LOS Is Not Just "Can I See It?": UAV Radio Link Engineering Explained

1. Start with the mission

A UAV radio link should not be selected by looking at peak data rate or advertised range alone.

A useful communication link has to satisfy the mission requirement:

• How far should the vehicle operate?
• What video resolution and bitrate are required?
• Is telemetry safety-critical?
• Is the payload controlled over UART, Ethernet, or IP?
• Does the operator need access to an onboard computer?
• What latency is acceptable?
• Is the link point-to-point, point-to-multipoint, or multi-node?
• Is the environment open field, urban, mountainous, maritime, or industrial?
• What antennas can physically fit on the UAV?
• What is the available power budget?

A small UAV doing 1 km inspection with 720p monitoring has a very different communication requirement from a long-range platform sending 1080p video, telemetry, and payload data over 5–10 km.

2. Start with the traffic budget

Before calculating range, calculate the data requirement.

A typical UAV data link may carry:

A practical data budget should include margin:

This does not mean the radio should advertise "12 Mbps." It means the link should provide at least 12 Mbps of stable application throughput under the expected field condition.

3. PHY rate is not application throughput

Radio datasheets often mention PHY rate. For example, an 802.11ac 2×2 radio may support a peak PHY rate of up to 867 Mbps under ideal conditions.

But PHY rate is not the same as usable throughput.

Application throughput is reduced by:

• MAC overhead
• TCP/UDP/IP overhead
• encryption overhead
• channel contention
• retransmissions
• modulation and coding changes
• packet loss
• antenna mismatch
• multipath
• aircraft orientation
• distance
• RF noise
• CPU and interface limits

A realistic planning table:

For UAV links, stable throughput matters more than peak throughput.

4. Video bitrate planning

Video is usually the largest part of the UAV data budget.

The required bitrate depends on resolution, frame rate, codec, scene complexity, encoder settings, GOP/keyframe interval, CBR/VBR mode, latency target, and error recovery behavior.

For UAV field use, bitrate should be selected based on link margin, not only image quality.

A common mistake is to test video at 1080p 15–20 Mbps near the ground station, then expect it to remain stable at several kilometers.

5. Latency is a system-level number

Radio latency is only one part of the total video/control latency.

End-to-end latency includes the entire chain:

Camera exposure
→ image pipeline
→ encoder
→ packetization
→ operating system queues
→ wireless radio
→ network jitter buffer
→ decoder
→ display

For UAV operations, latency should be considered relative to the task:

6. Link budget: the core RF calculation

A radio link budget estimates whether enough signal power reaches the receiver.

The simplified equation:

Received Power =
  Tx Power
+ Tx Antenna Gain
- Tx Cable Loss
- Free Space Path Loss
+ Rx Antenna Gain
- Rx Cable Loss

In dB form:

Prx(dBm) = Ptx(dBm) + Gtx(dBi) - Ltx(dB) - FSPL(dB) + Grx(dBi) - Lrx(dB)

The link works only if received power is above receiver sensitivity by a useful margin:

Link Margin = Received Power - Receiver Sensitivity

Example:

If receiver sensitivity for the selected modulation is −88 dBm:

Link Margin = −79.7 − (−88) = 8.3 dB

An 8 dB margin may work, but it is not comfortable for a moving UAV. Aircraft attitude, antenna nulls, multipath, and interference can quickly consume margin.

7. Free-space path loss at 5 GHz

Free-space path loss is the signal loss caused by spreading of electromagnetic energy over distance.

The commonly used formula:

FSPL(dB) = 32.44 + 20·log₁₀(distance_km) + 20·log₁₀(frequency_MHz)

For a 5.8 GHz link:

Two important observations:

1. Every doubling of distance adds about 6 dB of path loss.
2. Higher frequencies have higher path loss for the same distance.

This does not mean 5 GHz cannot be used for UAV links. It means antenna planning, channel selection, and link margin are important.

8. LOS is not just "can I see it?"

Visual line-of-sight and RF line-of-sight are not the same thing.

At radio frequencies, energy does not travel as an infinitely thin beam. The useful radio path occupies an ellipsoidal region around the straight line between transmitter and receiver. This region is called the Fresnel zone.

Even if the drone is visible, objects near the path can disturb the signal — ground, terrain slopes, trees, rocks, buildings, the aircraft body, the carbon-fiber frame, the battery pack, or the payload enclosure.

