farq.uz
Blog · Mobile

How to check an operator's coverage and speed yourself

B
Farq.uz research
September 8, 2026 · 16 min

Mobile plans

See all

0+ · 34 offers

Tekin
0
CDMA-180
6 300
Obod
6 300
Mobile

There is no single number that answers for your flat

Mobile speed depends on the address, the floor, the walls, the time of day, the phone, the tariff and how many people are on the same cell right now. Which is why any nationwide figure says nothing about one particular room, and why the check takes a week and is done with your own hands. Below: what to measure, how coverage differs from speed, and how to try an operator out before you move your number.

The question almost always comes out as “who has the better network”. Put that way it cannot be answered: a network is not one number but how it behaves at your four to six places. Home, work, the route between them and the spot you visit every week — that is the entire sample that matters.

The good news is that the check needs nothing but a phone and patience. The bad news is that one measurement is not enough, and that is exactly where every home-made comparison falls apart.

what 1080p video asks for, per YouTube
5Mbps
what 720p video asks for
2.5Mbps
what a one-to-one 720p Zoom call asks for
1.2Mbps

Coverage and speed are different things

Coverage

Whether there is a signal at this spot and whether it holds. The test is not the bars on the screen but whether the call connects and whether it survives when you move deeper into the room.

Speed

How much data gets through once the signal is there. It swings across the day far more than coverage does: in the evening the same spot can measure several times lower than at noon.

Latency and stability

How fast the reply comes back and how often the connection stumbles. For calls and video meetings this matters more than speed: a conversation breaks up from delay, not from a shortage of megabits.

Mixing these three up is what starts the arguments. “I get 80 Mbps” is about speed at one spot in one minute. “It does not work in the lift” is about coverage. “Video stalls even though the speed test shows plenty” is usually about stability and latency.

The practical conclusion: measure all three, and in the same places. Speed without coverage is useless; coverage without speed is a connection that will not open a page.

How to measure

The check takes a week

  1. 1Pick 4–6 places

    At home by the window and deep inside the room, at work or college, along your usual route. A measurement on the street says nothing about what happens indoors.

  2. 2Measure three times a day

    Morning, afternoon and evening. In a residential area the evening figure is usually the lowest — and it is the one that decides whether video plays.

  3. 3Log the failures, not just the speed

    A dropped call, a stalled video, a page that would not open. One such entry says more about a network than three successful measurements.

  4. 4Repeat for a week

    A single day can land on an outage, roadworks or a holiday. A week shows how the network behaves rather than one incident.

  5. 5Compare the median and the failure count

    Not the maximum. A network with a steady result beats a network with a record and evening collapses.

One test means nothing

A measurement captures one spot in one minute at one level of load. An hour later the same place can read several times differently, and the next room again. So a single screenshot with a big number is no argument for an operator or against one. The argument is a table covering several places over a week, showing both the typical result and how many times the connection let you down.

What to log at each place
What to logHowWhy
Place and timeSpot, floor, position in the room, hourWithout this the measurements cannot be compared with each other
Download speedThe same testing service, the same serverChanging the server moves the result more than changing the operator
Upload speedThe same testVideo calls and sending files depend on it
LatencyThe ping value in the same testCalls and video meetings break up from delay, not from a shortage of megabits
Network typeThe icon on the screen: 4G or 5GIt tells you what you have actually measured
FailuresA dropped call, a stalled video, a page that would not openThe only thing that describes the experience rather than a number

One row per place and time of day. After a week the table answers the question better than any ranking.

What to hold equal before comparing operators

  • The same phone: different handsets hold a network differently even with the same SIM
  • The same place and height: one step towards the window moves the result more than most other conditions
  • One testing service and one server — otherwise you compare routes, not networks
  • The same network generation: 4G cannot be compared with 5G
  • A package with data left on both SIMs: once it runs out, paid charging begins, and that is not a slow network
  • VPN switched off: it changes the route of the traffic and, on some tariffs, how it is charged

What speed you actually need

Before comparing measurements it is worth knowing what to compare them against. Services publish recommended speeds, and the figures turn out to be unexpectedly modest: 1080p video runs on five megabits a second, a video call on barely more than one. Everything above that is headroom, not necessity.

The speeds the services themselves name for video, calls and music
Even high-quality video asks for less than an ordinary speed test shows
Recommended speed by task
TaskRecommended speed
YouTube, 360p0.7 Mbps
YouTube, 720p2.5 Mbps
YouTube, 1080p5 Mbps
YouTube, 4K20 Mbps
Zoom video call, 720p1.2 Mbps
Zoom video call, 1080p3.8 up, 3.0 down
Zoom, screen sharing50–75 kbps
Zoom with the camera off60–80 kbps

1080p video runs on five megabits a second. Anything above that in a speed test is headroom, not a difference you will feel.

Separate out what the tariff is doing

Half of the “difference between operators” turns out to be a difference between tariffs, and it is written into the price lists.

Access to 5G is often a separate line in the tariff: Ucell Bor 110 adds 100 GB “on the 5G network” to its 200 GB, and Bor 160 VIP adds 150 GB to 350. Perfectum puts 5G into a tariff of its own, A'lo 5G, quoting speeds of up to 1 Gbps, and publishes its own measurements on the operator's site: 980 Mbps on 5G and 90 Mbps on LTE. The 5G icon on your screen promises neither figure — it only says which network the phone is attached to.

