Will That Colocated System Kill Our WiFi? How to Predict Interference Before Deployment

Understanding how to predict whether a new WiFi network will destroy your throughput, how energy detection thresholds fail to protect against adjacent-channel interference, why minimum distance calculations often use wrong assumptions, and when RF filters actually help versus when they’re expensive theatre that won’t solve your interference problem

👉 Not a member? Free link Here

A manufacturing facility shares a building with three other companies. Your WiFi network operates on channel 149 (5.8 GHz). The tenant upstairs just deployed their own WiFi on channel 144, directly above your APs. Your controller warns “adjacent channel interference possible” but provides no concrete answer: Will it actually interfere? How far should they be? Do you need RF filters? How much will performance degrade?

These questions repeat across collocated deployments: office buildings with WiFi on every floor, shared warehouse space with multiple operators, industrial parks with adjacent facilities. Traditional answer: “It depends on many factors”, followed by vague recommendations for separation distance or RF filters. A better answer requires understanding three mechanisms of interference and how to evaluate each. This article explains how to predict collocated system interference without months of testing, when to recommend RF filters versus when to recommend different channels, and what measurements prove whether your network survives coexistence.

The Three Mechanisms of Colocated Interference

When two wireless systems operate near each other, interference happens three ways. Most deployments ignore two of them, focusing only on SNR degradation while missing blocking and CSMA/CA false triggering that often cause bigger problems.

Mechanism 1: SNR Degradation (The Obvious One)

The adjacent system transmits a signal that arrives at your receiver as unwanted noise. Your signal-to-noise ratio (SNR) can drop. If a collocated system transmits continuously at +10dBm while 10 meters away, and your receiver sensitivity is -95 dBm, considering an adjacent channel of yours, the received power in the same central frequency you are using is close to -70dBm (dependent on the wireless mask), where this value of interference will raise your effective noise floor.

Your original SNR of 25 dB drops to lower than this (even zero). Performance degrades, modulation may drop from 64-QAM to QPSK (or lower), and throughput falls. This mechanism is well-understood.

Problem: Most interference predictions assume that interference only happens via SNR degradation. But two other mechanisms often cause more damage.

Mechanism 2: Energy Detection False Positives (The Hidden One)

WiFi uses CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance): before transmitting, a device listens for energy at its own centre frequency. If energy is detected above the Energy Detection (ED) threshold (typically -62 dBm for 20 MHz channels), the device thinks “channel is busy” and defers transmission. This prevents collisions with other WiFi devices.

However, adjacent-channel interference may or may not trigger ED depending on the received power at its own centre frequency:

Your receiver is tuned to your centre frequency. It only measures energy present at that frequency. When a colocated system transmits on an adjacent channel (±20 MHz away), the energy that reaches your receiver at your centre frequency is attenuated by the colocated system’s spectral mask (typically -25 dB at ±20 MHz offset) plus path loss.

Example — Adjacent Channel (Safe):

Colocated system: +10 dBm at their frequency

Spectral mask at your frequency offset: -25 dB

Path loss (10 meters): -50 dB

Energy received at your centre frequency: +10–25–50 = -65 dBm

Your ED threshold: -62 dBm

Since -65 dBm < -62 dBm: ED does NOT trigger

ED false triggers occur when the received interference power EXCEEDS (is stronger than) your ED threshold. Adjacent channels with proper spectral masks typically keep interference below the threshold.

Mechanism 3: Receiver Blocking (The Worst One)

Strong adjacent system signals overload your receiver’s front end even if the signals are on different frequencies. Receiver amplifier saturates, can’t amplify desired weak signal properly, desired signal becomes unreadable. This happens even when filter attenuation is significant (20–30 dB) because the strong interferer is so powerful.

👉 Buying RF filters before measuring colocated interference at your actual centre frequency — exactly like this article warns against — is how you end up spending $2,000+ on filters that don’t solve the problem. My 178-page Industrial Wi-Fi Guide includes ED threshold decision logic wireless engineers use to predict colocated interference correctly and negotiate with tenants BEFORE deployment — not after discovering filters didn’t help.[Get the guide and stop buying useless filters]

Measuring Interference Potential

Before the colocated system goes live, collect three pieces of information to predict interference severity. This requires no access to their equipment, only knowledge of what they’re deploying.

Step 1: Frequency Proximity Assessment

The distance between your centre frequency and their centre frequency determines spectral mask attenuation. WiFi channels are 20 MHz wide. Spectral mask attenuation is NOT linear — it’s standardised by 802.11 regulations.

