The Mobiilimittari measurements are active by nature, which means that in order to measure the quality, a specific test traffic is being generated. During the measurement, download and upload speeds are tested using HTTP (Hypertext Transfer Protocol) and, in addition, round trip time is tested. The pure performance metrics gives already a good understanding of fluency e.g. of web browsing since the used transfer protocol is the same. In addition, the measurement metrics are used for calculating estimations of what the quality of certain application types might be, and based on this, the Mobiilimittari application provides you an overall estimation of network connection quality.
The overall connection quality estimation is implemented considering requirements of the commonly used application types, nowadays. Applications, however, evolve all the time, which has resulted into more demanding requirements for the network connection, at least, in the past. If the similar behaviour continues, the overall quality estimation will be updated accordingly. The overall quality estimation is represented in a scale from 0 to 10, and it can be generally interpret as the following:
0: The connection does not work properly (some of / all tests have failed)
1 – 2: The connection is poor and usage of applications requiring the Internet connection is slow and unpleasant.
2 – 4: The quality is not in a good level but it may be usable for some selected applications such as low quality video services and text based social media applications.
4 – 6: The connection is moderate and most of applications (operating over the Internet connection) may be used, although not necessarily with the best possible quality.
7 – 8: The connection is good and almost all services are well usable. The most demanding applications like UHD video services, interactive network games or high quality video conferences may, however, require a connection that is even better.
9 – 10: The connection quality is excellent and service operation should not be limited due network connection capabilities.
While evaluating the measurement results, it is worthwhile to notice that these are affected by a number of factors. First of all, mobile broadband speeds are commonly advertised as the theoretical maximums that are achievable only in the optimal conditions. Therefore, you shouldn't be surprised if your results do not reach the maximum speed. However if your subscription always gives you significantly lower results than the maximum, even in different geographical locations, the matter could perhaps be worth of clarifying. Please notice that your mobile device can have its own restrictions even though the subscription speed would be unlimited. The measurement results itself are affected, at least, by the following: a number of other users in the network and their activity, location, external interference and velocity of your mobile device.
The communication network capacity is always limited. If the network has many users on the same area, the capacity must be divided among these in a way or another. For example, if many users are connected to the same base station, they might degrade capacity of others, and henceforth every users' quality get downgraded. This can cause a significant variation into Quality of Service and quite often almost regular time of day variation. During the night time typically there are only few users, therefore a single user can possibly get a very high capacity to its use, but during the evenings, when many users will use their wireless connections, it can be the opposite (i.e. capacity is very low compared to the maximum). Geographical location affects into the quality also, since received signal strength is dependent of distance to the base station, the terrain and possible obstacles in between. The lower the signal strength is, the lower modulation is forced to be used in radio layer, which will then affect the data speeds. If the signal strength is low enough, bit level errors begin to appear that can be seen as packet losses on the upper telecommunication layers. Finally, this results to slower and more unreliable operation of the networked applications of the users. If the user moves, for instance, Doppler effect will introduce an additional interference. Moreover, mobile device switching from a base station to another while moving around may cause more errors, and therefore can degrade the quality as well. Moreover, external interference sources may have a significant impact on the quality. This is a significant issue in WLAN networks that are operating on unlicensed frequency bands that can be used by anyone. The unlicensed bands are relatively narrow, so in practice, if there are many WLAN networks in the area, it is possible that they interfere each other. Hence, if your home WLAN network would work properly, the WLAN network of a neighbour may cause significant impact on your network.
Due to the previously described facts, wireless telecommunication has a lot of natural uncertainties. Therefore, the conclusion of connection performance should not be drawn based on a single or even couple of measurements. These give an overview on the network functioning at a given time and location. In order to get a more sophisticated view on your connection functioning and quality, it is advised to perform the measurements in different time of day and in multiple locations. It is clear however, that if you have ordered a mobile broadband subscription for your home with some sort of quality guarantee, then the connection should operate with the promised quality on the area.
The Mobiilimittari service is intented to be used in Finland. The background measurement servers are located within Finland, behind the high bandwidth Internet connection, and therefore measurements performed within Finland gives the most accurate results. A network connection is always an entirety including the end user device, an operator network, Internet path and finally the target destination, which is, in this case, one of the Mobiilimittari measurement servers. Also the Internet path sets its own limitations on that end-to-end connection. If the user is far away from Finland, this can restrict the connection remarkably. With fast LTE networks, the Internet path can already have a significant impact on the overall quality. So, you are allowed to perform measurements also abroad, and there are no technical obstacles in it, but please remember that longer the distance is to Finland, the results can become more unreliable.
The Mobiilimittari measurement results contain a significant amount of statistical information that may be used, for example, in network performance surveys or in identifying a problem in the network. More advanced users have a chance to download a limited amount of the more detailed measurement results (of the latest own results) in CSV format. If you are interested in professional usage of the measurement service and the detailed results, please contact us at email@example.com.
VTT has a long history with Quality of Service measurements, and the work on that has been started back in 2002. The Mobiilimittari service, and its professional version Moset, are active Quality of Service measurement tools but VTT has other measurement tools such as a passive Quality of Service measurement tool Qosmet. In addition, VTT performs tailored measurement campaigns for its customers. Further information on VTT's CNL group solutions and results can be found here.