Deauthentication Attack Explained: How It Works and PMF Defense thumbnail

Deauthentication Attack Explained: How It Works and PMF Defense

⏱ approx. 24 min views 50 likes 0 LOG_DATE:2026-05-11
TOC

Deauthentication Attack #

A Deauthentication Attack (often Deauth Attack / De-auth) sends spoofed IEEE 802.11 Deauthentication management frames (Subtype 0x0C) to forcibly disconnect Wi-Fi clients from their AP, or vice versa. Hardware: any cheap Wi-Fi card with monitor-mode + packet-injection support. Software: open-source tools — aircrack-ng / mdk4 / wifite. Despite being one of the oldest known Wi-Fi attacks, as of 2026 it is still routinely exploitable on a huge installed base.

The root cause is a 1997 IEEE 802.11 design decision: management frames are sent unencrypted and unauthenticated. Data frames have been progressively protected (WEP → WPA → WPA2 → WPA3), but management frames that handle connection lifecycle — Authentication, Association, Deauthentication, Disassociation — sat in the clear for over two decades. IEEE 802.11w (Protected Management Frames, PMF) was finalized in 2009 and made mandatory in WPA3 in 2018, finally fixing the spec — yet most legacy devices, older corporate APs, and public Wi-Fi still leave PMF disabled, keeping a huge attackable surface.

This article frames Deauth as a building block of larger attacks — outright DoS, WPA2 handshake capture, Evil Twin redirection, and band-downgrade — and walks through how the frame works (SVG 1), the four typical scenarios (SVG 2), the standard tools, the major real-world incidents, the technical fix via 802.11w PMF, and the legal framing under Japan's Radio Act §109(2) and the US Communications Act §333.

1. 802.11 frames and the historical flaw in management frames #

The 802.11 MAC layer defines three frame classes:

Class Role Encryption (WPA2 era)
Data Frame Actual user data (HTTP / TCP, etc.) Encrypted under WPA2
Control Frame Low-level RTS / CTS / ACK None
Management Frame Beacon / Probe / Association / Deauthentication / Disassociation Plaintext under WPA2 (without PMF)

The problem is the Management Frame class. Frames that control the connection lifecycle (Authentication / Association / Deauthentication / Disassociation) were specified to be plaintext and unauthenticated from 1997 onward, and that stayed unchanged for years.

The consequence: an attacker who can transmit on the right channel can impersonate the AP or the client, send spoofed Deauth frames, and:

  • Disconnect any active client
  • Make the AP believe a client has left
  • After disconnection, catch the 4-way handshake when the client auto-reconnects, or lure the client to an Evil Twin AP

No prior credentials, WEP key, or WPA passphrase are needed — being in radio range is enough. That's what makes the Deauth Attack so decisively powerful.

2. Structure of a Deauth frame — why plaintext is enough #

At byte level, a Deauthentication frame is small:

Structure of an 802.11 Deauthentication frame Frame Control / 4 MAC addresses / Sequence / Reason Code — only 26 bytes total Frame Control Type=Management Subtype=0x0C Duration 2 bytes Address 1 (DA) Destination 6 bytes Address 2 (SA) Source 6 bytes ★ forgeable Address 3 (BSSID) 6 bytes Sequence Control 2 bytes Reason Code 2 bytes — the entire payload FCS (CRC32) 4 bytes — integrity only, not tamper-resistant ▼ Common Reason Codes (defined in the 802.11 standard) 1 — Unspecified reason / generic, frequently used in attacks 2 — Previous authentication no longer valid 3 — Sending STA leaving (deauth/disassoc) 7 — Class 3 frame received from nonassociated STA ▼ Why plaintext / unauthenticated — the history ① 1997 — original 802.11 spec: management frames "are just control," no encryption needed Only data frames are protected (by WEP). The spec authors didn't anticipate attacks abusing the management plane. ② 2003 WPA / 2004 WPA2: AES-CCMP protects data, management frames still plaintext If an attacker spoofs the MAC, APs and clients accept the Deauth as legitimate control traffic. ③ 2009 IEEE 802.11w (PMF) — management frames finally protected BIP (Broadcast Integrity Protocol) and CCMP-MFP authenticate Deauth / Disassoc with a MIC. ④ 2018 WPA3 — PMF is mandatory for Wi-Fi Alliance certification With PMF enabled, Deauth attacks are neutralized — 21 years after the original flaw shipped. In practice, most installed base runs WPA2 / Mixed Mode; public Wi-Fi and old corporate APs commonly leave PMF off.

