Interactive Application Note — Surveying 5G Backhauling for Cameras
Deciding, from a ground survey, whether a 5G site can carry a pole-mounted camera's video uplink — and turning that into a Pass/Fail profile for the CellScout surveyor.
The application
Fixed cameras are mounted high on poles and stream video back over a 5G mobile network — the cell is the camera's backhaul. Each camera needs a dependable upload (e.g. 5 Mbps) to keep its stream alive. Before an installer commits a pole, a technician walks the site with a handheld surveyor and reads the radio conditions. The question is simple to ask and easy to get wrong:
The radio is 5G Standalone on an FR1 band (selectable below — the camera-backhaul default is mid-band TDD n77), measured through the four metrics a surveyor sees — SINR, RSRP, RSRQ and RSSI. The engine models how those translate into achievable upload, accounts for the fact that the surveyor stands on the ground while the camera lives 20+ m up its own pole, and adapts to the chosen band's duplex, sub-carrier spacing, resource blocks and power class. The device is the Quectel RM520N-GL inside CellScout (details in Section 6).
How to use this note
Pick the band and set your scenario in the controls, read the SINR the surveyor should see in Table 1, sanity-check the coverage gates in Section 4, then build one row per site/cell in the Site-Profiles tool and download a CSV for CellScout. The highlighted cell in Table 2 always shows where your target is first met.
Scenario controls
Primary parameters
Advanced parameters (defaults are fine for most surveys)
Height gain = coefficient × log₂(camera height / survey height), capped at 15 dB. It lifts the RSRP/RSSI ground lines always, and the required ground SINR only in the noise-limited regime (up high a cell can instead become interference-limited — that residual stays in the survey margin).
Parameter guide — what each control does, why it matters, and its impact
| Parameter | What it controls | Why it matters | Default | Impact |
|---|---|---|---|---|
| 5G SA band (FR1) | Carrier band & its duplex / SCS / power class | TDD vs FDD changes the whole UL-airtime model; sets SCS, PRB count and HPUE eligibility | n77 | High |
| Channel bandwidth | PRB count (via the band's SCS) | Throughput scales ~linearly with PRB — the biggest lever after the airtime model | 30 MHz | High |
| Target upload / camera | Required application-layer uplink per camera | The pass/fail target itself — scales every SINR threshold | 5 Mbps | High |
| Cameras on the cell | How many streams share the uplink | More cameras split the PRB, raising the SINR each needs; can make the target infeasible | 1 | High |
| Streaming at once (%) | Fraction of cameras active simultaneously | Sets the real concurrent load; CBR video ≈ 100% | 100% | Medium |
| Max UL modulation | Spectral-efficiency ceiling (64 / 256QAM) | Decides feasibility on high-target or many-camera sites | 256QAM | High |
| UL MIMO rank | Spatial layers (1 or 2) | Rank-2 ~doubles capacity where it locks near the cell | 1 | Med–High |
| Cell resources for cameras (%) | PRB left after other traffic | A loaded cell grants fewer PRB, raising the required SINR | 100% | High |
| Camera height (m) | Antenna height on the pole | Height coverage gain lifts RSRP/RSSI (and SINR when noise-limited) — the core reason ground ≠ pole | 20 | Medium |
| Surveying Antenna Gain (dBi) | Gain of the ground survey antenna | Shifts measured RSRP/RSSI 1:1; baseline for the antenna delta | 10 | Medium |
| Camera Antenna Gain (dBi) | Gain of the pole-mounted antenna | With survey gain, gives the delta translating ground reading to the camera | 10 | Medium |
| UE Tx power | HPUE 26 dBm vs Class-3 23 dBm (auto per band) | Shifts the RSRP power-limit gate by ~3 dB | band | Medium |
| SINR limiting regime | Whether antenna/height gain transfers to SINR | Gain helps SINR when noise-limited, not when interference-limited | Interference | Medium |
| Ground→pole survey margin (dB) | Residual uncertainty buffer | Absorbs what height/antenna modelling can't — interference up high, pattern, orientation | 3 | Medium |
| App→PHY overhead (%) | HARQ + IP gap between app and PHY rate | Sets the PHY target above the application target | 20% | Medium |
| Fading margin (dB) | Camera confidence margin above the floor | Separates "bare minimum" from "recommended" | 2 | Low–Med |
| Survey antenna height (m) (adv) | Height of the survey antenna | Baseline for the height gain vs the camera | 2 | Low–Med |
| Height gain (dB/octave) (adv) | How strongly height converts to coverage gain | Environment-dependent; set 0 to disable and rely on the margin | 2 | Medium |
| PUSCH data symbols/slot (adv) | Usable symbols after DMRS | Minor throughput scaling; stable for fixed cameras | 12 | Low |
| Custom TDD UL share (adv) | Exact operator UL % — adds a column (TDD) | Precise when you know the slot config | off | Med* |
| Transport (adv) | TCP vs UDP | TCP adds ~5% overhead (ACK/retransmit) | UDP | Low |
Impact = how strongly the parameter moves the required-SINR / feasibility result. *Custom TDD is Medium only when set, and applies to TDD bands only.