The first Fresnel zone radius can be approximated by:

r = 17.32 × √((d1 × d2) / (fGHz × D))

At the midpoint of the link, d1 = d2 = D/2.

For 5.8 GHz, approximate midpoint first Fresnel zone radius:

A common rule is to keep at least 60% of the first Fresnel zone clear.

So at 5 km, if the midpoint Fresnel radius is about 8 m, roughly 5 m of clearance around the path is desirable.

This is why a UAV can be visually visible but still have a weak radio link.

9. Antenna gain, beamwidth, and practical trade-offs

Antenna gain is not free power. It reshapes the radiation pattern — a higher-gain antenna concentrates energy in a narrower direction.

Air side

The UAV antenna should be lightweight, low-drag, mechanically stable, placed away from carbon fiber and batteries, and positioned to reduce blockage during turns and attitude changes.

Ground side

The ground antenna can be larger and higher gain because weight is less critical.

10. Cable loss matters at 5 GHz

At 5 GHz, RF cable loss can be significant. Thin coax over a long run reduces link budget quickly.

A 3 dB cable loss means half of the RF power is lost before reaching the antenna.

Keep RF cables short, use good-quality connectors, avoid sharp bends, secure cables against vibration, and place the radio close to the antenna where possible.

11. Channel bandwidth: range versus throughput

Wider channels can carry more data, but they need better signal quality and a cleaner spectrum.

Noise power increases with bandwidth. Thermal noise is approximated by:

Noise Floor(dBm) = −174 + 10·log₁₀(B) + NF

Example with NF = 5 dB:

Doubling bandwidth increases noise by about 3 dB. This means 80 MHz mode may show high peak rate at short range, but 20 MHz or 40 MHz may be more practical for field operation.

12. Receiver sensitivity and modulation

Modern radios change modulation and coding rate depending on signal quality.

This is why range and throughput are connected. The link may still exist at long range, but the usable throughput may be much lower.

For UAV data links, the important question is not "is the link connected?" — it is "does the link provide enough stable throughput and low enough latency for the mission?"

13. Link budget example for a 5.8 GHz UAV link

A simplified 5 km example:

Received power:

Prx = 29 + 3 − 1 − 121.7 + 12 − 1
Prx = −79.7 dBm

Comparing to receiver sensitivity:

At high data rate, receiver sensitivity may be worse. At lower data rate, sensitivity is usually better. The same physical link may support drone telemetry reliably while failing to sustain high-bitrate video.

14. Why UAV orientation affects link quality

A UAV is not a fixed tower. It moves, turns, vibrates, pitches, rolls, and yaws.

This affects antenna polarization, antenna pattern, body blockage, multipath, MIMO performance, cable stress, and connector stability.

During a turn, the aircraft body or battery may partially block the antenna. If the antenna has a null in that direction, the received signal can drop suddenly.

This is why flight testing is different from ground testing. A radio link that looks stable when both units are sitting on a table may behave very differently when the UAV banks during flight.

15. MIMO and diversity

A 2×2 MIMO radio uses two transmit/receive chains, which can improve throughput, diversity, multipath handling, and link stability — but only when the antenna system is designed properly.

Important points:

• use both antenna ports
• maintain physical separation where possible
• avoid placing both antennas in the same null region
• consider polarization diversity
• avoid blocking both antennas with the same airframe component
• test the link while the UAV changes orientation

For compact UAVs, MIMO antenna placement is often a mechanical integration problem as much as an RF problem.

16. QoS: protect telemetry before video

In a mixed-data UAV link, video can easily dominate bandwidth. If video consumes the entire link, telemetry and command traffic may suffer.

A good configuration should prioritize:

1. command/control
2. telemetry
3. payload safety/status data
4. video stream
5. dashboards and diagnostics
6. file transfer/log download

Practical strategies:

• set video bitrate limits
• use CBR or capped VBR
• avoid uncontrolled high-bitrate streams
• separate traffic classes where possible
• prioritize telemetry packets
• avoid large file transfers during flight
• reduce video bitrate at range
• monitor packet loss and latency

For many UAV missions, a slightly lower-quality video stream with stable telemetry is more useful than high-quality video with delayed aircraft state.

17. Range planning should include fade margin

A static link budget is only the starting point. Real field links need fade margin to account for multipath, aircraft rotation, temporary obstruction, atmospheric variation, antenna mismatch, polarization mismatch, connector aging, RF noise, and installation differences.

