Automotive networks are carrying more data than they did when classical CAN was first introduced.
Electronic control systems have become more complex, vehicles contain more ECUs, and modern functions require greater communication capacity between controllers.
CAN-FD was developed to address some of these communication requirements.
It increases the payload capacity of a CAN frame and allows a higher bit rate during the data phase while retaining the core CAN arbitration concept.
But there is an important distinction that engineers and security teams should not overlook.
More bandwidth does not automatically mean more security.
CAN-FD improves communication capability. Security still depends on the architecture, protocols, ECUs, gateways and additional security mechanisms built around it.
This article examines what changes from classical CAN to CAN-FD, what security risks remain, how CAN-FD traffic can be assessed and where organizations can begin building a practical security assessment process.
What Is CAN-FD?
CAN-FD stands for Controller Area Network with Flexible Data Rate.
It extends classical CAN by allowing larger payloads and a faster data phase.
One widely cited implementation characteristic is a payload capacity of up to 64 bytes compared with the 8-byte payload limit of classical CAN. CAN-FD can also use a higher bit rate during the data phase, while arbitration remains at the nominal bit rate. ([ScienceDirect](https://www.sciencedirect.com/science/article/abs/pii/S016740482300113X))
This makes CAN-FD useful for automotive systems that need to move more information without immediately abandoning the CAN ecosystem.
CAN vs CAN-FD at a Glance
| Characteristic | Classical CAN | CAN-FD |
|---|---|---|
| Maximum payload | Up to 8 bytes | Up to 64 bytes |
| Data phase speed | Single nominal rate | Higher data-phase rate possible |
| Arbitration | CAN arbitration | CAN arbitration retained |
| Communication model | Broadcast bus | Broadcast bus |
| Built-in cryptographic authentication | No | No |
| Built-in encryption | No | No |
| Security extensions | Possible through higher layers | Possible through higher layers |
The final two rows are particularly important.
CAN-FD is an improved communication protocol, not a complete cybersecurity architecture.
Why Was CAN-FD Needed?
Vehicle electronic architectures continue to generate more data.
Applications such as advanced driver assistance, diagnostics, body control, powertrain management and other electronic systems can place greater demands on in-vehicle communication.
Classical CAN’s 8-byte payload limitation can make some communication workloads inefficient.
CAN-FD addresses this by allowing a larger data field and a faster data phase.
That engineering improvement, however, creates an important cybersecurity question:
What happens when the same underlying trust assumptions are carried into a higher-capacity network?
Does CAN-FD Fix the Security Problems of CAN?
No.
CAN-FD does not inherently provide cryptographic sender authentication, confidentiality or message-level security.
Research on CAN-FD continues to identify security concerns involving spoofing, replay, fuzzing and denial-of-service attacks. ([PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC11175007/))
The larger payload provides more room for security information, but the protocol itself does not automatically determine how authentication or encryption should be implemented.
A recent security architecture study similarly describes CAN-FD as having limited intrinsic security and evaluates additional authentication and encryption mechanisms rather than treating CAN-FD itself as a complete security solution. ([ScienceDirect](https://www.sciencedirect.com/science/article/abs/pii/S016740482300113X))
The Core Security Problem
Consider a simplified CAN-FD network:
Gateway
|
+--------+--------+
| | |
v v v
ECU A ECU B ECU C
\ | /
\ | /
+------CAN-FD------+
|
ECU D
Every connected node participates in the shared communication environment.
If an unauthorized node obtains access to the relevant network, simply knowing a legitimate message identifier does not provide cryptographic proof that the message originated from the legitimate ECU.
This is one reason authentication mechanisms may be required for security-critical communication.
CAN-FD Attack Surface
The CAN-FD bus should not be viewed in isolation.
The actual attack surface can include the systems that provide a path toward the bus.
- Diagnostic interfaces
- Compromised ECUs
- Telematics systems
- Infotainment systems
- Wireless interfaces
- Vehicle gateways
- Development interfaces
- Service equipment
- Connected applications
The presence of an interface does not establish that a vehicle is remotely exploitable. The security assessment must establish whether the interface can actually reach the relevant network and what controls exist along the path.
Four Important CAN-FD Attack Classes
- Message injection
- Spoofing or masquerading
- Replay
- Denial of service
Research has also examined fuzzing and other anomalous message patterns against CAN-FD networks. ([PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC11175007/))
1. CAN-FD Message Injection
Message injection occurs when an unauthorized source introduces additional frames into the network.
Because CAN-FD retains the CAN arbitration and broadcast communication model, an attacker with appropriate network access can attempt to transmit frames that appear structurally similar to legitimate traffic.