The key vulnerability: Address 2 (Source Address) is arbitrary, so the attacker can impersonate either the AP or the client. With a Wi-Fi card in monitor mode + injection-capable, the tooling (Linux with iw / aircrack-ng) can transmit hundreds to thousands of spoofed Deauth frames per second.

3. Scenarios — Deauth as a building block, not just a DoS #

Most people hear "Deauth Attack" and think DoS, but in practice it's the entry move of a larger attack:

Four typical Deauth Attack scenarios

DoS / WPA2 4-way handshake capture / Evil Twin redirection / 5GHz→2.4GHz downgrade

① Pure DoS — continuous Wi-Fi disruption

Attacker: broadcast Deauth flood

All clients in range thrash between disconnect and reconnect

Electronic locks / IP cameras / IoT devices can fail open or shut

Tools: aireplay-ng / mdk4 / wifite

② WPA2 4-Way Handshake capture → dictionary attack

Attacker: one Deauth → client auto-reconnects → 4-way handshake is captured

capture.pcap → offline dictionary attack with hashcat / aircrack-ng

Short WPA2-PSK passwords (under ~10 chars) crack readily

Deauth is the trigger that forces the handshake to happen

③ Evil Twin / Rogue AP redirect

Attacker spins up a rogue AP with the same SSID, then Deauths clients off the real AP

Clients auto-reconnect to the rogue AP (stronger signal)

Attacker MITMs all traffic — SSL Strip / captive portal for credentials

Tools: airbase-ng / WiFi-Pumpkin / EAPHammer

④ 5GHz → 2.4GHz downgrade

Attacker Deauths the client off its 5GHz AP

Client rolls back to 2.4GHz — same SSID or attacker's rogue 2.4GHz AP

2.4GHz reaches further and through obstacles — attacker can stay distant

Especially effective when 5GHz has PMF but 2.4GHz doesn't

▼ Scenario ②: handshake-capture sequence

[Attacker] Put mon0 on the target AP's channel / use airodump-ng to learn the client's MAC

[Attacker → Client] aireplay-ng --deauth 1 -a [BSSID] -c [client MAC] sends one targeted Deauth

[Client] Auto-reconnects → completes the 4-way handshake (M1 → M2 → M3 → M4)

[Attacker] handshake.pcap → hashcat -m 22000 for dictionary / GPU cracking

WPA2-PSK with a weak password falls in hours-days; WPA3 SAE resists offline dictionaries

Strategic role of each scenario:

  • (1) DoS: deliberately taking down Wi-Fi-dependent physical security / IP cameras / electronic locks (legitimate only as a pentest scenario test)
  • (2) Handshake capture: keeps WPA2-PSK cracking viable today — Deauth is the necessary step to force a fresh handshake
  • (3) Evil Twin: the on-ramp for credential phishing on public Wi-Fi — Deauth gets the victim off the legitimate AP so they reconnect to yours
  • (4) Downgrade: an advanced move — even when WPA3 / PMF is in place, forcing a fallback to an unprotected band or mode preserves the attack surface

In real-world incidents, Deauth standing alone as the goal is the minority case — it's the first move of a bigger compromise.

4. Tools — readily available open source #

Standard on Kali Linux / Parrot OS:

Tool Purpose
aircrack-ng suite Industry standard — airmon-ng (monitor mode) / airodump-ng (scan) / aireplay-ng (Deauth injection) / aircrack-ng (cracking)
mdk3 / mdk4 More aggressive — large-scale Deauth Flood / Beacon Flood / Authentication Flood
wifite / wifite2 An aircrack-ng front-end — single command runs Deauth + handshake capture + crack
scapy Build custom 802.11 frames in Python — great for learning and bespoke Deauths
bettercap General-purpose MITM framework with a Wi-Fi module that includes Deauth
airbase-ng / hostapd-mana / WiFi-Pumpkin Build Evil Twin / Rogue APs (paired with Deauth)

Basic aireplay-ng Deauth injection:

# (1) Put the Wi-Fi card into monitor mode
sudo airmon-ng start wlan0
# → produces wlan0mon

# (2) Identify the target AP's channel / BSSID / client MAC
sudo airodump-ng wlan0mon
# → AP BSSID = XX:XX:XX:XX:XX:XX / Channel = 6 / Client = YY:YY:YY:YY:YY:YY

