The Role of the Interconnect Network in Enabling HD Voice Communications

By TMCnet Special Guest
Mohan Palat, principal product manager at Sonus Networks
  |  June 01, 2011

VoIP traffic has been growing exponentially in recent years in both fixed and mobile networks, as both enterprises and consumers seek to improve efficiency and reduce costs by increasing their adoption of IP services. At the same time, the emergence of HD voice promises to deliver on the true potential of VoIP by offering clear and life-like reproduction of audio.

Current narrowband telephone networks can only handle voice in the range of 300 Hz to approximately 3500 Hz. Voice outside this frequency range is discarded by the narrowband telephone network, compromising voice quality and degrading voice that is reproduced at the receiving end. The adoption of HD voice in the enterprise and by service providers effectively extends the voice frequency range from 30 Hz to 7000 Hz, which reproduces natural human speech more accurately.

One immediate benefit of this more accurate reproduction of the human voice is the ability to improve voice communication across cultures and countries. Most simply put, HD voice makes it easier to understand voice content despite the presence of different accents during a phone call. As important to improving overall call quality, HD has proven effective at eliminating some background noise, so voice calls from noisy locations are much more intelligible and natural. HD technology takes the stress out of phone calls because the participants do not have to repeat words or sentences, speak loudly, or strain to hear the spoken words. HD voice also makes it easier to identify who is talking during conference calls with two or more participants who have similar voices. In fact, informal surveys have shown that productivity increases when callers are able to hold conversations without asking each other to repeat a word or sentence. In addition, HD makes it easier to comprehend speakers who talk softly during phone conversations and to understand words when more than one person speaks at the same time during a call. HD is useful for speech-to-text applications where speech clarity reduces the number of false translations and improves the user experience.

In addition to the consumer benefits, HD technology offers benefits to service providers and enterprises. An improved user experience often results in longer call duration, leading to increased revenues. Increased customer satisfaction due to better quality of experience reduces churn. HD voice also can attract new customers who are drawn to the service by the promise of better quality voice. Higher voice quality leads to improved brand image, and businesses that rely on call centers will benefit from HD in such ways as improved order accuracy, faster problem resolution and improved quality of experience for customers. This can result in increased business for the company. Finally, HD improves the quality of audio broadcasts, resulting in improved customer satisfaction for such services. 

Moving from Standard to HD Voice

HD voice requires the support of both network and the endpoint devices (phones). It also requires the use of specialized HD voice codecs, which are different from the narrowband G.711 voice codecs currently used in traditional telecom networks. The International Telecommunication Union’s G.722 codec is the most widely used HD voice codec in telecom networks and enterprises. Since it was standardized in the 1980s, all the patents on G.722 have expired, and there is no license fee to use it, which makes it attractive to vendors. 3GPP-defined AMR-WB (also referred to as ITU G.722.2) is the standardized HD voice codec for 3G/UMTS mobile networks. Patents associated with this codec are still in force, making it less attractive to vendors. In addition to G.722 and AMR-WB, there are a number of proprietary HD voice codecs, such as Skype (News - Alert) SILK and iSAC, in use today.

Although most of the discussion on HD voice is focused on devices, codecs and access networks, it’s crucial to understand the role the IP core and interconnect network plays in enabling HD voice communications. To establish and maintain an end-to-end HD call, the IP core/interconnect network must enable the interworking of multiple heterogeneous devices and access networks, and it must do so efficiently and cost effectively. To accomplish this, the core/interconnect network must support functions that will ensure that an end-to-end HD voice call has the highest level of quality and operational efficiency, real-time transcoding, media pass-through, and intelligent routing.

Real-Time Transcoding

The interconnect network must provide real-time transcoding between different HD codecs during a VoIP call. For example, if the calling party originates an HD voice call using an AMR-WB phone and the called party is on a G.722 phone in another network, the interconnect network must determine that the two HD codecs are different and provide the necessary transcoding of the IP voice traffic between the two endpoints. Or, if the originating party initiates an HD voice call using a G.722 phone to a person who is on a G.722.1 or G.722.2 phone, the interconnect network must provide real-time transcoding of the G.722 audio to G.722.1 audio, while attempting to maintain the quality of the audio as it transits the network to the terminating device. Without the capability to transcode between the two codecs, the network will downgrade the HD voice call to a traditional narrowband call. Transcoding makes it possible to establish an end-to-end HD voice call between two HD-enabled devices, independent of the codecs being used.


Media Pass-Through

The interconnect network must also support media pass-through, which essentially allows the HD VoIP call to transit the network without transcoding (also referred to as transcoder-free operation). This is useful in cases where the calling and called parties support the same HD voice codec. It is also useful when the service provider wants to reduce the number of transcoding operations by avoiding any transcoding until the call reaches the terminating network. As transcoding introduces signal loss, limiting it or preventing it altogether preserves the quality of the transmitted audio signal. By supporting the pass-through mode, the interconnect network can ensure that the HD voice signal maintains the highest possible quality as it transits the network.

Intelligent Routing

The interconnect network must support intelligent routing of HD voice calls through the network. This requires the selection of an HD-enabled path if the incoming call is HD and a non-HD path if the incoming call is from a traditional narrowband device. This can be accomplished within the interconnect network by maintaining a list of all possible routes along with the HD voice status of each route (that is, if a route supports HD voice). This routing capability can determine the most optimal route to a given terminating device. Without this capability, it is difficult to guarantee that an HD voice call will be routed through an HD-enabled network. Routing an HD voice call through a non-HD path will downgrade it to a traditional narrowband call with reduced audio quality. In fact, this capability is a mandatory requirement for interconnect HD voice calls that involve multiple networks.

Intelligent routing also supports the mapping of codecs when negotiating VoIP SIP setup. This means aligning and selecting the same HD codec types at the calling and called parties and performing media pass-through functions, thereby minimizing the need to transcode the call.

Current solutions in the market generally support transcoding and media pass-through capabilities. Intelligent routing becomes critical, however, when the HD voice call involves multiple service provider networks and transits through an interconnect network. A solution based on a second-generation session border controller with advanced features, including centralized routing capabilities, is well positioned to support the HD voice requirements on the interconnect network. So with HD voice, VoIP is no longer just about reducing costs and providing new features like unified communications. It is also about an improved user experience through better voice quality.

HD voice is expected to revolutionize VoIP, both in the service provider and enterprise markets. Clear and life-like reproduction of audio will lead to increased usage and additional use cases for VoIP. The main impediment to mass adoption of HD technology, however, is simply that the various HD voice services that use different HD voice codecs cannot interoperate with each other. A solution that bridges this divide will help create a truly global HD voice service.

Mohan Palat is principal product manager at Sonus Networks (News - Alert).

TMCnet publishes expert commentary on various telecommunications, IT, call center, CRM and other technology-related topics. Are you an expert in one of these fields, and interested in having your perspective published on a site that gets several million unique visitors each month? Get in touch.

Edited by Rich Steeves