According to data collected over the last 12 months by Minacom, VoIP phone service now sounds better and connects faster than the standard public-switched phone network (PSTN).
Results from the company's standards-based, single-ended service quality test system show that VoIP service quality increased steadily over the last year, with an average Mean Opinion Score (MOS) of 4.2, compared to 3.9 for the PSTN.
MOS is a scale commonly used to describe speech quality, ranging from 1 (worst) to 5 (best). Minacom reports that, based on a MOS threshold of 3.6, only 1 out of 50 calls in North America were considered to be unacceptable (1 in 10 worldwide), while greater than 85% of VoIP calls exceeded average PSTN quality over the same period.
The company's detailed test results show that VoIP service bettered PSTN quality worldwide, and improved in all regions over the course of the survey. In addition to superior sound quality, calls over VoIP were seen connecting more quickly at an average of 8.2 seconds, compared to 8.9 seconds for those placed over the PSTN.
Regionally, the PSTN was faster to connect for calls placed to North America (4.3 seconds vs. 5.7 for VoIP), while international calls connected faster with VoIP (8.7 vs. 10.4 seconds for PSTN). According to the company, linear regression indicates that VoIP is closing the gap, connecting 2 seconds faster in July 2006 than a year earlier.
"Over the last few years, VoIP has undergone rapid deployment as service providers become more confident that the technology can deliver PSTN-quality service," comments Michel Nadeau, president and CEO of Minacom. "Our test results confirm that VoIP services available today can equal or exceed the quality of traditional PSTN offerings."
Minacom notes that a recent Internet Phone quality study by Brix Networks indicated that 1 in 5 VoIP calls were classified as unacceptable, and that call quality was steadily declining. As this study may have created the impression that VoIP service is not capable of delivering PSTN-grade phone service, Minacom says that it felt that these results should be clarified for both those in the VoIP industry, and individuals and enterprises considering VoIP service.
According to Minacom, the Brix report evaluated computer-to-computer (PC-PC) Internet phone services, similar to those offered by Skype, Google Talk, MSN, and Yahoo Messenger. Minacom contends that the quality and service reliability of these applications does not compare to that of the VoIP phone services offered by telcos, cable operators, and broadband VoIP providers who carefully deploy, monitor, and manage the quality of their services.
Minacom points out that PC-PC VoIP quality is subject to diverse impairments including firewall settings, computer performance, anti-virus installations, high-compression codecs, and Internet bandwidth shared among gaming, file download, web, and e-mail applications. By contrast, the company notes that VoIP as offered by service providers is switched using telecom grade equipment, uses lower-compression codecs, and is prioritized over regular Internet traffic using sophisticated, standards-based multimedia telephone adapters that are maintained and monitored by the operator.
Minacom notes that its tests were conducted over PSTN, managed broadband, and cable VoIP lines, the same services offered to residential and enterprise customers by phone, cable, and hosted VoIP providers.
According to a press release, each month, Minacom's PowerProbe 6000 service level test probe places hundreds of calls from the company's QoS labs in Montreal, Canada to public destinations worldwide over PSTN, broadband VoIP, cable VoIP, DSL, FTTP, and wireless networks. The results of these tests are published in Minacom's QoS Benchmark Reports, a free e-mail newsletter. The results shown in the current study are based on data published in these reports over a one-year period from July 2005 to July 2006.
Minacom's DirectQuality R7 test system uses a combination of ITU and industry-standard algorithms to calculate listening quality MOS using both analog and IP measurements. According to the company, MOS scores based only on IP packet statistics do not capture the effects of echo cancellers in network equipment and telephone adapters, noise introduced by copper wiring, or issues with call volume and delay. The company says its PowerProbe 6000 IVR Test Agent measures a wide range of analog and IP impairments, including noise, echo, delay, packet loss, call volume, jitter and loss, as well as an array of connectivity metrics including Post Dial Delay (PDD), Answer Seizure Ratio (ASR), and Dial-Tone Delay (DTD).
"Carriers are becoming increasingly educated about MOS scoring and want to know where MOS scores are coming from," notes Jessy Cavazos, a telecom industry analyst at Frost and Sullivan. "There are numerous products in the market that only look at the packet metrics. Hence, many carriers are starting to see degradation they should not see, or not seeing degradation they should see. False service quality alarms result in unproductive troubleshooting efforts by service providers, whereas unidentified quality issues ultimately leads to dissatisfied customers. That is why Minacom uses three different technology sources for MOS scoring instead of only one, so as to capture all possible service issues with the highest degree of accuracy available."
"Wireline copper telephony is subject to long, aging wiring to residences, quantization distortion introduced by narrowband PCM codecs, and least-cost routing decisions which can result in quality issues," concludes Minacom's Nadeau. "It is no surprise that VoIP delivered digitally from end-to-end can outperform the PSTN. With the introduction of wideband codecs and the ever-faster Internet backbone, we should expect that VoIP calls will soon be the next best thing to 'being there'."