To make practical and economic use of the advantages of test connections, a series of automatic, computer-controlled mechanisms is required.
Flexible measurement sets make kyago a universal and future-proof tool. Routine works should not affect the user. The specialist should only be brought in in cases where professional knowledge and experience are really needed. kyago has taken account of these requirements in the concept.
Further details on the technical structure and quality parameters are described in the kyago Whitepaper.
The days of ISDN and classic telephony are numbered, as telecommunications providers are gradually converting all their networks to All-IP.
All-IP refers to the conversion of the transmission technologies previously used in telecommunications networks to a uniform IP-based system. The abbreviation IP stands for "Internet Protocol" and describes a communication technology on the Internet with which information can be transmitted digitally and packet-switched.
All-IP now enables telecommunications providers to provide customers with all services such as telephony, fax, TV, Internet and mobile telephony in a single network, the so-called Next Generation Network (NGN). For this purpose, an Internet connection is completely sufficient. A separate ISDN connection is no longer required.
Against the background of this development, the question of the experienced service quality is becoming increasingly important for All-IP.
Voice under the microscope
The end-to-end speech quality measurements of the kyago platform can be performed between measurement systems at various locations. To this end, connections are set up via end customer interfaces.
To determine the end-to-end voice quality of the respective connections, standard ITU-T voice samples of male and female voices are repeatedly transmitted in both communication directions and compared with the originals in the subsequent automated assessment process. The evaluation of speech quality is based on ITU-T standards.
As a further QoE aspect relevant for end customers, voice quality measurements can be performed at the same time as bandwidth utilisation of the respective product. For this, parallel to the voice quality measurement, larger data amounts are transferred from and to data reference systems and IPTV measurements initiated.
Highspeed or lame duck?
In today's age of digitalization, fast Internet is the basis for the business success of companies in many industries. This is why the most important providers are eagerly pushing ahead with the expansion of networks, the laying of fiber optic cables and the increase in Internet speed.
However, the information provided by the providers on speed must often be viewed as theoretical values in practice. The actual speed of a broadband connection often turns out to be much lower than booked. Above all, cable and DSL connections are always below the theoretically possible maximum when many users call up large amounts of data at the same time, for example through video streaming.
High-speed Internet under the microscope
To evaluate the quality of high speed Internet, various measurements can be performed with the kyago platform. The available upstream and downstream bandwidth is determined by standardised upload and download measurements which are made to data reference systems. These measurements are also carried out at the same time as download or upload in order to determine the behaviour of the products under simultaneous utilisation of the bandwidth. As further QoE indicators, among other things, packet times, the frequency of packet losses, and the time for DNS name resolution are measured.
All Cloud or what?
Even with large numbers of visitors and temporary sales promotions, customers expect optimal performance. Unsatisfied customers are caused by long loading times, sluggish music and video clips or unreachable websites.
Average users wait no longer than four seconds for a website to appear in full. 90 percent of Internet customers turn their backs on an online shop after three failed access attempts at the latest.
Cloud under the microscope
Even with large numbers of visitors and temporary sales promotions, customers expect optimal performance. Dissatisfied customers are the result of long load times, faltering music and video clips or unreachable sites.
The kyago platform measures both performance-based QoS measurements such as the DNS resolution time and response times for gaming servers and user-oriented QoE measurements such as Kepler, website loading times and upload measurements for various photobook providers.
The use of DNS, as one of the most frequently used Internet services, is usually not noticed by the user. However, because this service is fundamental for interaction and a runtime delay affects a high number of Internet services, this measurement is carried out as an independent test.
A number of online games require above-average runtime performance to enable players to compete with other players on an equal footing in interactive games.
In the user-oriented QoE measurements, a website is accessed via a browser. Standardised test pages (ETSI Kepler reference page) from national and international webhosting-providers are accessed and measured at the transport layer.
In addition, various, frequently used websites are also measured at the application layer. The two leading browsers by market share in Germany (Firefox and Chrome) are used to determine the quality parameters. Key elements of the trigger points standardised according to W3C "Navigation Timing" are used to determine the measured values of website calls.
Global video traffic presents a major challenge for Internet service providers if they want to provide excellent quality of service to their customers.
They not only have to ensure quality of service in their own network, but also pay particular attention to the video content feed via various peering and content delivery networks (CDN).
WebTV and streaming under the microscope
The streaming test carries out a measurement using the over-the-top approach for different video content providers. A Chrome web browser with the appropriate plug-ins needed for all the major components of the video being tested, is used to access the videos.
The streaming services measured use adaptive streaming. Depending on the provider, different streaming approaches are offered using MPEG-DASH, HLS or progressive download (HTML5 video), among others. Video containers with different video codecs such as H.264, HEVC/H.265, VP9 or AV1 and audio codecs such as MP3, AAC or Opus are used.
Quality analysis is carried out according to the ITU-T Rec. P.1203 standard, which supports HEVC/H.265, VP9 and AV1 as well as 4K/UHD-1 resolutions and frame rates of up to 60 FPS in addition to H.264 with extensions.
As part of the YouTube (YT) video response time measurements, ten videos are retrieved every hour in a Google Chrome browser and the time it takes for the first video segment to arrive from the content server is measured.
In the age of modern network infrastructures and convergence to the Next Generation Network (NGN) or ALL-IP network, well-known services such as linear television, which has previously been aired by the TV stations on cable, satellite or terrestrial via broadcast, are now being ported via new transmission media.
The most common designation for television which is broadcast or received via the Internet is “Internet Protocol Television” (IPTV). IPTV generally designates the digital transmission of audio and video signals via a broadband Internet connection.
Through the broadband Internet connection, genuine interaction with the consumer is possible, through which new applications and services such as video-on-demand (VoD) can be realised.
IPTV under the microscope
In the increasingly complex world of next-generation video services, knowledge about the quality of service is a crucial factor in the fight for market share and customer satisfaction. With the IPTV measurements, measurements are carried out using set-top boxes supplied by the providers. For control of the test cases on the set-top box, primary control is via Bluetooth Low Energy (BLE).
The test cases mirror real scenarios of a user. After the set-top box is switched on, a total of four different channels are measured and the video quality of these is evaluated.
The IPTV quality benchmarks will also be measured with parallel HTTP upload and download, parallel telephony connections, and a second IPTV stream to exhaust the full bandwidth.
The subjective and objective quality of the media data is derived from the current IP data stream (live and non-reference measurements) and is based on the analysis of network parameters and the quality of service (QoS) as well as on an analysis of the video codec layers with the help of “deep packet inspection”.
Where possible, the measurements will be made using adaptive streaming. Depending on the IPTV provider, different streaming approaches are offered via HTML5 MPEG DASH or HLS. Typically, the MP4 video container is used with H.264 coding and MP3, AAC or OGG audio codec.
The quality analysis is based on OPTICOM’s latest Perceptual Evaluation of Streaming Video Quality (PEVQ-S) algorithm, which now supports VP9 and UHD resolutions in addition to H.265.
The test catalogue also includes advanced zapping, where different zapping scenarios are considered from a customer perspective and times are recorded.