Antenna Utilities
Co-location antenna isolation and physical separation for RF systems integration. Solve horizontal, vertical, and slant separation for a target isolation, budget the required isolation from desense and blocking, compare geometries, size filtering, and check a whole tower layout.
Free forever on a Standard account. No credit card.
Put two antennas on the same tower or rooftop and they couple to each other. When one is transmitting and the other receiving, that coupling can desensitise or block the receiver, and the fix is separation: the antennas have to be far enough apart that the isolation between them holds the interference below what the receiver can tolerate. The question every co-location engineer asks is simply how far, and the answer depends heavily on whether the antennas are side by side or stacked.
The noIM₃ Antenna Separation Calculator sizes that separation with the standard first-order co-siting isolation laws. Horizontal separation, antennas side by side, gives an isolation of 22 + 20·log₁₀(d/λ) less the two antenna gains, rising 20 dB per decade. Vertical separation, antennas stacked so they sit in each other pattern nulls, gives 28 + 40·log₁₀(d/λ), rising 40 dB per decade, which is why a few metres of vertical stacking can beat many metres of horizontal spacing. Slant separation, an antenna offset both ways, is interpolated between the two by the elevation angle. It is a systems integration tool, not an antenna design tool: the gains and the required isolation are inputs.
Five modes cover the co-location job end to end. Separation solves the spacing for a required isolation or the isolation for a spacing. Isolation Budget works out how much isolation you actually need from the interference scenario, receiver desense and receiver blocking. Geometry compares the separation each geometry needs so the vertical stacking saving is explicit. Filtering sizes the extra duplexer or cavity isolation when the available separation falls short. Tower Layout resolves a set of antennas by height and offset into a pairwise isolation matrix, flagging the problem pairs.
Enter the frequency and geometry and solve the minimum spacing two co-located antennas need for a required isolation, reported in metres and wavelengths, or switch to read the isolation an existing spacing achieves. Horizontal, vertical, and slant geometries are all supported.
Horizontal isolation is 22 + 20·log₁₀(d/λ) − Gₜ − Gᵣ, rising 20 dB per decade. Vertical isolation is 28 + 40·log₁₀(d/λ), rising 40 dB per decade because the antennas sit in each other pattern nulls. Slant is interpolated between the two by the elevation angle of the line joining the antennas.
Because the vertical law rises twice as fast, stacking reaches a target isolation in a fraction of the horizontal distance. The Geometry mode reports the separation horizontal, vertical, and slant each need for one target and states the vertical saving directly, so the layout decision is quantified rather than assumed.
A separation is only meaningful against a requirement. The Isolation Budget mode computes the isolation each interference mechanism demands: receiver desense from transmitter noise falling in the receive band, and receiver blocking from the transmitter carrier, and reports the governing one.
A given desensitisation corresponds to a specific interfering power relative to the receiver noise floor: a 1 dB desense to an interferer about 6 dB below the floor, a 3 dB desense to one equal to the floor. The tool computes the tolerable interference and the isolation needed to reach it, so the trade between allowed degradation and required isolation is visible.
When there is not enough room, the shortfall is made up with filtering. The Filtering mode takes the separation actually available, reports the isolation it achieves, and reports the extra duplexer, cavity, or band-reject isolation needed to close the gap, colour coded from sufficient separation to heavy filtering.
The Tower Layout mode takes a set of antennas by height, horizontal offset, and gain and resolves them into a pairwise isolation matrix, colouring every pair pass or fail against the requirement, calling out the worst-case pair, and ranking the pairs so the antennas that need to move or be filtered are obvious.
Runs entirely in your browser. No frequencies, separations, or site layouts are submitted to a server. Useful for commercially confidential work, classified projects, or any environment where information security policy prohibits sending engineering data to third party services.
Reach for the Antenna Separation Calculator in any of the following situations.
It depends on the required isolation, the frequency, and whether the antennas are side by side or stacked. Horizontal separation gives an isolation of 22 + 20·log₁₀(d/λ) less the antenna gains, so reaching 60 dB at 450 MHz side by side takes on the order of tens of wavelengths. Vertical stacking gives 28 + 40·log₁₀(d/λ), so the same 60 dB is reached in a few wavelengths. The calculator solves the exact distance for your frequency, geometry, and required isolation.
Because of the antenna patterns. Antennas mounted side by side point their main beams at each other in azimuth, so the coupling is high and falls off relatively slowly at 20 dB per decade. Antennas stacked vertically sit in each other pattern nulls, where the radiation is weak, so the coupling starts lower and falls off twice as fast at 40 dB per decade. The result is that a few metres of vertical separation can achieve what tens of metres of horizontal spacing would, which is the single most useful fact in co-location layout.
Enough to hold the interference below what the receiver tolerates, and that comes from two mechanisms. Receiver desense is the transmitter broadband noise falling in the receive band raising the noise floor; the isolation must bring that noise below the level that would degrade the receiver by more than the allowed amount, typically 1 dB. Receiver blocking is the transmitter carrier overloading the front end; the isolation must keep the carrier below the blocking level. The Isolation Budget mode computes both from your transmitter and receiver figures and reports the larger, governing requirement.
An interfering power about 5.9 dB below the receiver noise floor. Desensitisation is how much the interference raises the effective noise floor: an interferer equal to the floor adds 3 dB, an interferer 6 dB below adds about 1 dB, and so on. So allowing 1 dB of desense means the transmitter noise reaching the receiver must sit almost 6 dB under its noise floor, and the required isolation is whatever brings the transmitter noise down to that level. The calculator does this conversion for any allowed desense.
Make up the shortfall with filtering. A bandpass cavity, a band-reject notch, a duplexer, or a ferrite isolator adds isolation on top of the physical separation. The Filtering mode breaks the isolation into its contributions and stacks them: the spacing provides one part, any filter already fitted provides another, and the two together are compared against the requirement. So if 3 m of vertical separation gives 54 dB and a fitted cavity adds 15 dB, the achieved total is 69 dB, which the tool shows is 6 dB short of a 75 dB requirement rather than only telling you the raw filtering figure. Breaking the budget down this way makes it far easier to reason about what is limiting you and how much more filtering to buy.
Yes. The Tower Layout mode takes a set of antennas, each with a height, a horizontal offset, and a gain, and computes the isolation between every pair using the appropriate horizontal, vertical, or slant law. The result is a pairwise isolation matrix with every pair coloured pass or fail against the requirement, the worst-case pair called out, and the pairs ranked, so you can see at a glance which antennas on a busy structure are too close.
They are standard first-order far-field estimates, typically good to within about 5 to 10 dB of measured antenna-to-antenna coupling in benign conditions, and worse where there is significant scattering. They are excellent for sizing a mast, comparing geometries, and finding the problem pairs, but they ignore mast and structure scattering, near-field effects at very close spacing, the detailed antenna patterns, nearby buildings and reflectors, feedline and connector coupling, and common-grounding effects, and the vertical form assumes vertically polarised collinear antennas. A prominent planning-estimate confidence indicator carries the expected accuracy and these assumptions on every screen. For a critical or marginal installation, confirm the number with measured coupling or a full electromagnetic model.
No. The calculator runs entirely in your browser. No frequencies, separations, or site layouts are submitted to a server. Useful for commercially confidential work, classified projects, or environments where information security policy prohibits sending engineering data to third party services.
Free forever on a Standard account. No credit card.
