IoT Attack Surface
IoT & Embedded Security · Embedded Systems & Firmware track
4 min read
In this module
Mapping the Full Attack Surface
An IoT device is a sprawling system spanning hardware, firmware, wireless protocols, cloud services, and mobile applications. Each component carries its own vulnerabilities, and the seams between them spawn extra attack surface. Real IoT security requires understanding and defending the whole ecosystem.
The IoT Attack Surface Model
A typical IoT device has five major attack surface areas:
1. Device Hardware: The physical device itself. Attack vectors include debug interfaces (UART, JTAG), bus sniffing (SPI, I2C), fault injection, side-channel analysis, and outright physical tampering. An attacker with physical access can extract firmware, read encryption keys, and rewrite device behaviour wholesale.
2. Firmware: The software running on the device. Vulnerabilities include hardcoded credentials, buffer overflows, command injection, insecure update mechanisms, and backdoors. Firmware is often written in C without any of the modern security guardrails you'd expect.
3. Wireless Protocols: The radio communication channels. Vulnerabilities include unencrypted traffic, weak pairing mechanisms, replay attacks, and protocol-level flaws. Each protocol (WiFi, BLE, Zigbee, LoRa, Z-Wave) brings its own security model and its own weaknesses.
4. Cloud Backend: The server infrastructure IoT devices communicate with. Vulnerabilities include insecure APIs, weak authentication, broken authorisation (device A can read device B's data), and outright data exposure.
5. Mobile Application: The smartphone app used to configure and control the device. Vulnerabilities include insecure local credential storage, missing certificate pinning, and API keys hardcoded into the binary.
The most impactful IoT attacks chain vulnerabilities across multiple surfaces. For example: sniff the BLE pairing to get the device's WiFi credentials, then use those credentials to access the local network, then exploit a vulnerability in the cloud API using the device's authentication token. Each step builds on access gained in the previous step.
Here's that chain drawn out across the radio, the home network and the cloud. Each numbered arrow crosses a trust boundary that some team assumed somebody else was guarding.
Hardware Attack Surface
Physical access to an IoT device enables:
- Firmware extraction: Through debug ports, flash dumping, or chip-off
- Key extraction: Reading encryption keys from flash, EEPROM, or OTP fuses
- Cloning: Duplicating the device's identity by extracting and copying its credentials
- Modification: Altering firmware to add backdoors or disable security features
- Fault injection: Bypassing secure boot or authentication through voltage/clock glitching
Many IoT devices live in physically accessible places: on doors, in hotel rooms, bolted to industrial equipment, perched on street corners. Your threat model has to assume the attacker can pick up the device and take it home to their lab.
Firmware Attack Surface
Common firmware vulnerability patterns in IoT:
- Web interface vulnerabilities: Command injection, XSS, CSRF, directory traversal in the device's management web interface
- Network service vulnerabilities: Buffer overflows and protocol flaws in daemons listening on the network
- Update mechanism flaws: Unsigned updates, downgrade attacks, predictable update URLs
- Hardcoded secrets: Credentials, private keys, API tokens compiled into the firmware
- Insecure defaults: Default admin passwords, unnecessary services enabled, debug interfaces active
Wireless Attack Surface
Each wireless protocol has specific attack vectors:
- WiFi: Deauthentication attacks, rogue access points, WPA key cracking, traffic sniffing on open networks
- BLE: Passive sniffing of advertising data, pairing interception, GATT profile enumeration, replay attacks
- Zigbee: Network key extraction during join, insecure rejoin, frame replay
- Z-Wave: S0 security downgrade, key exchange interception
- LoRaWAN: ABP key reuse, counter reset attacks, gateway impersonation
Systematic Attack Surface Mapping
A structured approach to mapping the attack surface:
For each interface (hardware, network, wireless, cloud, mobile): 1. Enumerate all entry points (ports, protocols, physical interfaces) 2. Identify the trust boundary (what is authenticated vs unauthenticated) 3. Map data flows (what data crosses each boundary) 4. Identify assets (keys, credentials, personal data, control commands) 5. Enumerate threats (who would attack, what they would gain) 6. Assess existing controls (encryption, authentication, authorisation) 7. Identify gaps (unprotected interfaces, weak controls)This systematic mapping produces a complete picture of the device's security posture and prioritises which areas need attention.
IoT security is defined by the weakest component. A device with strong encryption but default credentials is insecure. A device with proper authentication but unencrypted wireless communication is insecure. A device with hardened firmware but accessible JTAG is insecure. Security must be consistent across all attack surfaces.