Wi-Fi Signal Strength in dBm Explained for Network+ (2026)
Last reviewed against the current CompTIA Network+ N10-009 exam page (Domain 2 wireless devices, Domain 5 performance problems) and objectives 2.3 and 5.4; Cisco site survey guidelines, VoWLAN troubleshooting guide and Wireless RF Reference Guide; MetaGeek Wi-Fi Signal Strength Basics; Apple Wi-Fi roaming documentation exam objectives (official objectives). Found an error? Tell us and we will correct it.

In Wi-Fi signal strength dBm terms, -30 is the strongest reading you'll see on a laptop, -67 is the floor for voice and video, -80 is barely connected and -90 is noise. The trap on Network+ is the minus sign. Every reading is negative, so the number closest to zero wins. Offer a student -30, -50, -65 and -80 dBm and ask for the best signal, and the one who picks -80 because "80 is bigger" has just missed an easy point.
This post explains why the numbers are negative, gives you the dBm chart worth memorizing, shows how noise changes the picture, and finishes with eight scenario drills. Want to see how you handle wireless troubleshooting right now? Our free Network+ N10-009 practice test puts a coverage-hole question in the troubleshooting domain. No signup, no email.
Independent study resource. Not affiliated with or endorsed by CompTIA. Exam codes, objectives and weightings are set by the vendor. Always confirm current details on the official exam pages before booking.
Key Takeaways
- Closer to zero is stronger. -30 dBm is about the maximum you'll see standing next to an access point. Every other reading is weaker than that.
- -67 dBm is the number to memorize. MetaGeek lists it as the minimum for VoIP and streaming video, and Cisco's site survey guidelines set the same -67 dBm minimum for voice, paired with a 25 dB signal-to-noise ratio.
- -70 dBm is the floor for reliable email and web, -80 dBm is basic connectivity with unreliable packet delivery, and -90 dBm is at the noise floor (MetaGeek).
- Signal alone doesn't settle a question. SNR is signal minus noise. Cisco wants at least 25 dB for voice and video and 20 dB for data. A "very good" -65 dBm signal next to a -80 dBm noise source is a 15 dB SNR, and voice falls apart.
- N10-009 tests this in objective 5.4 (performance issues, including interference and channel overlap, signal degradation or loss, insufficient wireless coverage, client disassociation and roaming misconfiguration) and objective 2.3 (channels, channel width and frequency bands). Domain 5 is 24% of the exam and Domain 2 is 20%.
Why Is Wi-Fi Signal Strength a Negative Number?
Because dBm measures power against one milliwatt on a logarithmic scale, and the power reaching your laptop is a tiny fraction of a milliwatt.
The formula is dBm = 10 × log10(power in milliwatts). Zero dBm is exactly 1 mW. An access point radio set to 20 dBm is transmitting 100 mW. By the time that signal has crossed the room, bounced off a wall and squeezed through a doorway, your card might receive one hundred-millionth of what left the antenna. Start at 100 mW (20 dBm), lose 80 dB on the way, and you land at -60 dBm, which is a millionth of a milliwatt. That's why the numbers are negative: the scale is anchored at 1 mW, and received Wi-Fi power is always below it.
| dBm | Power | Where you'd see it |
|---|---|---|
| +30 dBm | 1 W | Not a Wi-Fi client. Reference point only |
| +20 dBm | 100 mW | A common access point transmit setting |
| 0 dBm | 1 mW | The anchor of the scale |
| -30 dBm | 1 µW (one thousandth of a milliwatt) | Received signal right next to the AP |
| -60 dBm | 1 nW (one millionth of a milliwatt) | A good signal a room or two away |
| -90 dBm | 1 pW (one billionth of a milliwatt) | Down in the noise |
Two rules of thumb do almost all the exam arithmetic for you. Three dB doubles or halves the power. Ten to the power of 0.3 is 1.995, close enough to 2 that nobody argues. Ten dB is a factor of ten. So the step from -60 dBm to -70 dBm isn't "a bit weaker". It's one-tenth of the power. From -60 to -80 is one-hundredth.