1 · Required uplink SINR
For TDD bands the uplink gets only a slice of airtime (the biggest single driver of throughput), shown under three common slot patterns. FDD bands transmit uplink continuously on their paired carrier, so a single column applies. Camera SINR is what the modem on the pole must achieve; Ground-survey target is what the surveyor should read, after antenna-gain, height and margins.
| TDD pattern | UL share | Efficiency needed | Modulation | Camera floor | Camera recommended | Ground-survey target |
|---|
Unknown operator config on a TDD band → use the ~20% UL row. "Not achievable" means the target exceeds what this channel/rank/modulation can carry at that camera count — widen the channel, add rank-2, lift the modulation cap, or reduce cameras.
2 · Achievable upload vs measured SINR
Estimated per-camera application-layer upload at each SINR, after sharing the cell among simultaneously-active cameras. Highlighted cell = target first met.
SINR→modulation uses standard ~10% BLER operating points (±1–2 dB is noise). Rank-2 roughly doubles capacity where it locks near the cell.
3 · The ground survey is a proxy — the camera lives 20 m up
The surveyor stands on the ground with a paddle antenna on a ~2 m pole; the camera it is qualifying sits 20+ m up. Two different radio environments, now modelled explicitly through the height and antenna controls.
| Effect at height | Direction & how the tool handles it |
|---|---|
| Clearer line-of-sight, less ground clutter / vehicle / foliage blockage | Camera usually better — the height gain raises RSRP/RSSI (and SINR when noise-limited). |
| Sees more cells from up high (the "too-high" effect) | Camera can be worse — interference; left in the survey margin (not auto-credited to SINR). |
| Different antenna gain / pattern / orientation vs the survey paddle | Offset either way — the two Antenna-Gain controls handle the gain part deterministically. |
4 · Supporting metrics — RSRP, RSRQ, RSSI
These don't set throughput; they confirm the SINR figure can be trusted. Gate lines show the camera requirement and the ground-survey line (shifted for antenna-gain difference, height and UE power). RSRQ is antenna-invariant — it cancels.
| Metric & role | Yellow/Green | Orange/Yellow | Red/Orange |
|---|
5 · Combined verdict (SINR-primary)
| Condition — use the Ground-survey values for the site's airtime column | Verdict |
|---|---|
SINR ≥ ground-survey target
ANDRSRP ≥ ground Orange/Yellow
ANDRSRQ ≥ −13 dB
ANDRSSI consistent with RSRP/RSRQ
| RF PASS |
SINR between ground floor and target
ORRSRP in its conditional band
ORRSRQ between −13 and −16 dB
| BORDERLINE |
SINR < ground floor
ORRSRP below Red/Orange
ORRSRQ < −16 dB
| RF FAIL |
RF PASS = the air interface can carry the upload; a shortfall then implicates the IP/transport side. BORDERLINE = run the onboard speedtest or read the camera modem. RF FAIL = fix coverage/interference/config first.
Antenna-gain impact — why it isn't just "add dB to the margin"
Open: how survey antenna gain moves each of the four metrics
Antenna gain is a known, deterministic shift — unlike the survey margin, which absorbs the unknown offset. It hits the four metrics differently, so it can't be folded into a single SINR-style margin:
| Metric | Effect of survey antenna gain | Why |
|---|---|---|
| RSRP | Shifts ~1:1 (−3 dBi → −3 dB) | Gain lifts received signal power directly. |
| RSSI | Shifts ~1:1, same direction | Gain lifts total wideband power too. |
| RSRQ | ~Invariant | RSRQ = N·RSRP/RSSI — gain is in both, so it cancels. |
| SINR | Conditional: ~1:1 noise-limited, ~0 interference-limited | Gain lifts signal above the fixed receiver noise floor, but lifts interference along with it. |
The same logic governs the height gain: full effect on RSRP/RSSI, transferred to SINR only when you set the regime to Noise-limited.
Site profiles → CSV for CellScout
Capture one row per site/cell, then download a CSV of Pass/Fail profiles. CellScout's Gauge View takes user-defined Red / Orange / Yellow / Green thresholds per parameter (Poor / Fair / Good / Excellent) with a beeper — the CSV gives the three bin-edges per metric, ground-referenced, ready to load.
| Incl. | Site | Cell | Band | Airtime | Target | Cams | SINR R/O·O/Y·Y/G | Feasible |
|---|
No sites added yet. Set your scenario, enter a Site/Cell ID, pick the airtime, and click "Add current scenario".
