V2X antennas and RF — where the datasheet range goes.
A vehicle-to-everything (V2X) antenna is the part that turns a chipset's quoted range into the range you actually get, and it is the part most likely to be decided last. Two units built on identical silicon, running identical software, can differ by hundreds of metres of usable range purely on antenna choice, placement and the metal around them.
There is no universal V2X antenna design, and any supplier who offers you one has not asked what vehicle it is going on. A rooftop antenna on a bus, a shark-fin on a saloon car and whatever can be fitted to a scooter are three different RF problems with three different answers. What follows is what actually decides the outcome.
The radio is bought. The antenna is designed.
Every V2X unit in a given generation uses one of about five modem parts, and they perform within a few decibels of each other. The silicon landscape → What differs between a unit that works at a junction and one that does not is almost entirely the RF chain around that part: the antenna, where it sits, what metal is underneath it, and how much of the signal is lost before it reaches the connector.
This matters more for V2X than for most vehicle radios because of what the messages are for. A telematics link that drops for thirty seconds recovers and nobody notices. A safety message that fails to arrive at a blind junction has already failed. The relevant metric is not peak range in a car park — it is reliability at the geometry where the warning is needed, which is usually the geometry with a building in the way.
| Loss | Where it comes from | Why it is often missed |
|---|---|---|
| Ground plane | The conductive surface under the antenna shapes the radiation pattern. A monopole over a large roof behaves very differently from the same monopole on a plastic panel or a bench | Bench and anechoic figures are usually taken over an ideal ground plane. The vehicle is not one |
| Feed loss | Coaxial cable, connectors and any board-level transition between the module and the antenna. At 5.9 GHz thin cable is lossy per metre, and a vehicle run is not short | Quoted separately from the module's output power, if at all |
| Body obstruction | The vehicle's own roof, pillars and boot lid, plus other vehicles between sender and receiver | Line-of-sight figures do not include it, and traffic is the normal case |
| Non-line-of-sight geometry | Buildings at intersections, which is exactly where intersection-movement warnings are needed | The headline range figure is a line-of-sight figure |
| Self-interference | Other transmitters in the same enclosure — particularly a cellular transmitter degrading a nearby GNSS receiver | Only appears once everything is integrated, which is late |
Vertical, omnidirectional, and 5.9 GHz.
Industry context, not a product specification. These are the conventions an Intelligent Transport Systems band antenna is designed against.
- Band. The Intelligent Transport Systems allocation sits in the 5.9 GHz region, and the exact block differs by jurisdiction — which is a regulatory fact before it is an antenna fact. 3GPP designates it Band 47 for LTE-V2X and n47 for NR-V2X. Who has allocated what, band by band →
- Polarisation. Vertical, by convention across ITS deployments. A polarisation mismatch between two participants costs signal for no benefit, so the convention is worth more than any individual optimisation.
- Pattern. Close to omnidirectional in azimuth for a vehicle, because a vehicle cannot know which direction the next message arrives from and has no time to steer. Elevation is where the design freedom is: energy radiated into the sky is wasted.
- Diversity. Two antennas rather than one, giving transmit and receive diversity, is common on units that can afford it. It buys robustness against multipath fading rather than raw range — and multipath is the normal urban condition.
- Conformance. RF behaviour in the band is bounded by a harmonised standard rather than left to the designer: ETSI EN 302 571 covers radio spectrum access requirements for the 5 855–5 925 MHz range, and ETSI EN 302 663 defines the ITS-G5 access layer. On the North American side, SAE J2945/1 specifies on-board system requirements for V2V safety communications, including transmit-power behaviour. What to read, and what it costs →
None of the above fixes a design. It bounds one. Within those bounds the vehicle decides everything else.
The coexistence problem is the integration problem.
