Point-to-point wireless for sites without fiber
A wireless bridge can move Ethernet between buildings faster than trenching, but only when the path, spectrum, capacity, availability, mounting, and power are designed as one system.
Fiber is often the long-term benchmark for building-to-building connectivity. But trenching can be delayed, prohibited, temporary, or more expensive than the site can justify. A point-to-point wireless link can bridge Ethernet across that gap without treating the path like ordinary Wi-Fi.
The original rule still holds: line of sight comes first. But visible line of sight is only the beginning. Trees leaf out, cranes move, roofs flex in wind, the Fresnel zone needs clearance the eye does not see, and advertised throughput means little unless the link can hold the required modulation through weather and interference.
Use point-to-point wireless when two locations can support stable directional radios, a clear and future-safe RF path, legal spectrum, enough predicted IP throughput, and the availability your operation requires. Do not choose equipment from distance claims alone. Model the path, set the service target, build fade margin, verify mounting and power, then commission the link in both directions.
The four-gate go / no-go test
A point-to-point link is viable only when it passes four independent gates. A strong radio cannot compensate for a blocked path, an unstable mount, an illegal channel, or a capacity target the platform cannot sustain.
If one gate fails, stop selecting radios. Change the path, raise the mounting height, reduce the service target, choose another band, add an intermediate site, use licensed spectrum, or return to fiber. Hardware selection comes after path viability.
What a point-to-point wireless bridge actually does
A point-to-point (PtP) link uses a directional radio at each endpoint to connect two network locations. In a common Ethernet-bridge deployment, the wireless pair behaves like a controlled Layer 2 path between switches, buildings, cameras, access points, or other network equipment.
That is different from indoor Wi-Fi, where many mobile clients share an access point, and different from point-to-multipoint (PtMP), where one access point serves several remote sites. PtP dedicates the radio path to two endpoints, which makes antenna alignment, link budget, capacity direction, and availability easier to engineer.
Line of sight is not enough: protect the Fresnel zone
Seeing the far rooftop does not prove the RF path is clean. The first Fresnel zone is a three-dimensional area around the direct path. Objects inside it can create diffraction, absorption, and reflected paths that increase loss and reduce the fade margin available for the target data rate.
A common planning target is to keep at least 60 percent of the first Fresnel zone clear. Critical or long-distance links may justify more conservative clearance. Trees deserve special attention because height, leaf density, moisture, and wind change over time; a winter survey can hide a summer failure.
For longer paths, visual inspection is not enough. Use accurate endpoint coordinates, verified antenna heights, a terrain profile, and the correct earth-curvature or propagation model. Then update the profile with obstructions the map may not know about: parapets, trees, billboards, cranes, tanks, towers, and planned buildings.
Choose the topology before the product
The word "wireless" does not identify the right architecture. First decide whether the project needs a dedicated two-site bridge, one hub serving several locations, or a physical fiber path.
| Need | Best starting point | Why |
|---|---|---|
| Connect exactly two buildings | Point-to-point wireless | Dedicated directional path with capacity and alignment designed for two endpoints |
| One tower or rooftop must serve many sites | Point-to-multipoint | One sector or access point coordinates multiple subscriber locations |
| Highest long-term capacity with a practical physical route | Fiber | No RF interference, no spectrum constraints, and strong upgrade headroom |
| Primary path plus rapid backup | Fiber + diverse wireless | Media diversity can reduce common failure modes when routes and power are truly independent |
| Temporary site, construction phase, or lease restriction | Point-to-point wireless | Can be deployed and removed without trenching, subject to safe mounting and a viable path |
A PtP link is not automatically a replacement for fiber. It is a different transmission medium with different failure modes, deployment speed, regulatory requirements, and maintenance needs.
When fiber is still the better choice
Prefer fiber when a practical route exists and the project needs the strongest long-term capacity headroom, immunity from RF interference, simple symmetrical scaling, or a physical medium that can support future optical upgrades without replacing the radio path. Wireless remains valuable for rapid deployment, difficult crossings, temporary service, backup, and locations where trenching is not practical.