# (3) Lock to that channel and capture
sudo airodump-ng -c 6 --bssid XX:XX:XX:XX:XX:XX -w capture wlan0mon

# (4) In a second terminal, inject 10 Deauth frames
sudo aireplay-ng --deauth 10 -a XX:XX:XX:XX:XX:XX -c YY:YY:YY:YY:YY:YY wlan0mon

# (5) Verify the handshake is in capture-01.cap, then crack it
sudo aircrack-ng -w wordlist.txt capture-01.cap

scapy to build a custom Deauth frame (educational):

from scapy.all import *

bssid = "XX:XX:XX:XX:XX:XX"
client = "YY:YY:YY:YY:YY:YY"

# Deauth frame (subtype 12 = Deauthentication)
pkt = RadioTap() / \
      Dot11(type=0, subtype=12, addr1=client, addr2=bssid, addr3=bssid) / \
      Dot11Deauth(reason=7)

sendp(pkt, iface="wlan0mon", count=10, inter=0.1)

Legal note: running these against networks you don't own violates Japan's Radio Act and Unauthorized Computer Access Act (and in the US, Communications Act §333). Only use them against your own AP / home LAN / explicitly contracted pentest targets — covered below.

5. Real-world incidents — when Deauth caused real damage #

Year Incident What happened / lesson
2014 Marriott Hotel ($600,000 FCC fine) Marriott Gaylord Opryland (Nashville) deliberately Deauth-flooded guests' personal Wi-Fi hotspots to force them onto its paid Wi-Fi. FCC fined the chain $600,000 for "commercial Wi-Fi jamming." Established that disabling guest hotspots is illegal even on hotel property
2014 Hilton-related Wi-Fi jamming ($25,000 FCC fine) Same pattern at Hilton properties — Deauth on MiFi / personal hotspots to push guests onto paid Wi-Fi. $25,000 fine + cooperation order sent a clear message to the industry
2015 Smart-City IoT incidents Smart-city pilots saw outdoor IP cameras and Wi-Fi-controlled street lights knocked offline by Deauth tests — revealing that Wi-Fi-bound control planes don't belong in critical infrastructure
2016-2020 Credential theft on public Wi-Fi Multiple reports of Evil Twin + Deauth combinations at airports, cafes, and hotel lobbies capturing social / email credentials. Part of the urgency behind Wi-Fi Alliance mandating WPA3
2017 KRACK disclosure Mathy Vanhoef's WPA2 KRACK vulnerability works standalone, but real-world exploitation typically uses Deauth to force reconnection → KRACK to abuse the nonce reuse. Renewed attention to Deauth as the trigger for other attacks
2018- Physical-security research Office IP cameras / electronic locks / access-control systems running on Wi-Fi were repeatedly shown to fail under Deauth in seconds, reinforcing the view that Wi-Fi-dependent physical security is fragile

The pattern: Wi-Fi-dependent IoT and physical security need to be designed assuming Deauth can take them down at any time. That's why modern devices often add PoE Ethernet or 4G/5G cellular failover paths.

6. Defense — 802.11w (PMF) and detection #

The only technical root-cause fix is enabling 802.11w (Protected Management Frames, PMF):

Control Detail
Enable PMF (802.11w) Both AP and client must support and enable PMF — Deauth frames carry a MIC, and tampering is detected and dropped
Adopt WPA3 WPA3-certified gear mandates PMF, so Deauth Attacks are automatically neutralized
PMF Required mode Use "Required" rather than "Capable" — non-PMF clients are refused
WIDS / anomaly detection Wireless IDS flags "high Deauth rate from a single MAC" — built into Aruba / Meraki / Cisco APs
RF shielding Extreme but used in military / SCIF / certain data centers (Faraday cage, RF-absorbing paint)
Hardwire critical devices IP cameras / electronic locks / printers — anything where downtime has physical impact — should run on PoE Ethernet
Continuous monitoring Periodic airodump-ng / Kismet sweeps for unknown APs and unusual Deauth activity

AP-side PMF (example: hostapd):

# /etc/hostapd/hostapd.conf
wpa=2
wpa_key_mgmt=WPA-PSK
ieee80211w=2   # 0=Disabled / 1=Capable / 2=Required
# For WPA3
# wpa_key_mgmt=SAE
# ieee80211w=2 (SAE requires it)

Client-side (Linux NetworkManager):

# /etc/NetworkManager/system-connections/MyAP.nmconnection
[wifi-security]
key-mgmt=wpa-psk
pmf=3   # 0=default / 1=disable / 2=optional / 3=required

Windows 11, macOS, iOS, and Android all support PMF by default in recent versions — once the AP requires it, clients negotiate it automatically.