The security impact depends heavily on the affected identifier, transmission rate, ECU behavior and the presence of authentication or gateway controls.
2. CAN-FD Spoofing
Spoofing attempts to make unauthorized communication appear to originate from a legitimate source.
Without a mechanism that cryptographically verifies message origin, the CAN identifier alone should not be treated as proof of ECU identity.
This becomes particularly important when a message can influence a safety-relevant or security-relevant function.
3. CAN-FD Replay Attacks
A replay attack uses previously observed communication and transmits it again.
A message can be genuine when originally transmitted and still become dangerous when reused outside its intended context.
This is why secure communication designs often need a concept of message freshness in addition to message authenticity.
AUTOSAR SecOC, for example, uses authentication mechanisms together with freshness-related information to protect critical PDU communication. ([AUTOSAR](https://www.autosar.org/fileadmin/standards/R22-11/CP/AUTOSAR_SWS_SecureOnboardCommunication.pdf))
4. CAN-FD Denial of Service
CAN arbitration is designed to prioritize messages based on their identifiers.
An attacker able to transmit repeatedly using high-priority identifiers can potentially consume bus capacity and interfere with legitimate communication.
Research on CAN-FD intrusion detection explicitly examines denial-of-service scenarios in which high-frequency high-priority messages disrupt normal communication. ([PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC11175007/))
CAN-FD Fuzzing
Fuzzing sends systematically varied or unexpected inputs to a target in order to identify abnormal behavior or implementation weaknesses.
In an automotive laboratory, CAN-FD fuzzing can be used to test ECU handling of unexpected frames, payload patterns or protocol conditions.
Fuzzing should be conducted only within an authorized and controlled test environment because unexpected ECU behavior can affect system stability.
CAN-FD Traffic Analysis
One of the most practical security activities is understanding what normal CAN-FD communication looks like.
A traffic capture can provide useful information about:
- CAN-FD identifiers
- Timestamps
- Frame frequency
- Payload length
- Payload patterns
- Bus utilization
- Message bursts
- Unexpected identifiers
The exact fields available depend on the capture hardware, software and logging format.
Why a CAN-FD Baseline Is Important
An anomaly is meaningful only in relation to expected behavior.
For example, a message appearing at an unusual frequency might indicate an attack, but it could also be legitimate behavior during diagnostics, startup or a specific operating state.
A professional analysis therefore records the context of each capture.
- Vehicle or bench state
- Ignition state
- Diagnostic activity
- Relevant ECU state
- Test procedure
- Time of capture
CAN-FD Security Metrics
| Metric | Security Value |
|---|---|
| Identifier frequency | Helps establish normal communication patterns. |
| Inter-arrival time | Helps identify timing anomalies. |
| Payload length | Helps identify unexpected frame structures. |
| Payload behavior | Helps identify unusual data patterns. |
| Bus utilization | Helps identify congestion and flooding. |
| Identifier appearance | Helps identify unexpected communication. |
| Message bursts | Can indicate abnormal transmission behavior. |
CAN-FD Intrusion Detection
Intrusion detection systems can analyze CAN-FD traffic and identify deviations from expected communication behavior.
Research has explored machine learning and sequence-based approaches for detecting multiple CAN-FD attack classes, including replay, denial of service, fuzzing and spoofing. ([PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC11175007/))
However, a high laboratory detection accuracy should not automatically be interpreted as production readiness.
An automotive IDS must also consider:
- False positives
- False negatives
- Computational resources
- Latency
- Dataset quality
- Changing vehicle states
- Real-world traffic variability
The difference between a research model and an operational automotive security control is therefore significant.
Does CAN-FD Need Encryption?
Not every CAN-FD message necessarily requires encryption.
Security requirements should be derived from the threat model, data sensitivity, safety implications, architecture and applicable requirements.
For some applications, authenticity and integrity may be more important than confidentiality.
For others, confidentiality may also be relevant.
A recent CAN-FD security architecture study demonstrates one possible approach using encryption and authentication, but this represents a proposed security model rather than a universal requirement for every CAN-FD implementation. ([ScienceDirect](https://www.sciencedirect.com/science/article/abs/pii/S016740482300113X))
Authentication vs Encryption
| Security Property | Question |
|---|---|
| Authentication | Did an authorized source generate this message? |
| Integrity | Was the message modified? |
| Freshness | Is this message current rather than an old replay? |
| Confidentiality | Can an unauthorized party read the protected data? |
| Availability | Can the communication system continue operating under attack? |
These are different security properties.
Adding encryption does not automatically solve every automotive communication security problem.
