Karl Brown is the Vice President of Marketing at Ditech Networks. Mr Brown has over 15 years of experience in network communications, spanning wireline and wireless, voice and data, as well as equipment vendor and service provider environments. Prior to Ditech Networks, Mr Brown held marketing positions at ANDA Networks, Jetstream Communications, and Nortel Networks. Karl Brown earned Bachelor and Master of Science degrees in Engineering from the Georgia Institute of Technology as well as an MBA from Indiana University.
Even though 3G and 4G networks are deployed to provide high-speed data connectivity for emerging mobile applications, even a sophisticated multi-function phone is still a phone; users expect high quality voice service. The most efficient way to achieve this is by deploying a Voice Enhancement Device capable of mitigating internal and external voice quality impairments. These devices enable operators to unobtrusively monitor live calls on the network and aggressively deploy low bit-rate codecs to increase call capacity, while maintaining acceptable voice quality levels.
Although the driving force behind deployments is higher data speed to support emerging applications, the 3G and 4G infrastructures will still carry a significant amount of voice traffic. Voice traffic, in the form of data, will compete for network resources along with videos, music, email, text messages and map services. However, this will not change the critical importance of voice quality, and the need to maintain consistency and reliability no matter what technology is in use. Mobile subscribers consistently rank call quality as one of the most important criteria when choosing a service provider. The single biggest reason subscribers have mobile phones is still to make phone calls. Voice quality is especially challenging for mobile communications because subscribers make calls from a disparate locations, using a remarkable variety of devices, across multiple networks designed for different purposes. As a result, about 39 per cent of all mobile calls around the world are rated so poor that they are likely to cause churn. While traditional IP quality of service (QoS) provisions are necessary in 3G and 4G networks, these alone are not sufficient to ensure voice quality. The voice quality challenge will continue as new technologies emerge, including WiMAX, FMC/Unlicensed Mobile Access (UMA), IMS, UMTS, Transcoder Free Operation (TrFO) and others. Fortunately, there are cost-effective ways for carriers to measure and maintain acceptable levels of voice quality during the migration to IP and beyond. Measuring voice quality One of the most widely used methods to assess voice quality involves subjective testing, asking people to rate the quality of recorded samples. To standardize this critically important aspect of telephony, the International Telecommunications Union has developed a number of recommendations for subjective testing, including ITU-T P.800 and ITU-T P.835. However subjective testing is expensive and difficult to perform on live calls, making it unsuitable for large-scale testing and continuous monitoring. For many years, mobile carriers have used drive tests and similar methods to measure signal strength and other radio network-related voice quality impairments, but these techniques failed to take into account external conditions, such as ambient noise and acoustic echo, that degrade voice quality. To overcome these limitations, an industry standard was created several years ago that takes into consideration degradation that originates in the network, as well as voice quality problems that occur outside of the network. Called the ITU-T G.107 E-Model, this recommendation has the advantage of taking into account codec (coder-decoder or compression/decompression) type, frame loss, background noise level, mismatched speech levels, acoustic echo and transmission delay when calculating call quality. With accurate, continuous, voice quality measurements, carriers are better prepared to make the investments needed to guarantee quality, and to monitor the effectiveness of their measures. Voice quality enhancement There are numerous ways to enhance VoIP call quality in 3G/4G networks. Some handsets, for example, have basic noise suppression and acoustic echo cancellation built in. The performance of these solutions varies considerably, however, depending upon the handset. In addition, handsets have limitations in processing power and battery life that make it difficult to implement the most advanced noise suppression and echo cancellation algorithms. Another approach is to build noise reduction into the codec itself. For example, the CDMA standard has mandatory noise reduction integrated into its Enhanced Variable Rate Codec (EVRC). Supplementing this approach with echo cancellation systems and pervasive IP QoS provisions to minimize packet loss still falls short of addressing all voice quality impairments. Of course, carriers can always improve the quality of the radio network by deploying more base station sites and more transceivers. This approach, though, can lead to costly over-engineering, and fails to address the external impairments that are the root cause of an average of 39 per cent of all voice quality problems. The biggest problem with this traditional piecemeal approach to voice quality enhancement is that it makes it difficult, if not impossible, for the carrier to continuously