Building a Long-Range Video Link for UGV Operations: Antennas, Bands and Range Math

A link that streams 1080p from an aircraft at 5 km is the same link that dies at 400 m when you strap it to a robot dog between two warehouses. Ground vehicles change the physics of the connection: the antenna is low, the environment is full of edges, and the operator's position is dictated by the mission, not by radio. This article builds the UGV link from published component data — the AEROLINK ground series — and the range arithmetic that actually applies at 40 cm above the ground.

Why ground is harder than air

Radio line of sight from a vehicle antenna is a horizon problem. At 0.5 m antenna height, the geometric radio horizon is roughly 2.5 km over flat ground — but that is geometry, not engineering: real diffraction loss behind the first obstacle is 20–40 dB, which is 100× to 10,000× in power. An aircraft at 100 m clears the obstacle; a UGV drives into its shadow every time it turns a corner.

Three consequences for procurement:

1. Published ground-to-air range figures are upper bounds for elevated antennas. The HK-M30 mesh node specifies 50 km ground-to-air; behind a concrete tank at 50 m it is a different radio.
2. The vehicle's own body is the first obstacle. Mount height, mast design and antenna pattern on the chassis matter as much as the radio's watts — which is why the ground segment in our catalogue includes purpose-built antenna platforms.
3. Latency is a safety number. An operator walking a robot through a doorway is flying FPV at ground speed: the HM30's published 150 ms is the reference for close-in work; long-range links trade latency for penetration, and the trade should be deliberate.

The three-layer build

Layer 1 — the vehicle radio

For a single vehicle in open terrain (quarries, solar farms, pipeline rows), a point-to-point link with a mast-mounted ground antenna works: the HK-T27 at 30 km (40 Mbps, 8 W dissipation, 1.4 GHz with spectrum sensing and dynamic frequency hopping) is the light-duty choice, stepping to the HK-T30 at 50 km when the route is longer than the sightline.

For vehicles that will pass behind things — urban, industrial, indoor-outdoor transitions — mesh is the architecture: the HK-M30 (50 km, 40 Mbps/node) through HK-M46 (300 km) self-route through additional nodes, and the HK-D35 doubles bandwidth to 80 Mbps single-node when the payload is a thermal pod that cannot be compressed into mush.

Layer 2 — the ground segment

This is where UGV links are won. The catalogue's ground-side members exist because "hold the controller" is not an antenna system:

• HK-GC / HK-GD portable ground base stations — 9–36 V vehicle supply, RJ45 network access, 5–10 dBi antennas, in cases sized for a vehicle trunk. This is the operator's mobile command radio.
• HK-GE heavy-duty tracking gimbal — automatic pan-tilt with built-in RTK positioning and MAVLink control, directional + omnidirectional antennas: the antenna follows the vehicle so the beam stays pointed.
• HK-GQ base station with tracking camera and HK-GF hand-cranked mount — 12 dB / 16 dB directional antennas with IP66 enclosures for fixed posts and manual aiming respectively.

A 16 dB directional antenna aimed by hand at a vehicle driving behind a ridge is a different system from the same antenna on an RTK-cued tracking mount. Price the tracking before the mission says it needs it.

Layer 3 — the payload riding the link

The vehicle's camera is the throughput input. A D100 UGV carries 80 kg — the sensor stack (EO/IR pod, lighting, sometimes a LiDAR for mapping) sets the bitrate, and the bitrate sets the link. Run the same budget as any fleet: streams summed at transmitted bitrate, +30% headroom, and remember mesh divides per hop. For a two-sensor night-patrol load, the 80 Mbps class exists for exactly this case.

Range math you can audit

Before buying, write the four numbers:

1. Worst-case distance between operator and vehicle (not the site's longest diagonal).
2. Worst-case obstruction (one wall? a row of tanks? a hill?) — this decides point-to-point vs mesh more than distance does.
3. Required bitrate from the payload choice.
4. Latency tolerance from the driving task (teleop through doorways ≠ checkpoint review).

Then map: open + long → T-series point-to-point with mast. Occluded + mobile → M-series mesh with 2–3 relay nodes. Fixed watch-post → S-series long-range data link (HK-S46: 500 km class, 5 Mbps) where telemetry matters more than video. Every mission needing video and 100 km reach and 80 Mbps is a contradiction — the catalogue's five link families exist because those three numbers trade against each other.

What the quotation conversation looks like

The AEROLINK family is configured per mission — band options, antenna choice, node count — so it is quoted per SET with 14-day validity, EXW Hong Kong. Send: the site sketch, the vehicle model, the payload list, and the destination country (for the 1.4 GHz licensing question). You will get the member of the T/M/S/D/X/G line that matches, the ground segment sized to the terrain, and the link budget in writing — because a UGV link sold without one is just an expensive hope.

Meanwhile, the ground vehicles category and the security surveillance solution show how chassis, sensors and links are specified together in real deployments.