CAN-FD and AUTOSAR SecOC
CAN-FD security can be strengthened through security mechanisms implemented above the underlying communication technology.
AUTOSAR Secure Onboard Communication, or SecOC, is one example.
The AUTOSAR specification describes SecOC as a resource-efficient mechanism for authentication of critical data at the PDU level. It is designed to integrate with existing AUTOSAR communication systems and primarily considers symmetric authentication using message authentication codes. ([AUTOSAR](https://www.autosar.org/fileadmin/standards/R22-11/CP/AUTOSAR_SWS_SecureOnboardCommunication.pdf))
This is important because it demonstrates the difference between the communication bus and the security architecture built around that bus.
CAN-FD Is Not SecOC
These technologies should not be treated as interchangeable.
| Technology | Primary Role |
|---|---|
| CAN | Vehicle communication protocol. |
| CAN-FD | Higher-capacity CAN communication. |
| SecOC | Authenticated communication mechanism within AUTOSAR. |
| ISO/SAE 21434 | Automotive cybersecurity engineering lifecycle. |
| IDS | Detection of suspicious communication or behavior. |
A mature automotive cybersecurity architecture can use several of these technologies and processes together.
CAN-FD and Automotive Gateways
Gateways can play an important role in controlling communication between different vehicle networks.
Instead of allowing unrestricted communication between every network, a gateway can enforce routing and security policies.
Research into automotive TSN and CAN-FD security gateways has examined routing mechanisms and security controls for communication between network technologies. ([SAE International](https://saemobilus.sae.org/papers/routing-security-mechanisms-design-automotive-tsn-fd-security-gateway-2022-01-0113))
This makes gateway analysis an important part of a broader CAN-FD security assessment.
CAN-FD Security Assessment Methodology
A useful assessment should begin with architecture rather than immediately generating attack traffic.
Phase 1: Define Scope
- Identify the vehicle or ECU platform.
- Identify CAN-FD networks in scope.
- Identify permitted test activities.
- Define safety restrictions.
- Define data handling requirements.
Phase 2: Architecture Review
- Identify ECUs.
- Identify gateways.
- Identify external interfaces.
- Identify diagnostic paths.
- Identify security boundaries.
Phase 3: Traffic Collection
Collect authorized CAN-FD traffic from a laboratory bench, development environment, test vehicle or customer-provided capture.
Record the operating context for each capture.
Phase 4: Baseline Development
- Identify normal identifiers.
- Measure transmission frequency.
- Measure timing patterns.
- Analyze payload lengths.
- Document normal state transitions.
Phase 5: Security Analysis
- Review message injection exposure.
- Review replay protection.
- Review spoofing resistance.
- Review denial-of-service exposure.
- Review gateway controls.
- Review authentication mechanisms.
- Review key and freshness management where applicable.
Phase 6: Controlled Validation
Where authorized, selected findings can be validated in a controlled laboratory or test environment.
Live vehicle testing should be approached separately because unexpected communication can affect vehicle behavior.
Phase 7: Reporting
Document the evidence, affected systems, security impact, limitations and recommended remediation.
What Should a CAN-FD Security Assessment Examine?
| Area | Questions |
|---|---|
| Architecture | How are CAN-FD networks connected? |
| ECUs | Which nodes communicate? |
| Identifiers | Which IDs are legitimate? |
| Timing | What are normal transmission intervals? |
| Payloads | What data patterns are expected? |
| Gateways | Which messages can cross network boundaries? |
| Authentication | How is message origin verified? |
| Freshness | How are replay attempts addressed? |
| IDS | How is abnormal traffic detected? |
| Diagnostics | Which diagnostic paths can access the network? |
Tools for CAN-FD Security Analysis
The toolchain should match the assessment environment.
| Tool Category | Potential Use |
|---|---|
| CAN-FD interface | Capture and transmit authorized traffic. |
| CAN analysis software | Inspect and decode frames. |
| Wireshark | Analyze supported captures and higher-layer protocols. |
| Python | Automate log processing and anomaly analysis. |
| SocketCAN | Linux-based CAN/CAN-FD interface and testing. |
| Oscilloscope | Investigate physical-layer behavior when required. |
| ECU test bench | Provide a controlled validation environment. |
The important capability is not the number of tools. It is the ability to produce reproducible evidence from the data collected.
Can Wireshark Analyze CAN-FD?
Wireshark can be useful for supported CAN-related capture formats and higher-layer protocols carried over automotive networks.
However, it should be treated as one component of the analysis workflow rather than a complete CAN-FD penetration testing platform.
