Link Budget Calculator
An interactive downlink link-budget tool — EIRP, path loss, noise floor and required SINR to a live SNR and margin.
A link budget is the accountancy of a radio link: start with the power you transmit, add every gain, subtract every loss, and see whether what arrives at the receiver is enough to decode above the noise. Set the terms below and the downlink SNR and margin update live.
The Calculator
Enter the transmitter, the path loss, and the receiver/noise terms. The tool computes EIRP, the noise floor, received power, the resulting SNR, and the margin against your required SINR.
—
Defaults model a mid-band macro downlink: ~64 W (49 dBm) into a 24 dBi massive-MIMO array, 135 dB path loss (roughly a cell edge), 100 MHz, a 7 dB UE noise figure, and a −3 dB required SINR (QPSK with coding). Push the path loss up until the margin turns negative to find the coverage limit.
How the Budget Adds Up
Four steps take you from the power amplifier to a decodable signal:
1. EIRP — the effective radiated power in the beam:
2. Noise floor — the thermal noise the receiver must beat, set by bandwidth and the receiver's noise figure (kTB in dBm is −174 dBm/Hz at room temperature):
3. Received power — EIRP minus the path loss, plus any receiver antenna gain:
4. SNR and margin — how far the received power sits above the noise, and whether that clears the SINR the chosen MCS needs:
A positive margin means the link closes (the receiver can decode with room to spare); a negative margin means you are short — reduce the MCS (lower required SINR), add antenna gain, or reduce the path loss (smaller cell).
Where Path Loss Comes From
Path loss is the single biggest and most variable term. It grows with distance and frequency and depends heavily on environment (line-of-sight vs. buildings, indoor penetration). As a feel: free-space loss at 3.5 GHz is already ~63 dB at 100 m and ~83 dB at 1 km, and real macro-cell models (e.g. 3GPP UMa/UMi) add 20–40 dB for clutter, plus 15–25 dB more for building penetration. That is why sub-1 GHz bands (lower path loss) give coverage while mmWave (very high path loss) gives capacity over short ranges. This tool takes path loss as an input so you can plug in whatever propagation model or measurement you trust.
| Term | Typical DL value | Notes |
|---|---|---|
| Tx power | 43–49 dBm (macro) | Total across the array |
| Array gain | 16–26 dBi | Massive-MIMO beamforming gain |
| UE noise figure | 6–9 dB | Handset receiver |
| Required SINR | −6 to −3 dB | QPSK + strong coding (cell edge) |
Q. Why does a wider bandwidth raise the noise floor?
A. Thermal noise power is proportional to bandwidth (N = −174 + 10·log₁₀(BW) + NF). Doubling bandwidth adds 3 dB of noise, so for the same received power the SNR drops 3 dB — wider carriers need more received power to hit the same SNR. They still carry more total bits, but each Hz is a little noisier.
Q. How can SNR be usable at −3 dB (below the noise)?
A. Modern channel coding (LDPC/turbo) plus processing gain lets QPSK decode at negative SNR — the required SINR for the lowest MCS is around −5 to −3 dB. That's why cell-edge users fall back to QPSK: it tolerates the worst SNR.
Q. What's the difference between this and a coverage prediction?
A. This is a single-point budget: given a path loss, does the link close? A coverage prediction runs a propagation model over geography to turn "maximum allowed path loss" (where margin = 0) into a cell radius and a map. This tool gives you that maximum-allowed-path-loss point.
Related
Pair this with the measurement and adaptation pages to connect SNR to real reported metrics.