Cisco uses the 3 dB rule in its own RF guidance. The Cisco Wireless RF Reference Guide notes that when you bond two 20 MHz channels in 5 GHz under FCC rules, the signal on each channel drops 3 dB. The same transmit power is now spread across twice the bandwidth. Same power, half per channel, minus 3 dB. That's the whole rule in one sentence.
Distance follows the same logic. Free-space path loss grows with the square of the distance, so every time you double your distance from the access point you lose about 6 dB (20 × log10(2) = 6.02). Frequency matters too. At the same distance, moving from 2.4 GHz to 5 GHz adds roughly 6.4 dB of free-space loss (20 × log10(5 ÷ 2.4)). That is one reason the Cisco RF Reference Guide says 2.4 GHz travels farther than 5 GHz, with 5 and 6 GHz behaving similarly.
The Wi-Fi Signal Strength dBm Chart Worth Memorizing
Here's the scale worth memorizing. The thresholds come from MetaGeek's Wi-Fi Signal Strength Basics, a widely cited reference page. The -67 dBm voice line is also the figure in Cisco's site survey guidelines, which call -67 dBm the minimum recommended signal strength for voice applications, with a minimum SNR of 25 dB.
| Reading | What it means | What works |
|---|---|---|
| -30 dBm | Maximum achievable. You're standing next to the access point | Everything |
| -67 dBm | Minimum for applications that need reliable, timely packet delivery | VoIP, Wi-Fi calling, streaming video |
| -70 dBm | Minimum for reliable packet delivery | Email, web browsing |
| -80 dBm | Minimum for basic connectivity. Packet delivery may be unreliable | Not much. Pages load eventually |
| -90 dBm | Approaching or drowning in the noise floor | Any functionality is unlikely |
Notice there's no row for -50 dBm. That's deliberate. Between -30 and -67 you're in territory where the signal isn't your problem, and a question that offers -50 against -30 comes down to one thing: closer to zero is stronger. Both are excellent. -30 is better.
Signal Is Half the Story. Noise Is the Other Half
A reading of -65 dBm looks great on the chart. So why is the VoIP call on that laptop breaking up?
Because what a radio actually needs is a gap between the signal it wants and everything else on the channel. That gap is the signal-to-noise ratio, and it's simple subtraction: SNR = signal (dBm) minus noise floor (dBm). Cisco's wireless troubleshooting guidance is that real-time applications such as voice or video need at least a 25 dB SNR, versus 20 dB for data applications. The Cisco VoWLAN troubleshooting guide ties the two numbers together: the noise floor should not exceed -92 dBm, which is what gives you a 25 dB SNR at the -67 dBm cell edge.
Run the arithmetic on two rooms:
- Room A: signal -65 dBm, noise -92 dBm. SNR is 27 dB. Voice is fine.
- Room B: signal -65 dBm, noise -80 dBm (a microwave oven, a neighbor's hotspot on the same channel, a badly placed cordless phone base). SNR is 15 dB. Same signal reading, and the call stutters.
That's the pattern to watch for in scenario questions. When a question hands you both a signal number and a noise number, it's an SNR question, and the right move is to subtract before you judge. When a question says the signal is strong but throughput is poor, think interference first, not coverage.
One more vocabulary point, because it shows up in tool screenshots. RSSI and dBm aren't quite the same thing. MetaGeek's RSSI explainer describes RSSI as a relative index: the 802.11 standard allows a 0 to 255 scale and each chipset vendor picks its own maximum, so the raw number isn't comparable between adapters. dBm is absolute. Most tools convert RSSI to dBm so humans can read it, which is why "RSSI" and "signal strength in dBm" get used interchangeably in practice.
What Does N10-009 Actually Ask About Wireless Signal?
Two objectives carry most of it.