A C-V2X on-board unit is not a single-radio device, and its antennas are not interchangeable.
| Antenna | Roughly where | What depends on it |
|---|---|---|
| V2X / ITS | 5.9 GHz | The PC5 sidelink itself — every safety message sent and received |
| GNSS | L1 at 1575.42 MHz, often with additional bands | Position in the message payload, and the timing reference the sidelink needs. Degrade this and you degrade the radio, not just the map |
| Cellular (Uu) | Operator bands, low and mid | Certificate batch delivery, trust-list updates and firmware. The wide-area half → |
The GNSS row is the one that surprises people. In C-V2X the sidelink derives synchronisation from satellite time, so a GNSS antenna desensitised by a cellular transmitter twelve centimetres away is not merely a positioning problem — it can put the radio out of synchronisation with everything around it. This is also why GNSS spoofing is a V2X security question rather than a navigation one.
Isolation between these three, in an enclosure small enough to be styled onto a vehicle, is the part of the work that cannot be bought as a module. It is measured, not calculated.
Height, ground plane, and everything the stylist wants instead.
| Position | RF consequence |
|---|---|
| Roof centre | Best case. Maximum height, largest symmetric ground plane, cleanest 360° pattern |
| Roof-mounted shark-fin | The usual production compromise. Good height, shared housing with GNSS and cellular, and the coexistence problem concentrated in one place |
| Mirror or A-pillar | Asymmetric pattern with the vehicle body blocking one side. Sometimes acceptable for a specific approach geometry |
| Bumper or below waistline | Heavily obstructed by the vehicle's own structure and by other vehicles. Poor for V2V, occasionally defensible for a short-range use case |
| Inside the cabin | Behind glass with metallised coating, behind a dashboard, or inside a plastic housing surrounded by wiring. Convenient, and usually the largest single loss in a retrofit unit |
The two-wheeler problem
India's draft fitment proposal covers L category vehicles, which makes this the practically important case rather than an edge case. What the draft actually says →
A motorcycle or scooter has no roof, no significant conductive ground plane where an antenna could reasonably sit, a rider whose body absorbs 5.9 GHz energy and occupies a large fraction of the rearward hemisphere, far more vibration than a car, and a cost ceiling measured in tens of rupees rather than hundreds. It also leans, which rotates the antenna's pattern relative to the road in a way a car never does. Every one of those is an antenna and RF problem before it is anything else, and none of them is solved by choosing a better modem.
This is the clearest example of why there is no universal V2X antenna design: the same message set, the same radio and the same security stack, on a vehicle class where the standard mechanical answer does not exist.
A fixed position can aim. A vehicle cannot.
A roadside unit inverts the vehicle's constraint. It is at a surveyed fixed position, it knows which approaches matter, and it is not going anywhere — so the omnidirectional assumption stops being necessary.
- Sectoral or downtilted antennas can concentrate energy along the carriageway instead of radiating into the sky and the surrounding buildings.
- Multiple antennas per site can cover the arms of an intersection separately, which is often better than one antenna trying to cover all of them equally badly.
- Mounting height is a trade, not a maximum. Higher clears obstructions and increases the distance to the vehicles directly below; gantry, pole and signal-head positions all behave differently.
- The installation constraints are physical. Wind loading, lightning and surge protection, pole sway, and the fact that a technician with a ladder has to be able to reach it are all antenna-adjacent decisions.
The result is that RSU and OBU antenna work are genuinely different disciplines that happen to share a band. The roadside development kit →
What is engineering record, and what is programme work.
Everything above this section is industry context. This section is not, and the boundary is drawn deliberately.
The published record. Ambimat's antenna and RF work appears in its own published milestone history as dated outcomes rather than as a capability statement: antenna control units fabricated and supplied to Bharat Electronics Ltd in 1993, an antenna control unit designed in 2020, a QPSK modulator laid out as a six-layer board mixing an FPGA and digital section with an RF section carrying a 90-degree power combiner and splitter, an SPDT switch and a power amplifier, and FCC certification obtained on a wireless device in 2014. The record, with its sources →
What that establishes. Mixed digital and RF board layout, antenna-adjacent subsystem design, and taking a radio product through a certification process. Those are the disciplines an OBU or RSU antenna and RF chain needs.