Bin edges are the CellScout gauge boundaries: ≥ Yellow/Green = Excellent (green), ≥ Orange/Yellow = Good/Pass (yellow), ≥ Red/Orange = Fair (orange), below = Poor/Fail (red). Set the beeper at the Orange/Red edge for a hard Fail alert. SINR bins are site-specific (from the chosen airtime); RSRP/RSSI carry the antenna, height & power shift; RSRQ is universal.
6 · Assumptions & device facts
| Item | Basis |
|---|---|
| Bands | RM520N-GL 5G SA FR1 bands that carry uplink (TDD + FDD); SDL bands excluded. UL 2×2 on n38/n41/n48/n77/n78/n79. |
| SCS / slots | TDD bands 30 kHz (2000 slots/s); FDD bands 15 kHz (1000 slots/s). PRB per 3GPP 38.101-1. |
| UE Tx power | HPUE Class 2 ≈ 26 dBm on n38/n40/n41/n77/n78/n79; 23 dBm otherwise (overridable). |
| Airtime | TDD: uplink share of the slot pattern; FDD: continuous paired uplink. |
| Metrics source | AT+QENG="servingcell" gives RSRP/RSRQ/SINR in NR-SA; RSSI derived (RSRP − RSRQ + 10·log₁₀ N_RB). |
| Camera sharing | Simultaneously-active cameras split UL PRB equally; each judged at its own SINR. |
| Height model | Coverage gain = coef × log₂(camera/survey height), capped 15 dB; to RSRP/RSSI always, to SINR only when noise-limited. |
7 · Acronyms & glossary
| Term | Meaning |
|---|---|
| SINR | Signal-to-Interference-plus-Noise Ratio — sets achievable modulation/throughput (dB). |
| RSRP | Reference Signal Received Power — coverage strength (dBm). |
| RSRQ | Reference Signal Received Quality = N·RSRP/RSSI; load & interference (dB). |
| RSSI | Received Signal Strength Indicator — total wideband power (dBm). |
| FR1 | Frequency Range 1 — sub-7 GHz 5G NR (all the bands here). |
| 5G SA / NSA | Standalone (5G core) / Non-Standalone (5G anchored to LTE). |
| TDD / FDD | Time-Division Duplex (UL & DL share the channel in time) / Frequency-Division Duplex (paired carriers). |
| SDL / SUL | Supplemental Downlink / Uplink — single-direction bands (SDL can't carry camera uplink). |
| DL / UL | Downlink / Uplink. Camera video is uplink. |
| DDDSU | TDD slot pattern (Down×3, Special, Up) — ~20% uplink airtime; common public-5G default. |
| PRB / RB | (Physical) Resource Block — 12 subcarriers; the scheduling unit. |
| RE | Resource Element — one subcarrier × one OFDM symbol. |
| SCS | Sub-Carrier Spacing — 30 kHz on NR-TDD, 15 kHz on NR-FDD for this module. |
| MCS | Modulation and Coding Scheme — modulation + code rate (bits per RE). |
| QPSK / 16/64/256QAM | Modulation orders — 2 / 4 / 6 / 8 bits per symbol; higher needs higher SINR. |
| HPUE | High-Power UE — Class 2 (~26 dBm) uplink, allowed on n38/n40/n41/n77/n78/n79. |
| MIMO / rank | Multiple-Input Multiple-Output; rank = parallel spatial layers. |
| PUSCH / DMRS | Physical Uplink Shared Channel (carries the video) / its Demodulation Reference pilots. |
| BLER / HARQ | Block Error Rate (~10% target) / Hybrid ARQ fast retransmission. |
| gNB / UE | 5G base station / User Equipment (camera's RM520N-GL modem, or CellScout). |
| FWA | Fixed Wireless Access — often uses UL-heavy TDD configs. |
| LOS / NLOS | Line-of-Sight / Non-Line-of-Sight — central to the pole-height effect. |
| dBm / dBi / dB | Power vs 1 mW / antenna gain vs isotropic / a ratio. |
| QENG | Quectel engineering AT command reporting serving/neighbour cell metrics. |
CellScout Mobile Network Surveyor: https://telecomtest.com.au/product/cellsscout-mobile-network-surveyor/
Interactive application note · FR1 5G SA camera-backhaul survey. Figures are engineering estimates for field screening, not a throughput guarantee — judge coverage at the camera, and calibrate with CellScout's built-in Speedtest.