Objective 5.4, troubleshoot common performance issues, is where signal strength lives on the current exam. Its wireless sub-list names interference and channel overlap, signal degradation or loss, insufficient wireless coverage, client disassociation issues and roaming misconfiguration. Domain 5, Network Troubleshooting, is the heaviest domain on the exam at 24%, and CompTIA's Network+ exam page lists performance problems among its topics. A -85 dBm site-survey reading at the far end of an office, like drill 2 below, is a 5.4 scenario: the answer is a coverage hole, not a passphrase or a captive portal.
Objective 2.3, select and configure wireless devices and technologies, is the design side: channels and channel width, frequency options (2.4, 5 and 6 GHz), SSIDs, antenna types and access point placement. Domain 2, Network Implementation, is 20% of the exam and the CompTIA page lists wireless devices under it. The whole exam is up to 90 questions in 90 minutes, mixing multiple-choice with performance-based items, so a signal question you can answer in ten seconds is time you get back for a drag-and-drop.
Want to see how those two objectives feel in question form? Our Network+ practice test splits them out by domain: Domain 5 troubleshooting questions and Domain 2 implementation questions.
Where the Bands Come In: 2.4, 5 and 6 GHz
You can't troubleshoot a signal reading without knowing which band it came from, because the bands fail in different ways.
2.4 GHz has range but no room. The Cisco RF Reference Guide puts it plainly: the band has only three channels that don't overlap, 1, 6 and 11. 20 MHz channels are the norm there, and bonding channels at 2.4 GHz isn't recommended or typically supported in enterprise networks. So when three access points in one office sit on channels 1, 3 and 6, two of them are stepping on each other. That's adjacent-channel interference, and the fix is the 1-6-11 plan, not more transmit power.
5 GHz has room but less reach. Per the same Cisco guide, its 20 MHz channels are spaced 20 MHz apart, none of them overlap, and the 802.11 standard defines 23 of them. The usable list depends on the regulatory domain you're in. The catch is DFS. Access points on the UNII-2 channels must use Dynamic Frequency Selection to protect radar systems, and if an AP detects radar it has to stop using that channel and move. From the client's side that looks like every device on the AP dropping at once and reconnecting a minute later on a different channel. On the exam, that's a client disassociation symptom with a legitimate cause.
6 GHz (Wi-Fi 6E and Wi-Fi 7) is the newest and the most regulated. Cisco's 6 GHz unlicensed spectrum white paper explains that low-power indoor access points run at a fixed power spectral density of 5 dBm per MHz. Their allowed EIRP therefore ranges from 18 dBm on a 20 MHz channel to 27 dBm on a 160 MHz channel. Standard-power operation tops out at 36 dBm EIRP (Cisco AFC FAQ). Translation for the exam: in 6 GHz, a narrow channel means a quieter radio, and cell sizes are smaller than people coming from 2.4 GHz expect.
Both the 1-6-11 question and the "which standard uses 6 GHz" question are the kind of thing the Domain 2 practice questions linked above cover, if you want to check yourself.
Coverage Holes, Sticky Clients and Roaming
Have you ever walked across an office with a laptop and watched it cling to the far access point while a closer one sat right above your head? That's a sticky client, and it's the single most useful real-world picture for objective 5.4's "roaming misconfiguration".
Roaming is the client's decision, not the access point's, and clients are conservative. Apple documents the rule for its own devices: a Mac keeps its current connection until the RSSI crosses -75 dBm. Even then, it only moves to a candidate access point whose RSSI is at least 12 dB stronger than the current one. So if your handoff zone sits at -75 dBm and the next AP is only offering -70 dBm there, the Mac stays put. The user sees a weak, slow connection in an area that the floor plan says has coverage.
Cisco's voice design guidance is built around avoiding exactly that. The VoWLAN troubleshooting guide puts the cell edge at -67 dBm and says a wireless handset should roam before the RSSI reaches that level. It recommends roughly 20% cell overlap on 2.4 GHz and 15 to 20% on 5 GHz, so there's always a stronger neighbor to move to. Design the overlap, and roaming takes care of itself. Skip it, and you've built a coverage hole with a signal bar still showing.