A serious assessment may combine CAN-FD interfaces, CAN-specific analysis software, Linux tools, custom scripts, ECU documentation and controlled laboratory equipment.
Can CAN-FD Security Be Assessed Remotely?
Yes, when the assessment is designed around data and systems that can be accessed remotely.
Remote activities can include:
- CAN-FD log analysis
- Identifier inventory
- Traffic baseline development
- Timing analysis
- Anomaly investigation
- Security architecture review
- Gateway review
- IDS evaluation
- Security documentation review
- Risk reporting
This can be valuable for engineering teams that already operate a test bench or development environment but need independent cybersecurity analysis.
What a Remote CAN-FD Assessment Can Deliver
| Deliverable | Purpose |
|---|---|
| CAN-FD traffic profile | Document observed communication behavior. |
| Identifier inventory | Map observed CAN-FD identifiers. |
| Timing analysis | Identify unusual transmission patterns. |
| Anomaly report | Document traffic requiring investigation. |
| Attack surface review | Identify possible paths toward the network. |
| Security control review | Evaluate authentication and network controls. |
| Risk register | Prioritize findings. |
| Remediation plan | Translate findings into engineering actions. |
CAN-FD Security Assessment vs CAN Security Assessment
| Assessment | Focus |
|---|---|
| CAN security assessment | Classical CAN communication. |
| CAN-FD security assessment | CAN-FD communication and associated controls. |
| ECU security assessment | Individual ECU interfaces and implementation. |
| Gateway assessment | Segmentation and inter-network communication. |
| Automotive network assessment | Multiple vehicle communication technologies. |
In a modern vehicle, these assessments can overlap.
CAN-FD and Higher-Layer Protocols
CAN-FD should not be analyzed only at the frame level.
Higher-layer protocols can introduce additional security considerations.
- ISO-TP
- UDS
- XCP
- J1939
- Manufacturer-specific protocols
These protocols can expose additional functionality and therefore require their own threat analysis.
Current automotive cybersecurity testing guidance highlights ISO-TP, UDS and XCP as additional attack surfaces above the raw CAN/CAN-FD frame layer, particularly because diagnostic and measurement functionality can expose state changes, memory access and other powerful operations. ([CAN in Automation](https://www.can-cia.org/fileadmin/cia/documents/publications/cnlm/june_2026/cnlm_26-2_p31_automotive_cybersecurity_can_and_lifecycle_testing_dr_nils_weiss_dissecto_dr_matthias_pukrop_dspace.pdf))
Why CAN-FD Security Requires a Layered Approach
No single security mechanism addresses every CAN-FD threat.
A stronger architecture may combine:
- Network segmentation
- Secure gateways
- Message authentication
- Freshness protection
- Secure diagnostics
- ECU hardening
- Secure boot
- Secure software updates
- Intrusion detection
- Key management
The correct combination depends on the vehicle’s threat model and security requirements.
Common CAN-FD Security Mistakes
- Assuming CAN-FD is secure because it is newer than CAN.
- Assuming a larger payload automatically provides authentication.
- Treating the CAN identifier as proof of message origin.
- Ignoring gateway security.
- Ignoring diagnostic protocols.
- Building an IDS without a reliable traffic baseline.
- Testing only during one vehicle operating state.
- Ignoring key and freshness management.
- Assuming a research IDS accuracy figure represents production performance.
- Testing a live vehicle without appropriate safety controls.
How to Build a CAN-FD Security Baseline
A useful baseline should document both technical behavior and operational context.
CAN-FD Network
|
+-- ECU Inventory
|
+-- CAN-FD ID Inventory
|
+-- Message Frequency
|
+-- Timing Profile
|
+-- Payload Length
|
+-- Normal Operating States
|
+-- Gateway Rules
|
+-- Authentication
|
+-- Freshness
|
+-- IDS
|
+-- Diagnostic Access
Once this baseline exists, future security testing becomes more measurable.
CAN-FD Security Assessment Checklist
| Area | Question | Status |
|---|---|---|
| Architecture | Are CAN-FD networks clearly documented? | ☐ |
| ECUs | Are communicating nodes identified? | ☐ |
| Identifiers | Are expected CAN-FD IDs documented? | ☐ |
| Traffic | Is normal behavior baselined? | ☐ |
| Injection | Is unauthorized frame injection addressed? | ☐ |
| Spoofing | Is message origin authenticated? | ☐ |
| Replay | Is message freshness protected? | ☐ |
| DoS | Is flooding exposure understood? | ☐ |
| Gateway | Are network boundaries enforced? | ☐ |
| Diagnostics | Are diagnostic paths protected? | ☐ |
| IDS | Can abnormal traffic be detected? | ☐ |
| Testing | Is validation performed in a controlled environment? | ☐ |
Where CAN-FD Fits in the Automotive Security Stack
Vehicle Cybersecurity
|
+----------------------+
| |
v v
Cybersecurity Lifecycle Technical Controls
| |
v +-- ECU Security
ISO/SAE 21434 +-- Secure Boot
| +-- Gateway Security
| +-- SecOC
| +-- IDS
v +-- Secure Diagnostics
Threat Analysis +-- CAN / CAN-FD Security
and Risk Assessment
CAN-FD security is therefore one technical component within a larger automotive cybersecurity program.
