Changing support requirements for the wireless end user.
The service requirements remain the same; ensure the same high level of service for voice quality that the subscriber has always enjoyed. It should be recognized that "voice quality" refers to a rather subjective measurement referring to the sound quality in a telephone conversation. Because the telephone network was built primarily for voice applications, the emphasis has always been on the transmission of real time voice signals across the network. However, in recent time, we have seen the convergence of voice and data over the public network has changed forever the way in which quality of voice traffic is controlled.
In order to capture and maintain market share, service providers and their equipment providers must now take additional steps to ensure that while they can support and integrated network they will be able to support the high level of voice traffic the end user has continued to expect. End users--particularly the business end users--expect nothing less. This, in turn, has forced service providers and their supporting vendors to continue their ongoing search for effective tools, which enable them to evaluate the voice quality of their networks.
That Which Affects Voice Quality
The PSTN network is engineered to accommodate the human voice, and the user's perception of the quality of a telephone conversation depends upon the real-time transmission of the voice signals. The PSTN channels are specifically designed to address those factors, which effect quality: delay (including jitter), echo, and clarity.
Delay or Latency: This is the amount of time a voice signal takes to travel from one caller to another caller in a telephone conversation. This includes the time required to convert the signal from analog to digital to analog, the time required by switching equipment to buffer and route the signal, and the time to travel across a specific geographical distance. When a packet network is involved, a delay is involved to divide a signal into multiple packets for transportation across the network. Another form of delay is jitter, which refers to variations in delay caused by fluctuating signal strengths.
Echo: Echo is exactly that and is produced by the crossing of (electrical) voice paths, which happens when a signal moves from one type of network to another. The result of this situation is the reflection of the speaker's voice to the speaker's ear and is made more pronounced by excessive delay.
Clarity: Clarity is fidelity of the signal itself, which can affect voice quality as well as numerous other factors, which contribute to clarity. For example, the network across which a voice signal must travel actually consists of multiple carrier networks. The quality of the circuit, which affects the clarity of that voice signal, can vary from provider to provider. Another condition that can impair clarity is transcoding, which involves the use of multiple compression algorithms on a voice signal. This compression of an already compressed voice signal degrades voice clarity.
Testing for Voice Quality
Voice quality is a subjective measurement and depends upon individual perceptions of what is an acceptable or unacceptable level of voice quality. Traditionally, voice quality was always tested using large panels of human users to derive a "mean opinion." These measures were later dropped and, in the mid-1990s, standards were developed to automate the testing and measurement of end-to-end voice quality on both circuit and packet switched networks. Three algorithms have emerged thus far for automated and repeatable testing.
PAMS: Perceptual Anaysis/Measurement System predicts overall subjective listening quality. Originally developed by British Telecom, PAMS measures the voice signal in terms of both listening effort and listening quality. It compares two signals: an original unprocessed signal and a version of the signal, which has been degraded. PAMS analyzes the amount of different types of errors found I the degraded version and predicts a Mean Opinion Score (MOS). The degraded signal can receive an MOS of 1-5, with 5 denoting the best score.
PSQM: Perceptual Speech Quality Measurement predicts subjective quality of speech codecs. KPN research in the Netherlands developed PSQM to measure the objective quality of codecs. PSQM compares an original signal with a degraded version to measure the latter's distortion, noise and fidelity. The ITU has also approved an enhanced version of the algorithm, PSQM+, which manages large transient distortions more effectively. Neither version copes adequately with filtering, variable delay or short localized distortions.
PESQ: Perceptual Evaluation of Speech Quality is an objective measurement of subjective listening tests on telephony systems. A combined effort of KPN-BT in 2001, it compares a reference and a degraded signal to produce a quality score analogous to the MOS produced by a panel of human listeners. The PESQ incorporates some capabilities that make it a more powerful algorithm, which tests the effects of network elements such as codecs, but also measures end-to-end speech quality. It also considers various sources of signal degradation to include coding distortions, errors, packet loss, delay, jitter and filtering. PESQ test results are measured in the range of pf -0.5 to 4.5. The normal range in MOS values found in listening quality tests will be from 1 to 4.5, with a score of 1 as the lowest and 4.5 the highest.
Voice Quality Testing
Given the competitive importance of predictable voice quality in the next generation, networks carriers and their equipment suppliers are turning to test equipment service suppliers for solutions that can help them deliver the high quality of service that today's customers demand. Specifically, they are seeking voice quality testing tools that will provide the following levels of support:
* Execute simultaneously all standard testing algorithms
* Provide the capability for customizing to fit individual network and testing requirements
* Are easy to use and deploy
* Are portable for supporting technicians in the field
* Are modular in design, easily scaled to network and traffic growth patterns
* Offer remote access capabilities
There are a number of vendors that have developed system tools to support various requirements as defined above. However, three vendors that have developed systems that support the requirements as required are Acterna Systems, Empirix Systems, and GL Communications who produces out of the box solutions aimed at supporting an array of test solutions. These solutions are based on a familiar user interface that can transform PC-based tools into cost-effective testing tools. GL Communications also supports a number of portable versions to support field testing and remote access for field testing.