optimize call quality and capacity end-to-end in the network. Something better is needed. Voice enhancement and quality Voice enhancement devices, or VEDs, provide a more comprehensive approach to reducing voice quality impairments in GSM and 3G/4G networks. VEDs integrate multiple voice quality enhancement features, including adaptive noise cancellation, acoustic and hybrid echo cancellation, and automatic level control, in a single, scalable system. Some VEDs even compensate for inevitable packet loss, and a few offer transcoding among multiple codec types. The basic VED removes noise and echo, and matches levels on the uplink. The more advanced systems have the ability to do this on the downlink, as well. The ability to process downlink speech also allows the VED to enhance off-net calls that originate in other mobile carriers’ networks. More advanced VEDs have additional functions, such as intelligent packet restoration, to compensate for inevitable packet loss. Some of these advanced systems may even have features like enhanced voice intelligibility that process and enhance the downlink speech path to improve voice clarity for callers in noisy environments. Some of the more advanced VEDs also integrate support for the ITU-T G.107 E-Model, thereby creating a system that both measures and maintains voice quality continuously and automatically. When combined with existing drive tests and consumer surveys, these Quality of Experience measurements can establish performance indicators that track the effect of impairments on delivered voice quality over time. The most advanced VED systems even offer transcoding among a wide range of compressed and uncompressed codec types. The rapid increase in calls that traverse IP networks has increased the deployment of codecs, making interconnecting calls among service providers a significant challenge. Today, the most common method of interconnection is to send the VoIP call to the PSTN (public switched telephone network), then back again to the originating IP network. This approach is both inefficient and expensive, because it requires additional media gateway ports and associated operational expenses. Use of the PSTN for transcoding also causes additional delay in the call path, which exacerbates echo problems in most networks. By deploying transcoding on the VED platform, network operators can reduce the required equipment session capacity for supporting VoIP services by up to 50 per cent. VEDs can be deployed either in the network’s core or at the border. When deployed at the border, a VED with transcoding capabilities can convert codecs used on the access and peering links to a commonly used codec (so-called codec normalization) in the core network. This protects the carrier’s core network from a proliferation of codecs, and makes the deployment of additional core voice services less costly. Voice quality and churn The ability to overcome all major voice quality impairments, while concurrently performing all transcoding, is what gives the VED its primary advantage: churn reduction. Satisfying these requirements on a single platform with a single management system keeps both capital and operational expenditures at a minimum. Another key benefit of implementing a VED is its ability to substantially improve the overall voice quality for all calls. This has the effect of reducing the percentage of calls that fall below the minimum allowable quality. This, in turn, moves most calls out of the ‘churn zone’ of a typical GSM network or that of a 3G/4G network with VoIP communications. The diagram below depicts this effect graphically by showing how the mitigation of both internal and external impairments minimizes the percentage of calls that experience unacceptably poor voice quality, thereby raising most calls to an acceptable quality target range. The better voice quality provided by a VED moves most calls out of the “churn zone” that exists below the Minimum Allowable Quality level. The diagram shows how traditional voice quality enhancements techniques can achieve the same reduction in unacceptable calls, but only with a significant increase in Capex and Opex by over-engineering the network. Another benefit of the VED is its ability to support low bit-rate codecs with acceptable levels of voice quality, which enables carriers to increase call capacity on the existing infrastructure. Subjective testing by Dynastat, a leading listening test lab in the United States, proved that the VED was able to substantially improve voice quality even for low bit-rate codecs in noisy environments. The final major advantage of the VED is its superior scalability over piecemeal solutions. Voice quality is still king for subscribers on 3G and 4G networks. By deploying a Voice Enhancement Device capable of mitigating internal and external voice quality impairments, carriers will be fully equipped to maintain consistent and acceptable levels of voice quality, and to reduce the cost of churn associated with dissatisfied subscribers. VEDs also enable better test methodologies using ITU-T G.107 E-Model to unobtrusively monitor the customer experience for all live calls on the network. VEDs enable carriers to aggressively deploy low bit-rate codecs to increase call capacity, while maintaining acceptable voice quality levels. This, coupled with their capacity for high-performance transcoding, makes VEDs a critical element in the architecture of all 3G and 4G networks.