Distance matters because it determines what attenuation applies:

• Same channel (0 MHz separation): Spectral mask = 0 dB attenuation. Interference arrives at FULL power. Very dangerous.
• Adjacent channels (±20 MHz separation, e.g., 149 ↔ 144): Spectral mask = approximately -28 dB attenuation at your centre frequency. The colocated signal is significantly reduced before even considering path loss.
• 30 MHz speration: Spectral mask = -40 dB or greater attenuation. Interference is already heavily reduced at the source.

Figure 1 — Wireless standard mask

Step 2: Transmit Power and Distance Analysis

Ask the co-located tenant: What’s their AP transmit power and approximate distance from your APs? Then calculate the received power at YOUR center frequency (not just the raw signal).

Formula (considering spectral mask):

Received Interference Power = Their Transmit Power — Spectral Mask Attenuation — Path Loss

Examples with spectral mask included:

Scenario A: Adjacent Channel (+20 dBm at 5 meters)

• Transmit: +20 dBm
• Spectral mask (adjacent, -20 MHz): -25 dB
• Path loss at 5m: -47 dB (typical 5 GHz)
• Received at your frequency: +20–25–47 = -52 dBm
• Above the ED threshold (-62 dBm), it may cause some false deferrals. Problematic.

Scenario B: Adjacent Channel (+20 dBm at 30 meters)

• Transmit: +20 dBm
• Spectral mask: -25 dB
• Path loss at 30m: -60 dB
• Received at your frequency: +20–25–60 = -65 dBm
• Verdict: ✅ Below ED threshold (-62 dBm), but very close.

ED Threshold vs Signal Detection: Why Both Matter

Energy Detection (ED) Threshold: Related to different central frequencies and/or different channels. Typical -62 dBm. Any detected energy above this, the device defers transmission (CSMA/CA mechanism). Protects against collisions with other WiFi but also causes false deferrals when non-WiFi energy triggers it.

Signal Detection (SD) Threshold: Related to the same central frequency and the same channel. Typical -95. Only signals above this are decoded as valid WiFi frames.

Signals between ED and SD thresholds trigger “channel busy” (deferral).

Adjacent channel interference often lands in the ED-SD gap. Strong enough to trigger CSMA/CA deferral (-62 dBm exceeded) — Result: Your device thinks the channel is busy and defers. Throughput collapses due to false deferrals.

Minimum Distance Calculation: How to Estimate Safe Separation

Distance determines safety. A colocated system at 5 meters is dangerous. At 50 meters, usually safe. The question: what’s the minimum distance for your specific situation?

The Simple Model

Start with interference at their transmitter: typically +20 dBm. For every meter of distance, the signal weakens. Precise weakening depends on the environment (open space, walls, metal), but average weakening is significant.

Calculate the distance using the attenuation per distance formula, adding the attenuations of the elements. Use the same model as we used in previous examples.

With RF filter isolation: Subtract filter isolation from interference power. 30 dB filter means interference reduced 30 dB. Their -70 dBm after path loss becomes -100 dBm after 30 dB filter (below noise floor).

When RF Filters Actually Help vs When They’re Theatre

RF filters are expensive ($200–2000 per AP depending on quality) and sometimes necessary, sometimes useless theatre. Understanding when saves money and frustration.

RF Filters HELP with ED False Triggers when:

• Colocated interference power is between -62 dBm and -50 dBm (above ED threshold but not too strong)
• Filter provides enough isolation (20–30 dB) to push below -62 dBm

· Gap between your channel and theirs is small (adjacent channels): 20–40 MHz separation means filter leakage is significant. 30–40 dB filter helps.

· You can’t change channels or distance: If physically unable to increase separation or use different frequencies, an expensive RF filter may be the only option (though not perfect).

RF Filters DON’T Help with ED False Triggers when:

• Colocated interference is extremely strong (like -30 dBm before filtering)
• Even after 30 dB filter isolation, the result is still -60 dBm (above -62 dBm threshold)
• You need more than filter isolation — you need distance increase or channel change

👉 Deploying WiFi in colocated space and unsure if adjacent channels will interfere with your network? I work 1:1 with network engineers to calculate interference power at your actual centre frequency, compare it to ED thresholds, assess whether filters help or distance negotiation is required, and validate predictions through pilot testing — transforming RF filter guesswork into data-driven decisions that save thousands and prevent post-deployment surprises.[Check coaching availability]

Conclusion

Colocated system interference is not a weak signal problem — it’s a problem of where the signal arrives at your frequency. The spectral mask acts as your first defence: -25 dB for adjacent channels, -45 dB for 2+ channel separation.

Calculate interference at YOUR centre frequency: Colocated Power — Spectral Mask — Path Loss. Compare the result to the ED threshold.

Most expensive mistakes happen because engineers buy filters without measuring the actual interference at their frequency. The trick: negotiate distance or channels BEFORE the colocated tenant arrives. Measuring correctly saves $2,000+ in unnecessary filters.