A good example of voice quality testing systems can be seen in a typical GL Communications system comprised of PC-based software, which is capable of executing PAMS, PSQM/PSM+, and the PESQ algorithms which are capable of comparing a test signal and a degraded version that has been passed through a distorting system. Technicians are then able to run random voice quality checks manually or in automatic mode to perform a series of random checks over a user-defined time period.
These voice quality test results can appear in both tabular and graphical formats. For example, when running in automatic mode the system can produce a customized rolling graph which shows scores on one axis and the elapsed time on another. This allows the technician to obtain both historical and real-time test results from all three algorithms over a user-defined time frame.
Another customized graph can be developed which is bar form and contains the number of test results on one axis and the scores on the other. Using this graph, a technician can determine very easily how many tests, for instance, scored between 2 and 3, or between 3 and 4.
Three other components of this system function as individual "engines" depending upon a service provider's particular testing requirements. There is the T/1E/1 and T3 Analyzer, the Digital Central Office Switch Simulator & Bulk Call Generator (DCOSS), and the Analog Test Set (ATS). A standard element of the ATS, the Voice Quality Testing (VQT) application is available as an option with either the Analyzer or the DCOSS systems.
The Analyzer Solution
An enhanced third-generation enhancement, the ULTRA T1/E1 PCI card consolidates the essential elements of industry standard test equipment into a powerful PC-based T1/E1 solution. This element is used to monitor in-service T/1E/1 circuits, including 64Kbps channels within these connections performing routine tests for errors, signal characteristics and timing. Added to this is an optional client/server application that allows a service provider to control it remotely. When used with VQT software, the Analyzer solution is capable of tapping into a circuit and record voice files, allowing the VQT to capture the voice file and apply the three ITU algorithms to it.
The DCOSS Solution
This option is used to simulate and test advanced telecom networks and related products, including switches, gateways and transmission systems. The PC-based DCOSS system basically functions as a central office, placing and answering calls. It can be used to simulate one or both ends of a complex switching network with a variety of internal and external traffic, i.e., faxes, modem calls, voice files and call verify T1/E1 signaling protocols such as RI, PRI, ISDN, GR 303, SS7, and SS5. A service provider has the option of running VQT software with the DCOSS solution to assess voice quality on their network. Running in manual or automatic mode, the DCOSS generates multiple voice files, which in turn are recorded for simultaneous PAMS, PSQM/PSQM+, and PESQ processing by the VQT software. Like the Analyzer, the DCOSS solution also offers a remote access capability. A service provider can, via a GUI-based client interface and the Internet, execute from any place in the world all of the DCOSS functions required to complete a test series.
The ATS Solution
The ATS solution, using VQT software, provides the analog or optional cellular interface, together with the voice quality measurement and analysis tools that service providers need to ensure clear voice signals over any type of network. Typical applications of the ATS solution include VoIP PSTN, ATM, Frame Relay, and wireless networks. A service provider can connect the ATS solution which supports two analog ports with Rj-11 interfaces to any U.S. or European PSTN. With user-friendly GUIs, technicians can place calls to any telephone number as well as answer incoming calls. They also can configure it to send and record voice files manually or automatically for use in the VQT algorithms.
Assuring Voice Quality
With changes taking place in the advancing number of wireless users, there is an ongoing requirement to maintain a high degree of service on present-day and next-generation networks. This can only be accomplished through a high level of network maintenance across all networks to ensure an ongoing level of voice service. To maintain this high quality voice service, the industry will require easy to use testing services that deliver the degree of testing services required at all times. Whether they need a specific type of service or a variety of service solutions, what will be paramount will be the need for service and equipment providers to remain up to date with the technology.
In this system description, we outlined some of the features of the DCOSS monitor for simulating and testing telecom networks and products. The DCOSS system is a comprehensive telecom testing solution that can be deployed locally and remotely, making it a good solution for the service providers, carriers, testing and development labs, as well as testing local networks. We have seen how, over the years, the wireless mode of operation has grown. With this growth will come an ongoing demand for service, which in turn will require continual monitoring by all parties concerned. This brief outline is one level of service that we can expect to see more of in the future.
Andy Llana is a consultant with Vermont Studies Group Inc., a technology consulting group. He has contributed articles on various technology issues for WestWorld publications in the past.
|Printer friendly Cite/link Email Feedback|
|Publication:||Computer Technology Review|
|Date:||Apr 1, 2004|
|Previous Article:||Fast lane to China: companies that never thought of doing business overseas are now looking to the thriving Chinese economy.|
|Next Article:||Why WiFi? Productivity and convenience.|