly improving, as is server virtualization technology. Taken together, they make the case for software-based transcoding more appealing. Regardless, media transcoding is becoming not simply an operational imperative but a competitive benefit as well. The ability to support new codecs quickly allows CSPs to offer new revenue-generating services to their customers and provide superior voice quality across more applications and devices. Consider, for example, that the Opus audio codec released at the end of 2013 is one of the codecs supported for WebRTC applications. CSPs that cannot support the Opus codec might find themselves at a competitive disadvantage if their media transcoding infrastructure cannot support this and other WebRTC applicable codecs. 4 Which Media Transcoding Strategy Is Right for Your Network? White Paper (Hint: It might not be the strategy you’re using today) Media Processing Requirements In A Modern Network Voice Quality Enhancement Techniques The role of media transcoding is paramount to ubiquitously delivering end-to-end next Silence Suppression, Voice Activity Detection and Comfort generation services. However, CSPs should also look for other capabilities in a media Noise Generation processing solution to achieve superior sound quality and to support more calls across Marking and eliminating the silent periods between words the network. In addition to transcoding, media processing solutions in a modern network and sentences can reduce bandwidth consumption by should include: fifty percent or more. Modern media gateways and servers should be able to detect periods of silence and suppress • Voice quality enhancements those packets and inject the reassuring comfort noise at the far end to deliver the same call experience in half the • Support for the latest codecs bandwidth. • Support for video codecs Adaptive Jitter Buffer • Scalability Media gateways and servers should provide adaptive • Support for new standards correction to the play out of packets subject to IP transport impairments that cause variable delay in the packet arrival Voice Quality Enhancements time (jitter). Small buffers are used that provide smoother and more seamless media packet delivery. The introduction of technology and techniques to minimize the required bandwidth Packet Loss Concealment and address packet impairment effects inherent in IP networks is integral to VoIP High levels of packet loss can result in unacceptable service, communications. Silence suppression, Voice Activity Detection, Comfort Noise, adaptive so media processing solutions should have effective packet jitter buffering and packet loss concealment techniques are all important features to loss concealment (PLC) algorithms to replace the lost enable a richer user experience. packets in a voice stream in order for the sound produced to be natural and seamless. Support for the Latest Codecs High-definition (HD) voice codecs are used in Voice over LTE (VoLTE) and they are also of particular interest to enterprises and consumers because of their ability to deliver superior customer experience. Adaptive Multi-Rate Wideband (AMR-WB, also known as G.722.2), Internet Speech Audio Codec (iSAC), Internet Low Bitrate Codec (iLBC), and Skype’s SILK are examples of high-quality voice codecs. In addition, WebRTC, a game changing technology that is fast revolutionizing communications, utilizes Opus. Future-ready support for newer codecs on media gateways is critical for CSPs looking to expand their services portfolio. Support for Video Codecs Video communications are notorious bandwidth “hogs,” so in order to reduce the amount of bandwidth that video consumes while preserving the quality of service, CSPs look to video codecs such as H.263, H.264 and MPEG-4 to compress and interwork video sessions between different devices such as smartphones, tablets, videophones and PCs/laptops. (See Table 1) Network Bandwidth available Terminals Codecs 3G-324M 64 kbps Video