Know-how · Diagnostics
DoIP (ISO 13400)
The same diagnostic services, moved onto Ethernet — and a discovery handshake that is responsible for most of the trouble people have with it.
In short
DoIP (Diagnostics over Internet Protocol) is defined by ISO 13400 and carries diagnostic traffic — normally UDS — over automotive Ethernet instead of CAN. The tester finds the vehicle through a discovery step using vehicle identification messages over UDP, then opens a TCP connection and performs routing activation, which authorises the tester and binds its source address. Only after routing activation succeeds can diagnostic messages be exchanged. Each ECU behind the gateway has a logical address, and the gateway routes messages to it; the tester never addresses an ECU on the Ethernet layer directly. DoIP removes the frame-size and bandwidth limits that make CAN flashing slow, which is why it is the normal transport for large software downloads. It does not change the diagnostic services themselves — a UDS request is the same request whether it arrives over CAN or DoIP.
01
Discovery, routing activation, then diagnostics
A DoIP session has a shape that CAN does not. First the tester has to find the vehicle: it sends a vehicle identification request over UDP, or waits for the vehicle announcement the gateway broadcasts after link-up. The response carries the VIN, the logical address of the gateway and its entity identifiers.
Then the tester opens a TCP socket and sends a routing activation request carrying its own source address and an activation type. The gateway either accepts it or refuses with a code — already activated by another tester, unknown source address, missing authentication. Nothing diagnostic happens until this succeeds, and a tester that skips straight to a UDS request on a fresh socket simply gets nothing back.
After activation, UDS requests are wrapped in a DoIP header carrying source and target logical addresses, and the gateway routes them onward to the addressed ECU.
- Vehicle announcementGateway broadcasts after link-up; the easiest way to be found.
- Vehicle identificationUDP request/response carrying VIN, logical address and EID.
- Routing activationTCP, authorises the tester and binds its source address.
- Diagnostic messageUDS payload with source and target logical addresses.
02
Why a DoIP tester finds nothing
Almost every DoIP connection problem is in the first two steps rather than in diagnostics. The link may not have come up in the right mode — activation line not asserted, or the gateway still in a state where the diagnostic port is inactive. The tester may have no address on the vehicle subnet, or be on the wrong one. The announcement may have been broadcast before the tester was listening, and the tester may never send an identification request of its own.
Routing activation adds its own failure modes. Another tester may hold the session. The source address may not be one the gateway accepts. Some gateways require authentication before activation and refuse silently-looking responses otherwise. A tester that reports "no response from ECU" is usually reporting a socket that was never activated, not an ECU that is not there.
- Check link stateActivation line and gateway diagnostic port state before anything else.
- Check addressingTester IP on the vehicle subnet; correct logical source address.
- Check activation responseA refusal code is far more informative than a timeout.
- One tester at a timeAn already-active routing session will refuse the second.
03
What DoIP changes, and what it does not
It changes throughput. A software download that takes a long time over CAN with ISO-TP segmentation completes far faster over Ethernet, which is the main reason DoIP exists and the main reason it appears in production flashing. It also changes topology: a gateway and logical addressing replace the flat broadcast of a CAN bus.
It does not change the diagnostic model. Sessions, security access, data identifiers, routines and fault memory behave as ISO 14229 describes. If you understand a UDS exchange on CAN you understand it on DoIP; what you additionally need is the discovery and activation sequence in front of it.
Standards & references
Check the source.
These primary sources define the formats and regulatory context discussed above. They do not certify or endorse VehicleLab products.
Questions
Frequently asked.
What does DoIP stand for?
Diagnostics over Internet Protocol. It is specified in ISO 13400 and carries diagnostic communication — usually UDS — over automotive Ethernet.
Is DoIP a replacement for UDS?
No. DoIP is a transport. UDS is the diagnostic service layer that rides on it. A UDS request is the same request whether it arrives over CAN or over DoIP.
What is routing activation?
A mandatory handshake over TCP in which the tester presents its source address and an activation type, and the gateway authorises the connection. Until it succeeds, diagnostic messages are not routed.
Why is DoIP used for flashing?
Ethernet removes the frame-size and bandwidth constraints of CAN with ISO-TP segmentation, so a large software download completes in a fraction of the time.
Can DoIP and CAN diagnostics coexist on one vehicle?
Yes, and they commonly do. A gateway may expose the same ECUs over both, with DoIP used for high-volume work such as programming and CAN retained for other functions.
Next step
Open one of your own files with us.
Send a measurement, an A2L or an EHANDBOOK container. We will open it in the instruments on a call, with your engineers in the room.