Operational obstacles:

  • Old APs (more than ~5 years) sometimes have no PMF support and never will, even with firmware updates
  • Public Wi-Fi prioritizes compatibility and tends to ship with PMF Optional, which Deauth still bypasses
  • WPA2 + PMF Required breaks old IoT devices that don't support PMF — an existing-asset compatibility problem
  • Enterprises often have 2.4GHz legacy clients without PMF, requiring split SSID strategies

The litmus test for real Deauth defense is "can we set PMF Required?"

7. Detection — what a Wireless IDS sees #

A WIDS is the easiest place to catch a Deauth attack:

Signal What it suggests
Large bursts of Deauth from a single BSSID Broadcast Deauth Flood (e.g., mdk4)
Repeated Deauths to one client MAC Targeted Deauth — usually a handshake-capture preface
Multiple BSSIDs beaconing the same SSID Possible Evil Twin
Anomalously strong signal nearby Attacker may be physically close
Plaintext Deauth in a PMF-required environment Definitely forged — dropped due to MIC failure but worth logging

Open-source WIDS / monitoring:

  • Kismet (the standard, Linux, free) — Deauth flood / rogue AP / anomaly detection
  • WAIDPS — built by the wifite author
  • Snort + wireless preprocessor

Commercial:

  • Cisco CleanAir / Aruba RFProtect / Meraki Wireless Health — enterprise APs with detection + automatic countermeasures (RF containment against rogue APs)

SOC integration: feed Kismet logs into Splunk / Elastic, alert on "more than 100 Deauths/minute on one SSID." That's the modern operating model.

8. The legal side — Japan's Radio Act and the US FCC #

Japan:

  • Radio Act §109(2) ("interfering with the function of radio equipment"): emitting radio waves with intent to interfere with another's communicationsup to 1 year imprisonment or a ¥1,000,000 fine
  • Radio Act §4 (licensing): operating above Wi-Fi-band power limits in pursuit of jamming may also count as unlicensed operation
  • Unauthorized Computer Access Act: Deauth → Evil Twin → credential theft escalates into §4 (unauthorized access) and §5 (facilitation) territory
  • Obstruction of Business by Force (Penal Code §234): deliberately jamming Wi-Fi at commercial premises — up to 3 years imprisonment or a ¥500,000 fine

United States:

  • Communications Act of 1934, §333: interfering with another's licensed communications is a federal violation
  • The Marriott / Hilton fines were issued under this provision
  • FCC clarified in 2015 that interfering with personal Wi-Fi hotspots is illegal even on hotel or corporate property

For pentest / research, written authorization is non-negotiable:

  • An AP you own / your own home LAN — legal
  • Inside a customer's Rules of Engagement with written scope — legal
  • Anything else — illegal

There is a wide gap between "technically easy" and "legally permitted" here. The single most important rule when working with this attack is "only on your own AP, or within a signed pentest contract."


A Deauthentication Attack exploits a 1997-era design choice — 802.11 management frames are unencrypted and unauthenticated — to forcibly disconnect Wi-Fi clients. The flaw was technically fixed in 2009 with 802.11w (PMF) and mandated in 2018 with WPA3, yet most installed gear runs without PMF, leaving the attack viable in 2026.

Deauth is rarely the goal by itself — it's the building block for WPA2 4-way handshake capture (offline cracking) / Evil Twin redirection (credential theft on public Wi-Fi) / band-downgrades. Hardware costs a few dollars and the tooling (aircrack-ng, mdk4, wifite) is open source, so the on-ramp is extremely low, and real-world damages — Marriott / Hilton FCC fines, credential theft on public Wi-Fi, IoT physical-security failures — keep accumulating.

The single root-cause defense is PMF in Required mode plus migration to WPA3. Legally, it's covered by Japan's Radio Act §109(2) and the US Communications Act §333 — running these tools against any network you don't own is straightforwardly illegal. The only safe places to experiment are your own AP and contracted pentest targets.