Build the path before choosing hardware
A useful link plan starts with site data, not a product page. The planner should know the endpoints, mounting heights, target service, spectrum environment, and required availability before selecting a radio, antenna, or channel width.
What a real site survey should capture
Tools such as Cambium LINKPlanner model path profile, equipment, antennas, height, predicted throughput, fade margin, and availability before purchase. The model is only as good as the coordinates, heights, clutter, obstruction data, and service target entered into it.
Choose frequency by path and service target
No frequency band is universally best. The correct choice depends on distance, capacity, channel availability, interference, weather, antenna size, regulatory rules, and the availability target. Compare actual certified products in the jurisdiction where the link will operate.
A common starting point for commercial unlicensed backhaul because the product ecosystem is deep and the band supports many link lengths and antenna options.
Can provide additional spectrum for compatible fixed-wireless platforms. In the United States, outdoor standard-power operation is controlled by an Automated Frequency Coordination system.
Useful for high-capacity, narrow-beam links when the path and distance fit the selected platform. The small Fresnel zone can help in dense sites.
Often considered when the operation requires predictable interference conditions, engineered availability, longer paths, or service-provider-class backhaul.
Do not set the country code, frequency, transmit power, or antenna gain to values that do not match the certified equipment and local rules. Unlicensed does not mean unregulated. DFS, AFC, EIRP, channel, and professional-installation requirements still apply.
Plan capacity from applications, not the advertised radio rate
Start with payload traffic. A warehouse extension, a camera backhaul, and a remote office may all request "one gigabit," but their direction, burst behavior, latency sensitivity, and growth are different.
The PHY rate is not the same as delivered IP throughput. Protocol overhead, channel conditions, modulation, duplex scheduling, packet size, encryption, wired port speed, and traffic in the opposite direction all affect what applications receive. Use manufacturer-specific IP-throughput data and a link planner at the expected modulation, not the largest number on the box.
Is the target aggregate, symmetrical, or strongly asymmetric?
What latency and jitter can voice, control, and video tolerate?
How much growth is expected during the useful life of the link?
Does the wired interface, PoE injector, switch, or SFP create a bottleneck?
What throughput remains at the minimum acceptable modulation mode?
Link budget and fade margin in plain English
A link budget predicts whether enough signal reaches the far radio to support the required modulation. It combines transmitter power, antenna gain, path loss, equipment and cable losses, receive antenna gain, receiver sensitivity, and the margin reserved for fading and uncertainty.
Fade margin is not simply "extra signal." It is the difference between the predicted receive condition and the threshold needed for a target mode after accounting for the design model. More demanding modulation usually needs a stronger, cleaner signal. A link can remain connected at a lower mode while losing much of the throughput the application expected.
Translate availability into downtime
Availability is the predicted percentage of time the link remains above a defined threshold. Choose the target from the operational consequence of an outage, then design the path and redundancy around it.
May fit noncritical connectivity when interruption is tolerable and recovery is simple.
A stronger target for important business links, subject to path and redundancy design.
Requires disciplined engineering and often redundancy beyond one unprotected path.
Mathematical equivalents before planned maintenance, SLA exclusions, power failures, network-device failures, or repair time. A prediction is not a guarantee.
Design for weather, interference, and regulatory events
Weather affects bands differently. Rain attenuation becomes increasingly important at higher microwave and millimeter-wave frequencies, while atmospheric oxygen absorption is especially significant near 60 GHz. At lower unlicensed bands, interference and foliage may dominate instead.
On 5 GHz DFS channels, compliant devices must detect protected radar and stop or move as required. That may produce a startup wait, channel change, or interruption. On U.S. outdoor standard-power 6 GHz, AFC determines permitted frequencies and power based on geolocation and incumbent protection. Those behaviors belong in the availability plan, not in post-installation troubleshooting.
RF weather margin does not protect against a failed PoE supply, an unplugged switch, a saturated uplink, a moving mast, a cut outdoor cable, or a single power circuit. Critical links need power, network, physical, and sometimes path diversity.
The physical installation is part of the RF design
A link that aligns perfectly on installation day can degrade if the mast twists, the parapet mount flexes, the cable fills with water, the connector is strained, or the radio loses clean power. Narrower beams and higher-capacity links make mechanical stability more important, not less.
