There have always been more emotions than facts around the topic of communication quality. Not calling a friend or colleague in a critical situation is always very annoying, as is the failure of the Internet at the time of sending an important letter.
How is communication quality measured?
The measuring mobile complex moves along predetermined routes, which include the main city highways and central streets.
Test calls are made using the ASCOM TEMS Automatic complex to a landline telephone number (call generator on the territory of the RFC), which is also connected to the digital E1 lines of the fixed communication network (PSTN). At the same time, the complex includes 4 identical call devices equipped with SIM cards of four federal operators. Thus, calls are made synchronously at the same time, from the same place, which excludes privileged conditions for any operator.
Voice quality tests record the following parameters:
Voice quality. Includes unsuccessful call attempts, dropped calls, speech intelligibility. Speech quality is measured using POLQA (Perceptual Objective Listening Quality Assessment) technology, approved by the International Telecommunication Union. Each voice call is evaluated for three minutes. The result of calling the “Big Four” is a set of log-files of measuring complexes. Each voice call is evaluated for 3 minutes.
Delivery of SMS (the proportion of messages that did not reach the addressee, as well as the average delivery time). More learn here
Data transfer: percentage of unsuccessful TCP / IP connection to the server (HTTP IP-Service Access Failure Ratio); the proportion of unsuccessful sessions over the HTTP protocol (HTTP Session Failure Ratio); the average value of the data transfer rate to the subscriber (HTTP DL Mean User Data Rate) in kbit / s; duration of a successful session (HTTP Session Time) in ms. The procedure is performed for three standards: GSM (2G), UMTS (3G), and LTE (4G). The file size for HTTP download for GSM and UMTS is 3 MB and for LTE 100 MB.
Overall coverage scan, signal presence, and level (measured in dBm).
The received materials are analyzed, a protocol is formed, after which they are sent to the site of quality communication. The resource then publishes the data.
Measurements cannot be taken during abnormal network load: for example, on holidays, during mass events or major incidents. Also, night measurements cannot be taken.
What is the difference between crowdsourcing
It is believed that specialized services, such as the SpeedTest application, may show more reliable results. After all, measurements in this case are made by the audience of thousands of applications in all environments, at different times and in any premises.
But despite all these advantages, crowdsourcing measurements contain one big disadvantage: their results are very easy to dispute. The operator will always be able to refer to the factors distorting the signal. After all, a simple user, unlike a special laboratory, will not be able to measure a “clean signal” in any way.
Mobile laboratories carry out simultaneous parallel testing in the networks of all telecom operators, using the same type of equipment. Thus, the assessment results are not influenced by factors such as different UE capabilities and different subscriber traffic profiles in time and space.