What Engineers Should Ask Before a CAN-FD Assessment
- Which CAN-FD networks are in scope?
- Which ECUs transmit on each network?
- Which identifiers are safety or security relevant?
- Which gateways connect the network to other domains?
- Which external interfaces can reach the network?
- Are messages authenticated?
- How is freshness managed?
- Are diagnostic services protected?
- What monitoring exists?
- What happens when abnormal traffic is detected?
Final Perspective
CAN-FD solves an important engineering problem: classical CAN’s limitations in payload capacity and data-phase speed.
It should not, however, be mistaken for a cybersecurity solution.
The security of a CAN-FD network depends on what surrounds the protocol.
ECUs, gateways, diagnostics, authentication, freshness protection, intrusion detection, key management and secure software practices all contribute to the final security posture.
The most useful assessment therefore starts with evidence.
Capture the traffic. Understand the baseline. Map the communication. Identify the attack surface. Evaluate the controls. Validate findings safely.
CAN-FD provides more communication capability. Security still has to be engineered.
Need a CAN-FD Security Assessment?
Cautivra provides remote and laboratory-oriented automotive cybersecurity analysis for organizations working with CAN and CAN-FD networks.
Depending on project scope, the assessment can use customer-provided CAN-FD logs, architecture documentation, ECU information, laboratory captures or an authorized test environment.
Potential assessment areas include traffic analysis, identifier profiling, timing analysis, anomaly investigation, gateway review, security-control assessment and technical reporting.
Request a CAN-FD Security Assessment
Frequently Asked Questions
Is CAN-FD more secure than CAN?
CAN-FD improves payload capacity and data-phase speed, but it does not inherently add cryptographic message authentication or encryption. Security depends on additional controls and the overall vehicle architecture.
Can CAN-FD be hacked?
CAN-FD can be exposed to security attacks when an attacker obtains an appropriate path to the network or a connected ECU. Research has examined injection, spoofing, replay, fuzzing and denial-of-service attacks. Actual exploitability depends on the specific vehicle architecture and security controls.
Does CAN-FD support encryption?
Encryption can be implemented as part of a CAN-FD security architecture, but CAN-FD itself does not inherently provide encryption. The appropriate security mechanism depends on the system’s requirements and architecture.
What is the difference between CAN and CAN-FD security?
The fundamental security issue remains similar: neither classical CAN nor CAN-FD inherently provides complete cryptographic authentication or encryption. CAN-FD provides a larger payload and faster data phase, which can provide additional design flexibility for security mechanisms.
Can CAN-FD traffic be monitored for attacks?
Yes. Intrusion detection systems can analyze identifiers, timing, frequency, payload behavior and other traffic characteristics. Research has demonstrated multiple approaches for detecting CAN-FD anomalies, but production performance must be evaluated against realistic traffic and operating conditions.
Can CAN-FD security be assessed remotely?
Yes. Remote assessment can include CAN-FD log analysis, traffic profiling, architecture review, anomaly investigation, gateway review, security-control analysis and reporting. Physical testing requires access to an appropriate authorized test environment.
What tools are used for CAN-FD security analysis?
Depending on the environment, engineers may use CAN-FD interfaces, CAN analysis software, Linux SocketCAN, Wireshark, Python-based analysis tools, oscilloscopes and ECU test benches. The toolchain should be selected according to the assessment objective.
What is the role of AUTOSAR SecOC?
AUTOSAR SecOC provides a mechanism for authenticated communication at the PDU level and is designed to integrate with AUTOSAR communication systems. It can provide security properties such as authentication and protection against replay when correctly implemented and configured.
Sources and Further Reading
- AUTOSAR: Specification of Secure Onboard Communication
- SAE International: Automotive Cybersecurity Research
- ScienceDirect: Security Architecture for Automotive Communication Networks with CAN FD
- Scientific research on multi-attack intrusion detection for CAN-FD
- CAN in Automation: Automotive cybersecurity and lifecycle testing