- Mount to structure, not convenience. Verify wind loading, bracket capacity, corrosion environment, roof rules, and safe service access.
- Confirm the exact PoE requirement. Outdoor radios may use IEEE PoE or model-specific passive PoE; the wrong injector can fail to power or damage equipment.
- Use outdoor-rated cable and proper strain relief. Protect connectors, maintain bend radius, add drip loops, and support the cable so its weight does not pull the radio.
- Ground and protect the system as specified. Follow the equipment manual, local electrical code, and qualified installer requirements for mast, cable, surge protection, and bonding.
- Plan maintenance before the lift leaves. Document access, spare hardware, alignment method, replacement procedure, and who owns the roof or tower work.
Do not stop at RF: design the network path
The radio link may behave like an Ethernet cable, but the network still needs deliberate design. Decide whether the link will bridge VLANs, route between sites, carry a management network, transport multicast, or participate in a redundant topology.
Bridge only what must cross
Define VLAN trunks, native VLAN behavior, MTU, spanning-tree role, broadcast control, and loop protection. A transparent bridge can also transparently extend a mistake.
Route when separation helps
Routing can contain broadcasts, clarify failure domains, support resilient paths, and make policy easier to manage. Match the design to the radios and the wider network.
- Change default credentials and use supported management encryption.
- Separate management access from ordinary user traffic where practical.
- Document IP addresses, VLANs, routes, NTP, DNS, monitoring, and alert thresholds.
- Use QoS only after identifying the actual critical traffic and bottleneck.
- Keep firmware and configuration backups under change control.
- Monitor signal, noise, modulation, throughput, latency, errors, temperature, and uptime.
Commission the link before calling it complete
Association is not acceptance. A link is complete only after the measured baseline matches the design closely enough, applications pass, and the documentation allows another technician to recognize future degradation.
What common failure signatures usually mean
| Symptom | Likely areas to investigate | First checks |
|---|---|---|
| Link is connected but slower than planned | Low modulation, interference, misalignment, channel width, TDD split, wired bottleneck | Compare measured RSL/SNR and mode to the plan; test both Ethernet ports and both directions |
| Signal changes with wind | Mount movement, mast twist, cable load, narrow-beam alignment | Inspect structure and brackets; correlate signal history with weather |
| Performance falls after trees leaf out | Fresnel-zone foliage and seasonal excess path loss | Update the obstruction profile; raise, move, or redesign the path |
| Unexpected channel changes or downtime | DFS radar detection, AFC constraints, automatic channel behavior | Review logs, regulatory settings, selected channel, and local requirements |
| One direction is weaker or slower | Asymmetric interference, power, antenna, cable, alignment, or scheduling | Compare both ends separately; do not rely on one aggregate metric |
| Good RF metrics but poor application performance | Switching, routing, VLAN, MTU, duplex, QoS, congestion, or server limits | Test hop by hop and isolate the wireless link from the rest of the path |
When fiber or another path is the better choice
Common questions
Can a point-to-point wireless link replace fiber?
It can replace or supplement fiber for many building-to-building, campus, industrial, surveillance, temporary, and backup connections when the path, capacity, availability, spectrum, mounting, and power all meet the requirement. Fiber may remain the better choice when a practical route exists and the project needs very high long-term capacity or freedom from RF interference.
How far can a point-to-point wireless bridge go?
There is no useful universal distance. Range depends on frequency, certified power, antenna gain, path clearance, terrain, climate, interference, channel width, target throughput, and availability. Use the exact equipment in a path-planning tool rather than buying from a maximum-range headline.
Does point-to-point wireless require line of sight?
For predictable high-capacity backhaul, clear line of sight and Fresnel clearance are the preferred design. Some platforms can operate near line of sight, but obstruction loss is harder to predict and usually reduces margin, modulation, throughput, or availability.
What is the Fresnel zone?
It is a three-dimensional RF region around the direct path between antennas. Objects inside the first Fresnel zone can increase loss through diffraction, absorption, and reflections even when the far antenna is visibly clear.