For UAV operations, a margin of 10 dB or more is desirable when possible.

18. Field test methodology

A good UAV radio test should be progressive. Do not start with the maximum range claim.

Step 1: Bench test

Validate power stability, boot behavior, Ethernet connection, UART settings, video stream, telemetry flow, IP access, web dashboard, SSH/diagnostics, and thermal behavior.

Step 2: Short outdoor test

At 50–100 m, test antenna orientation, video latency, telemetry update rate, packet loss, and basic mobility.

Step 3: Medium range test

At 500 m to 1 km, test video stability, link recovery, aircraft turns, antenna blockage, RF noise, and telemetry continuity.

Step 4: Mission range test

At 3 km, 5 km, or target range, test required video bitrate, link margin, latency under load, packet loss during turns, Fresnel clearance, ground antenna height, and failure behavior.

Step 5: Edge-of-range test

Intentionally reduce margin to understand when video degrades, when telemetry becomes unstable, whether the link recovers automatically, whether bitrate adaptation is needed, and what the operator sees during degradation.

The goal is not only to prove the link works — it is to understand how it fails.

19. Practical configuration for a 5 GHz UAV/UGV data link

For a 5 GHz 2×2 MIMO UAV/UGV data link carrying video, telemetry, UART, and IP traffic, a practical starting configuration:

80 MHz channel operation can be useful at short range, but for longer range and field use, narrower channels may provide a more stable link.

20. Where CY-2 fits in this engineering discussion

CY-2 is a compact UAV/UGV radio link for mixed traffic: video, telemetry, UART, and IP data between a vehicle-side unit and a ground-side unit.

The engineering principles in this article apply directly to CY-2 deployments:

• do not plan only around peak PHY rate
• define a traffic budget before field testing
• keep telemetry protected
• use practical video bitrates
• choose antenna placement carefully
• treat 5 GHz as a high-bandwidth LOS link
• calculate link budget for the target range
• remember that LOS is not just visual visibility
• validate performance with real field tests

For drone companies, the value of an integrated link is not just that it can carry data — it is that it can carry the right mix of data reliably under realistic mission conditions.

21. Frequently asked questions

Is visual line-of-sight enough for UAV radio range?

No. Visual line-of-sight is helpful, but RF line-of-sight also depends on Fresnel zone clearance, antenna height, terrain, frequency, and link margin.

Why does a drone radio work nearby but fail at range?

At longer distance, free-space path loss increases, received signal drops, modulation rate may fall, packet loss can increase, and video throughput may become unstable.

Why does video fail before telemetry?

Video uses much more bandwidth than telemetry. A weak link may still carry low-rate telemetry while failing to sustain high-bitrate video.

Is 5 GHz good for UAV communication?

5 GHz is useful when higher bandwidth is needed, especially for video and IP data. However, it requires better line-of-sight, antenna planning, and Fresnel zone clearance than lower-frequency links.

Should I use 80 MHz channel bandwidth for maximum throughput?

Not always. 80 MHz can provide high peak throughput at short range, but it is more sensitive to noise and requires better signal conditions. For field UAV use, 20 MHz or 40 MHz may be more stable.

What is a good video bitrate for UAV testing?

For initial field tests, 720p at 2–3 Mbps or 1080p at 4–6 Mbps is usually a safer starting point than high-bitrate video.

What link margin should I target?

For moving UAVs, aim for at least 10 dB of practical margin when possible. More is better, especially in difficult terrain or noisy RF environments.

Why does aircraft orientation affect the radio link?

The UAV body, battery, carbon frame, and payload can block or detune antennas. During turns, the antenna pattern and polarization alignment can change, reducing received signal strength.

22. Conclusion

UAV radio link engineering is not just about selecting a radio with a high data rate or a long range claim.

A good UAV data link must be designed around traffic budget, video bitrate, telemetry reliability, latency target, link budget, receiver sensitivity, antenna gain, cable loss, channel bandwidth, Fresnel zone clearance, and field testing.

The most important lesson is simple:

> LOS is not just "can I see it?"

For radio communication, visibility is only the beginning. The RF path must have enough clearance, enough signal strength, enough margin, and enough usable throughput for the mission.

For UAV and UGV companies building real products, this engineering discipline is what separates a demo link from a dependable field communication system. Read more on the UAV radio link and drone data link pages, or reach out if you are working through a specific deployment.