Two more things the exam folds into this objective:
- Roaming misconfiguration in the strict sense is access points that don't match. If the SSID, security type or passphrase differs between APs, a client can't hand off cleanly. It has to disconnect and authenticate from scratch, and any live session drops.
- Insufficient coverage versus unusable coverage. Cisco's site survey guidelines warn that the -67 dBm coverage area is much larger at 1 Mbps than at 12 Mbps. That is why they tell you to disable the low data rates during a 2.4 GHz survey. A heat map that looks fine at the lowest rate can be a coverage hole at the rate your applications need.
And a security aside, since Security+ candidates read this too: a rogue access point broadcasting your SSID at a stronger signal than the real one can win the roaming decision for nearby clients. That's an evil twin, and it's covered in our SY0-701 network attacks walkthrough.
How to Read Your Own Signal Strength
You'll remember the chart faster if you check your own laptop against it tonight.
Windows. Open a command prompt and run netsh wlan show interfaces. Microsoft's netsh wlan reference documents the command. The output's Signal field is a percentage, not dBm. To compare the access points around you, Microsoft's wireless troubleshooting guide points to netsh wlan show networks mode=bssid, which adds the BSSID, signal strength, channel and radio type for every network in range. The percentage is Windows' signal-quality value. Microsoft's WLAN API documentation defines 0% as -100 dBm and 100% as -50 dBm, with a straight line between them. So 40% works out to roughly -80 dBm (dBm ≈ quality ÷ 2 − 100). Treat it as an estimate, because adapters report it with varying accuracy.
macOS. Hold the Option key and click the Wi-Fi icon in the menu bar. Apple's Mac Help documents this as the way to open Wireless Diagnostics. The expanded menu also lists the current connection's RSSI and noise in dBm. Subtract one from the other and you've got your SNR without any extra software.
Anything else. A Wi-Fi analyzer app on a phone shows every nearby network's signal in dBm along with its channel, which is the quickest way to spot a 2.4 GHz channel pile-up. It's also a tool worth knowing by name for objective 5.5 (use the appropriate tool or protocol to solve networking issues).
Eight Drills With Answers
Cover the chart and answer these cold. These are our own practice scenarios written to the objective wording, not questions from any live exam. Answers follow each one.
- A technician is comparing four readings from a site survey: -30 dBm, -50 dBm, -65 dBm and -80 dBm. Which represents the best signal strength?
-30 dBm. Every dBm reading on a client is negative, so the one closest to zero is the strongest. -50 is also excellent, -65 is still above the -67 dBm voice floor, and -80 is basic connectivity at best. - Wireless clients at the far end of an office lose connectivity while clients near the access point are fine. A survey shows -85 dBm in the far area. What's the problem?
Insufficient wireless coverage, a coverage hole. -85 dBm sits between MetaGeek's -80 "basic connectivity" line and the -90 noise floor. The fix is placement or another access point, not a passphrase change or a captive portal setting. - A VoIP handset reports a signal of -62 dBm and a noise floor of -75 dBm. Calls are choppy. What's the most likely cause?
Interference. SNR is -62 minus -75, which is 13 dB, far below the 25 dB Cisco recommends for voice. The signal looks fine on its own. The noise is the problem. Find the source or move the AP to a cleaner channel. - Three access points in a dense office are set to 2.4 GHz channels 1, 3 and 6. Users complain about slow Wi-Fi. What should the technician do?
Move the APs to channels 1, 6 and 11. Those are the only non-overlapping 20 MHz channels in 2.4 GHz. Channel 3 overlaps both 1 and 6, so two of the three APs are interfering with each other. - Every client on a 5 GHz access point disconnects at the same moment, then reconnects about a minute later on a different channel. Nothing changed in the configuration. What happened?
