RSRP, RSRQ, SINR and RSSI explained

A signal screen that shows four numbers is not showing the same thing four times. Each field answers a different question about the radio link, and the reason engineers keep all of them is that a connection can fail in several unrelated ways. Strength tells you how much of the wanted signal arrives. Quality tells you how much of what arrives is wanted at all. The signal-to-noise figure tells you how cleanly the modem can decode it, and total power tells you how loud the channel is in general, useful noise included. Learn which field moves when a problem appears and you can usually name the cause without leaving the room.

Reference power: the strength field

The reference power figure is measured only on the pilot symbols a base station transmits at known positions in the frame. Because those symbols are known in advance, the receiver can measure them without any traffic present and without other users interfering, and the result is a clean estimate of how much of this particular cell's energy reaches your antenna. It is averaged per resource element, which is why it does not rise when the cell gets busier. This is the field to watch when you are asking a question about distance, walls and geometry — the things that attenuate a signal on its way to you.

Total received power: everything in the channel

The wideband received strength counts all the energy inside the measured bandwidth: your serving cell, every neighbouring cell using the same frequency, and the noise floor of the receiver itself. On its own it is a poor guide, because a large value can mean a strong wanted signal or a room full of interference. Its usefulness is comparative. On older networks it is the only strength figure available, which is why second and third generation readings sit around twenty decibels above modern reference values for the same conditions.

Quality: the ratio that exposes congestion

The quality field is defined as the number of resource blocks multiplied by the reference power, divided by the total received power, expressed in decibels. In plain terms it asks what fraction of everything arriving is actually your cell's pilot energy. Values near minus ten are healthy; minus fifteen is mediocre; minus twenty or worse means the channel is crowded with energy that is not helping you. Because the divisor includes traffic from your own cell, quality falls naturally in the evening peak even if nothing about your position has changed. That single behaviour explains most complaints of the form nothing moved but everything got slower.

Signal to interference and noise: the throughput predictor

The signal-to-interference-and-noise ratio is the number the modem effectively bargains with. It decides which modulation and coding scheme can be used, and therefore how many bits per symbol you get. Below zero decibels a link survives only with the most robust coding, giving very low rates. Around ten decibels the connection is comfortable. Above twenty the modem can use the densest modulation the standard allows, and that is where headline speeds come from. This field is not part of every reporting interface on every handset, so it may appear as unavailable on some devices while the others are present.

Reading them together

The combinations are what make a diagnosis. Strong reference power with poor quality means interference or a loaded cell: you are close to a tower, and so is everyone else. Weak reference power with acceptable quality means you are simply far away or behind a wall, on a cell that is otherwise calm — moving helps, waiting does not. Weak on both means edge of coverage with neighbours bleeding in, the classic case for a dropped call at a cell border. Strong on both with slow data points away from the radio entirely, toward the core network, the backhaul, or the server at the other end.

Why a field sometimes shows nothing

Not every chipset reports every field, and not every network populates them. Older devices may expose strength but not quality; some report a signal-to-noise figure only while a data session is active. A meter that invents a plausible value in those cases is lying to you, so the honest behaviour is to print a placeholder and leave it empty. If a field stays empty on your device across several networks, it is your handset's reporting interface that is limited, not the coverage.

A worked example

Suppose the screen shows a reference power of minus eighty-two, a quality figure of minus eighteen and a signal-to-noise ratio of two decibels. Strength says you are close to the mast and nothing structural is in the way. Quality says most of the energy arriving is not your cell's pilot signal. The low ratio confirms it: whatever is arriving is competing with something of a similar size. The likely story is a crowded cell in a dense area, possibly with a neighbour on the same frequency bleeding in from across a square. Nothing about your position will fix that, and a speed test at three in the morning would look completely different. Now flip it: minus one hundred and six, quality minus nine, ratio twelve decibels. Very little energy is reaching you, but what does arrive is clean, which is a rural or deep-indoor picture on a calm cell. Here movement is the whole game, and two metres toward a window can add ten decibels.