CAN (Controller Area Network)

📘 CAN (Controller Area Network)

1) Concept

CAN is a multi-node serial communication bus designed for reliability and real-time messaging in embedded systems. It is used in automotive ECUs, industrial control, robots, drones, medical devices, and aerospace.

Multiple devices share one bus and communicate using message IDs.

Unlike UART or SPI, CAN is not point-to-point — it is a shared bus network.

---

2) Physical Layer (Differential Signaling)

CAN uses two wires:

Differential signaling provides:

• High noise immunity

• Long cable tolerance

• Stable operation in harsh environments

This is why CAN is strong in automotive + industrial fields.

---

3) Message-Based (No Device Addresses)

CAN does not use device addresses. Instead, messages are identified by message IDs.

Example:

All nodes receive all messages, and each node decides whether to accept or ignore based on ID filtering.

This is publish-subscribe communication at the hardware level.

---

4) Arbitration (Collision-Free Bus Access)

If multiple nodes transmit at the same time, CAN resolves conflicts using bitwise arbitration:

• IDs are compared bit-by-bit

• Lower ID value = Higher priority

• Higher priority frame continues; others automatically back off

Example:

This ensures real-time guarantee:

Critical messages always win the bus.

---

5) Data Structure

A CAN frame includes:

[ID][Control][Data (0–8 bytes)][CRC][ACK][EOF]

For Classical CAN, max payload is 8 bytes.

For CAN FD (Flexible Data Rate), payload can be:

8, 12, 16, 24, 32, 48, 64 bytes

and data phase bitrate can be much faster.

---

6) Data Rates

If you’re doing new development, CAN FD is the correct choice.

---

7) Why CAN Is Strong

In short:

CAN = Real-time, reliable, hardened communication bus.

---

8) Example Topology

     Node A
       |
Node B─+─Node C──── Termination (120Ω) at both ends
       |
     Node D

Only the ends of the bus have 120Ω terminators.

---