Why Your Signal Looks Good and Your Speed Does Not

The metric that misleads operators most often

RSSI (Received Signal Strength Indicator) is the number most operators look at first. It is the number that is easiest to pull from any radio management platform. It is also, by itself, almost useless as a diagnostic tool for throughput problems.

RSSI tells you one thing: how much signal power arrived at the CPE antenna. It does not tell you whether that signal is clean, whether it is being degraded by interference, or whether the modulation scheme the radio selected to encode that signal is capable of delivering the throughput the subscriber is paying for. For that, you need SINR.

Core Distinction

RSSI determines whether a link connects. SINR determines whether it performs. A link with strong RSSI and poor SINR is connected, stable-looking, and chronically underperforming. This is the most common diagnostic blind spot in fixed wireless operations.

What SINR actually measures

Signal-to-Interference-plus-Noise Ratio is the ratio of the desired signal power to the combined interference and noise power at the receiver. Expressed in decibels. A SINR of 20 dB means the desired signal is 100 times stronger than the combined interference and noise. A SINR of 7 dB means it is five times stronger, which sounds adequate until you see what MCS index that produces under real sector load.

The practical implication is this: two subscribers at different distances from the tower can have very different performance outcomes even if their RSSI readings are similar. The subscriber closer to the tower with moderate SINR may actually perform worse than the subscriber farther away with clean SINR, if the closer subscriber is sitting in an interference zone: a sector overlap area, an azimuth edge where two RN sites compete for CPE association, or a location where co-channel interference from a neighboring operator arrives at similar power to your own signal.

The MCS index: where SINR becomes throughput

The Modulation and Coding Scheme (MCS) index is the number your radio scheduler selects based on link quality. Higher MCS values use more complex modulation, encoding more bits per symbol, and deliver higher throughput. Lower MCS values use simpler modulation that is more resilient to noise but delivers a fraction of the peak throughput.

MCS Index	Min SINR Required	Sustained DL (Typical)	Deployment Context
MCS 9	25+ dB	120 to 180 Mbps	Tier 1, LOS, clean channel
MCS 7	18 to 24 dB	80 to 130 Mbps	Tier 1 to Tier 2, clean channel
MCS 5	12 to 17 dB	50 to 90 Mbps	Tier 2, moderate obstruction
MCS 3	7 to 11 dB	20 to 45 Mbps	Tier 3. Marginal. Unstable under load.
MCS 1	Below 7 dB	Under 15 Mbps	Tier 4. Do not approve.

A subscriber with -75 dBm RSSI and 7 dB SINR is at MCS 3. Their sustained throughput under real sector load is 20 to 45 Mbps. If they are paying for a 100 Mbps plan, they are calling your support line every time they run a speed test. Their RSSI looks fine. Their link is not.

Field Rule

A subscriber at -87 dBm with 20 dB SINR will outperform one at -75 dBm with 7 dB SINR in every real-world condition, every time the sector is under load. RSSI gets the link up. SINR determines whether it stays useful.

Why SINR degrades without RSSI changing

This is the scenario operators find most confusing: RSSI is steady, SINR is dropping, and nothing appears to have changed in the RF environment. The subscriber's signal looks the same as it did six months ago. Their speeds are half what they were.

Several things can cause SINR degradation without RSSI movement:

A new co-channel operator deployed nearby. Their signal arrives at the CPE at a lower power level than yours, not enough to affect RSSI materially, but enough to raise the interference floor and compress the SINR by 4 to 8 dB. This is the GAA interference scenario. It happens without warning and shows up as sector-wide throughput degradation during the hours when the neighboring operator's subscribers are active.

Sector utilization crossed a threshold. As more subscribers are added to a sector, the aggregate interference from their own uplink transmissions increases. A sector at 70% busy-hour utilization generates materially more internal interference than the same sector at 40%. SINR drops across the sector. RSSI does not move. Throughput degrades for every subscriber simultaneously during peak hours.

An adjacent sector began overlapping your coverage area. If a new site was deployed nearby with coverage that reaches into your existing sector, CPEs in the overlap zone will see competing signals at near-equal levels. The CPE scheduler treats this as interference. SINR collapses in the overlap zone. RSSI from your sector is unchanged.

Backhaul reached saturation during the busy hour. This one does not affect SINR at all, but it presents identically in speed tests. If your backhaul link to the RN site is saturated, CPEs will show clean RF metrics and deliver slow speeds. This is not an RF problem and cannot be solved with RF adjustments. Checking backhaul utilization before dispatching field resources for throughput complaints is Field Decision Rule 04 in the standards reference for exactly this reason.

The three-step diagnostic sequence

When throughput complaints arrive, this is the correct sequence. Every step takes under five minutes. Following it in order prevents dispatching field crews to sites where the RF is not the problem.

Check backhaul utilization first. Pull peak-hour utilization for the RN site serving the complaining subscriber. If it is above 70%, the problem is infrastructure, not RF. Dispatching a technician to adjust CPE alignment will accomplish nothing. Address the backhaul constraint. If backhaul is healthy, proceed to step two.

Pull SINR for the specific CPE and sector-wide average SINR. If the individual CPE SINR is below 10 dB and the sector average is above 14 dB, the problem is localized to that subscriber's link: installation quality, a new obstruction, or incorrect sector association. If the entire sector SINR is degraded, the problem is sector-level: interference, capacity saturation, or a competing signal source.

Check MCS index for the affected subscriber against their historical baseline. If MCS has dropped by two or more indices without a change in RSSI, SINR degradation from external interference is the most likely cause. Pull a spectrum scan for the sector channel. Identify any new energy on or adjacent to your operating channel. If you cannot pull a spectrum scan from your NMS, that is a separate problem worth solving before the next one like this arrives.

What to look for in your NMS

The metrics that reveal this problem are not always surfaced by default in network management platforms. Operators who want to catch this pattern before subscribers call need these visible on their sector dashboards:

Per-CPE SINR, not just RSSI, logged at least hourly, visible per subscriber

Sector-average MCS index: a drop in sector average MCS is the earliest visible indicator of interference or capacity saturation

Busy-hour utilization per sector, peak hour not daily average. Daily average hides the problem entirely.

Backhaul utilization per RN site, same logic: peak hour measurement required

Subscriber packet loss per CPE: above 0.5% sustained is a link quality signal, not a throughput signal

Operator Note

Most operators monitor average throughput. Average throughput is useless for FWA quality management. The metric that drives subscriber experience is busy-hour utilization at the sector level, combined with per-subscriber SINR. If you are not monitoring both of these, you are not monitoring your network. You are monitoring a summary that will not tell you anything is wrong until your support queue tells you first.

The bottom line

Signal looks good. Speed does not. The cause is almost never what it appears to be on the surface. The diagnostic sequence above takes fifteen minutes to run through completely. It will identify whether you have an RF problem, an interference problem, a capacity problem, or an infrastructure problem. It will keep you from sending a technician to a site where the answer is not in the RF environment.

The frameworks for SINR thresholds, tier classification, and sector capacity management are in Parts 3, 4, and 6 of the NGWA Standards. If your operation does not yet have documented signal acceptance thresholds or a formal capacity monitoring process, that is where to start before the next throughput complaint arrives.