A DFS event. The AP was on a UNII-2 channel, detected radar, and was required to vacate the channel. Clients were disassociated during the move. If it recurs, pick a non-DFS channel for that AP. - A MacBook stays connected to a distant AP at -72 dBm even though a nearer AP is available at -55 dBm. The user reports slow speeds. What's the best description of the issue?
A sticky client caused by cell design. Apple's documented behavior is to hold the current connection until RSSI crosses -75 dBm, so at -72 dBm the Mac hasn't started looking yet. The fix is design. Lower the far AP's transmit power so its signal falls past -75 dBm sooner, and size the cells so a candidate 12 dB stronger is always available near the -67 dBm edge. - A Windows user reports "Signal: 40%" from netsh and asks what that is in dBm. What's the right answer?
Windows defines the percentage as a straight-line scale from -100 dBm (0%) to -50 dBm (100%), so 40% is about -80 dBm: basic connectivity at best. It's only an estimate. For a precise dBm reading, use a Wi-Fi analyzer, or compare networks relative to each other withnetsh wlan show networks mode=bssid. - An administrator raises an AP's transmit power from 20 dBm to 23 dBm expecting to double the coverage distance. What actually happens?
Power doubles, distance doesn't. Three dB is twice the power, but free-space loss costs 6 dB per doubling of distance, so 3 dB extends range by roughly 41% (103/20 = 1.41), not 100%. The clients' own transmit power is unchanged, so their uplink doesn't improve at all, and the louder AP now reaches further into neighboring cells and adds co-channel interference.
FAQ
Is -50 dBm a good Wi-Fi signal strength?
Yes. It's well inside the range where signal strength isn't limiting anything. MetaGeek's scale puts -30 dBm as the maximum achievable and -67 dBm as the floor for voice and video, so -50 is comfortably between the two. If a connection at -50 dBm is slow, look at noise, channel overlap or the wired side, not the signal.
Is -70 dBm bad?
It's the edge. MetaGeek lists -70 dBm as the minimum for reliable packet delivery, enough for email and web browsing. It's below the -67 dBm line that MetaGeek and Cisco both use for voice and video, so a VoIP call at -70 dBm is living on borrowed time.
What SNR do I need for Wi-Fi?
Cisco's guidance is at least 25 dB for real-time traffic such as voice and video and 20 dB for data. With a noise floor at -92 dBm, which Cisco's VoWLAN guide treats as the ceiling for voice deployments, a -67 dBm signal gives you exactly 25 dB.
What's the difference between RSSI and dBm?
RSSI is a relative index. Per MetaGeek, the 802.11 standard allows a scale from 0 to 255 and each chipset vendor sets its own maximum, so RSSI from one adapter can't be compared with RSSI from another. dBm is an absolute measure of power referenced to one milliwatt. Most tools convert RSSI to dBm, which is why the two terms get mixed in everyday use.
Why does Windows show a percentage instead of dBm?
Because netsh wlan show interfaces reports Windows' 0 to 100 signal-quality value, which Microsoft's WLAN API maps from -100 dBm (0) to -50 dBm (100). That makes it a rough estimate: fine for "is it better here or there", too coarse for a design threshold like -67 dBm (about 66%). A Wi-Fi analyzer gives you the real dBm figure.
Make the Minus Sign Automatic
Here's the honest summary. Wi-Fi signal strength in dBm is one scale, six numbers and a subtraction. Closer to zero is stronger. -67 dBm is the voice line. Subtract the noise floor before you judge a reading. Three dB doubles the power, and every doubling of distance costs about 6 dB. Learn those and most of objective 5.4's wireless list, from coverage holes to sticky clients, becomes a reading-comprehension exercise.
Then make it reflexive. Run the eight drills again tomorrow without the chart, and take the free Network+ N10-009 practice test. The wireless questions arrive mixed in with routing, cabling and services, so you practice switching topics. When the troubleshooting domain feels comfortable, the next Network+ skills most people need are in our subnetting cheat sheet and the ports and protocols cheat sheet. Both are built the same way: one table, one method, drills with answers.