The second thing mistaken for a bad network is an exhausted package. Under the terms of these tariffs the speed is not throttled once the limit is passed: standard charging begins, 100 UZS a megabyte on Mobiuz and 25 on Ucell. So slow internet at the end of the month is sometimes not the network but the bill. Working out your own volume is covered separately: how many gigabytes a month you need.

The advertised network speed next to the speed a task actually asks for
The advertised speed and the speed a task needs are an order of magnitude apart

How to test an operator before switching

A trial run without losing your number

01

Take a separate number for the trial

A second SIM or an eSIM in the same phone gives an honest comparison: one device, the same places, the same times. Your main number stays where it is.

02

Connect a comparable tariff

Comparing a package with a 5G quota against a tariff without one is not comparing networks. Put tariffs of the same network generation on both SIMs, each with data left.

03

Live a week at your own places

Home, work, the journey. Measurements three times a day, and a separate line for every failure — a dropped call, a stalled video, a page that would not open.

04

Decide from the table, not the impression

The winner is the network with the steadier result and fewer failures at your places. If the networks are close, price, package size and terms decide the rest — not megabits.

If the check picks a different operator, you can take your number with you: a number transfer takes a few minutes, but you cannot transfer again for another 30 calendar days, and the balance is wiped when you move. Which is why the trial happens before the application, not after. What the tariff you are moving to will really cost is a separate calculation — based on what you actually use.

What to do if the signal is poor at home

Find the spot that works

Signal inside a flat is spread unevenly: by the window, in another room and on another floor the result differs. One spot found solves the problem faster than any change of operator.

Check what is left of the package

An exhausted package produces similar symptoms: the speed is not throttled, but every megabyte starts costing money. The operator's app or USSD shows the balance — on Mobiuz that is *102#.

Weigh it against home internet

If the gigabytes go at home, a mobile network is rarely the cheaper option: home tariffs start at 165,000 UZS a month with unlimited traffic, and their speed depends on the technology at your address and on the time of day.

Account for tethering

A laptop on a hotspot treats mobile data as an ordinary connection and pulls updates at full size. What feels slow afterwards is the queue on the network, not the network.

Home internet and mobile solve different problems, and their prices are worth comparing head to head: what speed you need at home is a separate guide with provider prices. And if the signal disappears when you travel, the cause is usually not coverage but which network the phone has landed on: roaming, eSIM and picking a network by hand.


Test yourself

  1. Which result from a week of testing matters most when choosing a network?
    Почему

    A maximum is one lucky minute. What describes the experience is the typical result and the number of times a call dropped or a video stalled.

  2. From what speed does 1080p video play without stopping?
    Почему

    YouTube recommends 5 Mbps for 1080p and 2.5 for 720p. Higher numbers in a speed test are headroom, not something you will notice while watching.

  3. The internet went slow at the end of the month. What to check first?
    Почему

    Under the terms of these tariffs the speed is not throttled after the limit: charging begins at 100 UZS a megabyte on Mobiuz and 25 on Ucell. The symptom looks like a bad network, but the cause is on the bill.

Frequently asked questions

Which operator has the best coverage?

There is no single answer for the country: coverage is the state of the network at one particular spot, and it differs between neighbouring buildings. The question is settled by a week of measurements at your own four to six places: home, work, the journey, the spot you visit regularly. Log not only the speed but the failures — dropped calls and stalled video. Advertising rankings do not answer this, because they do not measure your room.

How many megabits are enough for a phone?

By the services' own recommendations, not many: 2.5 Mbps for 720p video, 5 for 1080p, 20 for 4K, 1.2 Mbps for a one-to-one video call at 720p and 60–80 kbps for a call with the camera off. So the question is usually not about megabits but about stability: five megabits that hold all evening are worth more than fifty that collapse.

Does the 5G icon on the screen mean high speed?

No, it only means which type of network the phone is attached to. Access to 5G can also be a condition of the tariff: Ucell Bor 110 and Bor 160 VIP list a separate quota “on the 5G network” — 100 and 150 GB — while Perfectum puts 5G into a tariff of its own, A'lo 5G. Networks are worth comparing on tariffs of the same generation; otherwise what is being compared is the terms, not the networks.

Why does the speed drop in the evening?

In the evening a residential area puts more people on the same cell at once, and the capacity is divided between them. Which makes the evening measurement the one to decide by: it describes the hours when the network is actually used. The daytime figure at the same spot is usually higher and settles less.

What can I do if the signal is bad at home?

Start by finding a spot inside the flat that works: by the window, in another room and on another floor the result differs. Then check what is left of the package — once the limit is passed the speed is not throttled, but every megabyte starts costing money. The third step is a week of measurements three times a day, with failures logged. And if the internet is mostly spent at home, compare the price against home internet.

How can I try an operator out before moving my number?

Put a second SIM or an eSIM in the same phone and run both networks side by side for a week: one device, the same places, the same hours. Give both SIMs tariffs of the same network generation, each with data left, or you will be comparing terms rather than networks. Run the trial before you apply: a transfer wipes the balance, and the number cannot be moved again for 30 calendar days.

What to read next

Compare operators once you have your own results

Tariffs, internet packages and the official terms of seven operators — with filters for price and volume.

Operator terms as of 9 September 2026.

Want to see the honest price?

Save a product in farq.uz — we show the 90-day low next to the current price tag.

Download
Repost — please link back to farq.uz/blog

Read next

Mobile
How to pick a mobile tariff by what you actually use
17 min
Roaming
Internet on a trip: roaming, a local SIM or an eSIM
19 min
Home internet
What home internet speed you actually need
18 min
How to check an operator's coverage and speed yourself | farq.uz