Will trees block a wireless bridge?
They can. Foliage absorbs and scatters RF, especially when wet, and seasonal growth changes the path. A link that barely clears winter branches may lose throughput or fail after leaves return. Avoid designing through trees when reliability matters.
Does rain affect point-to-point wireless?
Yes, but the impact depends strongly on frequency, path length, rain climate, antenna, and fade margin. Rain attenuation becomes more important at higher microwave and millimeter-wave frequencies; 60 GHz links require model-specific rain and oxygen-loss planning.
Should I use 5 GHz, 6 GHz, or 60 GHz?
Choose from the path and service target. Five GHz has a broad ecosystem but can be congested and may use DFS. Outdoor standard-power 6 GHz in the United States uses AFC. Sixty GHz can deliver high capacity on suitable shorter paths but has stronger atmospheric and rain effects. Licensed microwave may be better for critical engineered backhaul.
How much real throughput will I get?
Use manufacturer IP-throughput data at the predicted modulation, channel width, distance, and traffic split. Do not equate PHY rate with application throughput. Validate both directions under simultaneous load during commissioning.
Do I need an FCC license?
Many 5 GHz, 6 GHz, and 60 GHz systems can operate under unlicensed Part 15 rules when certified equipment and all conditions are followed. Other microwave bands require licensing and coordination. Unlicensed systems still have channel, power, DFS, AFC, EIRP, and installation requirements.
Is a wireless bridge secure?
It can be part of a secure design when the platform supports appropriate encryption, credentials are changed, management access is controlled, firmware is maintained, and VLAN or routing policy is deliberate. The RF link does not replace normal network-security controls.
Can the link carry IP cameras, VoIP, and VLANs?
Usually yes when the selected radios and network design support the required Ethernet features, capacity, latency, multicast behavior, and VLAN configuration. Camera traffic is often strongly asymmetric, so plan the uplink direction and retention architecture carefully.
What is the difference between PtP and PtMP?
Point-to-point connects two endpoints with a dedicated path. Point-to-multipoint uses one access point or sector to serve several subscriber sites. PtMP is more scalable for many locations, while PtP is easier to dedicate and engineer for one pair of sites.
Do outdoor wireless radios need grounding and surge protection?
Follow the exact equipment installation manual and local electrical code. Outdoor radio systems commonly require a deliberate approach to mast bonding, shielded outdoor cabling, surge protection, grounding electrodes, cable entry, and qualified installation.
Technical references
- Cambium Networks: LINKPlanner network design tool
- Cambium LINKPlanner: Fresnel Zone
- Cambium LINKPlanner: Path Loss and Fade Margin
- Cambium LINKPlanner: IP Throughput, Mode Availability, and Fade Margin
- eCFR: 47 CFR § 15.407 U-NII, DFS, and 6 GHz operating requirements
- eCFR: 47 CFR § 15.255 operation in the 57–71 GHz band
- FCC: Licensed point-to-point microwave under Part 101
- Ubiquiti: DFS channel behavior and considerations
- Ubiquiti: 6 GHz AFC operation
- ITU-R P.530: Terrestrial line-of-sight propagation planning
- ITU-R P.676: Atmospheric gas attenuation
- ITU-R P.525: Free-space attenuation
Regulatory conditions vary by country, band, device class, antenna, installation type, and location. Confirm the current rules and certified configuration for the exact equipment before deployment.
The bottom line
A reliable wireless bridge is not two radios pointed at each other. It is a documented path with clearance, capacity, fade margin, legal spectrum, stable mounting, correct power, network policy, commissioning data, and an owner responsible for monitoring it.
Send EhubAmerica the endpoint coordinates, proposed mounting heights, distance, target throughput, traffic direction, application type, and availability requirement. We can help narrow the right platform and build a procurement list around the path instead of around a datasheet headline.
When you are ready, browse enterprise networking equipment, explore Cambium wireless platforms, or request a link-planning quote.
We stock and support networking, wireless, two-way radio, security, and life-safety equipment for integrators and enterprise teams across the US. Questions on a path or bill of materials? Reach us at operations@ehubamerica.com.