Radio laboratory “MegaFon” – what is tested in devices and how

Hello.

Having started discussing the peculiarities and subtleties of the MegaFon laboratory in St. Petersburg with its head Alexander Dzhakonia, I collected material for several articles. You can find the beginning of our conversation in the previous article.

To understand what exactly the laboratory does, let’s dwell on what exactly it tests and how. For ease of perception, I built this text exactly as we said, in an interview format, although it was rather an ordinary conversation.

EM. In one of the comments to the previous article, there was a phrase that your laboratory is only an image for the operator and nothing else. And you cannot or don’t want to make the phone work in the network in any conditions. Parry?

HELL. For the operator, the correct operation of the device in the network is in many ways an image story. In theory, you can make your phone stick to the network in the most difficult conditions, but the battery will drain three times faster. It is important for us to correctly configure not only phones and smartphones, but also the network itself – to give good coverage, to adjust the switching of the terminal between base stations when you move around the city.

EM. People often say they want VoWiFi, VoLTE. Are you testing these features, is this a mandatory requirement for your network?

HELL. Every three months we release updated operator requirements for mobile equipment. Together with our colleagues from MegaFon’s infrastructure development, we are creating a consolidated document. It is very interesting to observe him in retrospect, what we demanded three years ago and what we demand now. For example, now all smartphones of any price category must support VoLTE. In the laboratory, we test the performance of both VoLTE and VoWiFi, helping vendors to properly configure these functions.

One of the main trends in MegaFon is the transition to VoLTE, we want the maximum number of people to use this service. Interestingly, many people for some reason believe that this is a paid service. We even turned to the manufacturers of smartphones so that in the menu, where VoLTE was turned on, there was a small line that this service is free in the MegaFon network. Unfortunately, the release of a new firmware is always an additional cost, and manufacturers were not enthusiastic about our request.

EM. Okay, you are testing a device, VoLTE does not work in it, what happens next? Do you not let it be sold in your retail?

HELL. There is an interweaving of both commercial and technical issues. We invite manufacturers to us and help to correct the shortcomings. Right now, in our laboratory, there are colleagues from a large Chinese company. Correction, verification and debugging happen faster when both our engineers and the manufacturer’s engineers work at the same stand, this speeds up the process. Sending logs to the manufacturer’s headquarters, feedback – all this is very long, and when everyone is in the same room, corrections take place before our eyes.

The sooner the device goes through the lab and hits store shelves, the better. It is important for us that there are no congestions, we do not have infinite throughput, the laboratory must work efficiently.

A failed test program for a particular device does not mean that it is given up. There can be two or three such iterations, and with the joint interest of MegaFon and the manufacturer, we can bring almost any device to the correct state.

EM. Manufacturers are very different – someone wants to bring the device to condition, someone is trying to make the minimum. How does working with manufacturers differ?

HELL. The more famous the company is, the easier it is to negotiate with it. We speak the same language with large manufacturers. Small companies are not so free in financial resources, sending an engineer to us is the cost of a business trip to another country. It is also money to release new firmware with fixes.

EM. From the conversation, it seems that you are considering the devices in general, but this is not entirely true? What exactly do you test in the laboratory?

HELL. We are looking at a modem and a chipset-antenna bundle, and we also test various user scenarios. For example, that all indications are working correctly. It is clear that the testing program is different for a smartphone and a router. The same test can run differently for different devices. For example, you have a router nailed to a post in your summer cottage, which is suitable for outdoor use. It is not so important how this router can switch between base stations, since it does not move and is fixed in one place. On the other hand, this is important for a smartphone.

We have the same speed tests for different types of devices, we are looking at the same indicators. One way or another, the speed of the device in the network is determined by the category – Cat.4, Cat.6, and so on. If the manufacturer claims that it supports speeds, for example, up to 150 Mbps, then this speed should be in the device.

Sometimes there is controversy with manufacturers when they say that on a real network the maximum speed declared in the same Cat.6 cannot be achieved and therefore there is no point in looking at it. They are partially right, but if in the laboratory we see that under ideal conditions there is no declared speed, then this means that something is wrong somewhere, and the task arises to find this incorrectness. In a commercial network, such a device will also behave incorrectly, our experience suggests this. We cannot name the absolute value of the speed with which such a device will operate in a loaded network, in a bad signal. But we can say for sure that in the operator’s network such a device will work poorly, worse than it should.

EM. Look, Alexander, I have an inexpensive router with a weak processor, the modem is declared fast. Receiving a stream of information from the network at high speed, the device does not have time to store and distribute it, it literally chokes. It turns out that you are not catching this moment in any way?

HELL. We also test Wi-Fi. It makes no sense when the router receives 300 Mbps from the LTE network, but does not have the ability to distribute to several devices at maximum speed. And this is wrong.

We have a separate Wi-Fi network, separate Wi-Fi points, and we check the quality of the connection, including for Wi-Fi. To check the operation of the equipment, we load them during the day, at the same time you can see how the battery of the device works, heating occurs and how it affects the speed of the modem. Multilevel verification, the results of which can be used to see what happens to the device under real load. We always try to create conditions that are as close to real as possible. Simulating in a lab is good, but you need to recreate real-world conditions. Many manufacturers do not conduct such complex tests, do not drive their devices in such modes. And this allows you to catch more problems that you can then fix. I repeat that our job is to make the devices as good as possible and thereby improve the customer experience of MegaFon subscribers.

P.S. I still have about an hour’s conversation with Alexander in my store, I will isolate all the most interesting from it and share with you in a separate text. This is a rare case when you can look into how a radio laboratory works and what the people in it are guided by, what they check in the devices.


Source: Mobile-review.com — Все о мобильной технике и технологиях by mobile-review.com.

*The article has been translated based on the content of Mobile-review.com — Все о мобильной технике и технологиях by mobile-review.com. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!

*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.

*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!