What it does not establish. It is not a record of 5.9 GHz V2X antenna design, and no such record is claimed. A V2X antenna qualified for a specific vehicle class — its pattern measured on that vehicle, its coexistence verified against that unit's GNSS and cellular radios, its mechanical survival demonstrated on that mounting — is development work on a programme, not something the lineage already carries.
That work sits inside the hardware and product development engagement, alongside the enclosure, thermal and manufacturing decisions it is inseparable from. Antenna placement is not a late detail on an OBU programme; it constrains the enclosure, which constrains the board, which constrains the thermal budget.
Where this fits.
5.9 GHz spectrum
The band this antenna has to work in, jurisdiction by jurisdiction, with the real MHz figures.
Chipsets and hardware
The silicon on the other end of the feed line, and which parts support antenna diversity.
C-V2X
The PC5 sidelink and the Uu path, and why the radio needs satellite timing.
OBU development kit
The vehicle node, its published GNSS baseline and the antenna it carries.
RSU development kit
The roadside node, where a fixed position lets the antenna be aimed.
Hardware development
Where antenna, enclosure, thermal and manufacturing decisions get taken together.
Antenna constraint deciding your unit?
A vehicle class with no roof, an enclosure already tooled, a coexistence problem between the V2X and GNSS antennas, or a range figure that does not survive the vehicle. Any of those is a better opening than a general enquiry, because it says which part of the design is actually binding.
Questions this page answers.
What is a V2X antenna?
The antenna that couples a vehicle-to-everything (V2X) radio to the air at around 5.9 GHz. In the Intelligent Transport Systems band it is conventionally vertically polarised and close to omnidirectional in azimuth, because a vehicle has no idea which direction the next message will come from and cannot steer. It is not the same antenna as the vehicle's GNSS or cellular antenna, though all three often sit in one housing.
How many antennas does a V2X on-board unit need?
Usually at least three, at three different frequencies. One for V2X at 5.9 GHz; one for GNSS at L1 (1575.42 MHz) and often other bands, because C-V2X needs satellite timing as well as position; and one for the cellular Uu path that carries certificate batches and firmware. Units that use transmit and receive diversity carry two 5.9 GHz antennas rather than one. Fitting them in one enclosure without them degrading each other is the actual design problem.
Where should a V2X antenna be mounted on a vehicle?
As high as possible and as close to the centre of a conductive roof as the vehicle allows, because that gives the cleanest ground plane and the most symmetric pattern. Roof-centre beats shark-fin, shark-fin beats mirror or bumper mounting, and anything below the waistline of the vehicle is compromised by the vehicle's own body. The constraint is rarely radio engineering — it is styling, tooling, cable routing and wash-plant survival.
Why does V2X range in the field differ from the datasheet?
Because the datasheet figure is a link budget measured in favourable conditions, and the field is not. The losses that account for most of the difference are the ground plane the antenna actually sits on, the cable and connector loss between the module and the antenna at 5.9 GHz, obstruction by the vehicle body and by other vehicles, and non-line-of-sight geometry at junctions — which is precisely where the safety case is strongest.
Do a Roadside Unit and an On-Board Unit need the same antenna?
No, and this is one of the few places where the two genuinely diverge. A vehicle is omnidirectional because it does not know where the next participant is. A roadside unit is at a surveyed fixed position looking at known approaches, so a sectoral or downtilted antenna can put the energy where the road is instead of into the sky, and two sectors can cover an intersection better than one omnidirectional antenna. The trade is coverage shape against installation complexity.
Is V2X affected by Wi-Fi or other radios nearby?
It can be. The Intelligent Transport Systems band at 5.9 GHz sits directly above the unlicensed 5 GHz Wi-Fi bands, so out-of-band emissions and receiver desensitisation are real design considerations rather than theoretical ones — ETSI EN 302 571 exists to bound exactly that. Inside a single unit the bigger risk is self-interference: a cellular transmitter a few centimetres from a GNSS antenna will degrade the position and timing the V2X radio depends on.
Last updated 2026-08-19 · Technical reference maintained by Ambimat Electronics, Ahmedabad, India. Corrections: neel.shah@ambimat.com