Automotive Gateway Reference Architecture: Choosing Between Domain and Zonal Routing with Standards-Aligned Interoperability

The core architecture decision for a vehicle gateway is whether to keep signal-based CAN/CAN FD routing (ISO 11898-1:2015) in a central domain gateway or move to zonal controllers that translate legacy signals into service-oriented Ethernet traffic (AUTOSAR SOME/IP, IEEE 802.1Q time-sensitive networking with IEEE 802.1AS-2020 time synchronization). The trade-off is deterministic, well-validated legacy routing versus reduced harness weight and software-defined flexibility, at the cost of a bigger security and timing verification effort. Diagnostics (ISO 14229-1:2020 UDS over ISO 13400-2:2019 DoIP), in-vehicle message authentication (AUTOSAR SecOC, IEEE 802.1AE MACsec), and cybersecurity/update governance (ISO/SAE 21434:2021, UNECE R155/R156) should be fixed as interface contracts before topology is frozen.

Decision criteria (domain vs. zonal)

Keep a central domain gateway when most ECUs are legacy CAN/CAN FD with stable signal databases and end-to-end latency budgets are already validated; move to zonal + Ethernet backbone when you have several high-bandwidth sources (cameras, radar, central compute), plan frequent OTA feature delivery under UNECE R156, and can validate TSN scheduling (IEEE 802.1Q traffic shaping) and gPTP time sync (IEEE 802.1AS) on the target silicon.

Pre-implementation checklist

Collect the full signal/PDU inventory with cycle times and ASIL levels, the worst-case bus load per CAN segment, the list of safety-relevant routes needing E2E protection (AUTOSAR E2E profiles), the freshness/key-management model for SecOC, DoIP entity and logical addressing per ISO 13400-2, the TARA outputs required by ISO/SAE 21434, and the AUTOSAR release (Classic and Adaptive) each supplier actually supports, so routing tables and service interfaces are generated from one source of truth.

Illustrative scenario and recovery (analytical example, not measured data)

A platform with 6 CAN FD segments at 2 Mbit/s data phase and 4 cameras migrates to 4 zonal controllers on a 1 Gbit/s Ethernet backbone. The rollout keeps a fallback signal-gateway path for powertrain and braking routes during the first program phase, adds zones one at a time behind feature flags, and defines a safe-state rule: if gPTP sync is lost or SecOC verification fails beyond a configured threshold, the zone falls back to the last validated static routing table, logs the event for the vehicle SOC, and blocks OTA activation until diagnostics confirm recovery.

How do we verify that a zonal gateway migration keeps legacy CAN behavior while adding SOME/IP services and security?

Treat the legacy signal database as the reference contract: generate the zonal routing and SOME/IP service mappings from it, then run back-to-back tests that compare signal timing, cycle and content on the legacy bus against the translated path under worst-case bus load. Check that E2E protection and SecOC freshness hold across the CAN-to-Ethernet boundary, that gPTP sync and TSN schedules meet the latency budget on the real hardware, and that UDS over DoIP reaches every ECU through the new topology. Recovery has to be designed in: keep a validated static fallback routing table, define safe states per ASIL route, and record the evidence in the ISO/SAE 21434 cybersecurity case and the UNECE R156 software update management records before you enable OTA.

Back to AutomotiveGateway · Domains

← Back to AutomotiveGateway
Engineering & Architecture Brief

Automotive Gateway Reference Architecture: Choosing Between Domain and Zonal Routing with Standards-Aligned Interoperability

The core architecture decision for a vehicle gateway is whether to keep signal-based CAN/CAN FD routing (ISO 11898-1:2015) in a central domain gateway or move to zonal controllers that translate legacy signals into service-oriented Ethernet traffic (AUTOSAR SOME/IP, IEEE 802.1Q time-sensitive networking with IEEE 802.1AS-2020 time synchronization). The trade-off is deterministic, well-validated legacy routing versus reduced harness weight and software-defined flexibility, at the cost of a bigger security and timing verification effort. Diagnostics (ISO 14229-1:2020 UDS over ISO 13400-2:2019 DoIP), in-vehicle message authentication (AUTOSAR SecOC, IEEE 802.1AE MACsec), and cybersecurity/update governance (ISO/SAE 21434:2021, UNECE R155/R156) should be fixed as interface contracts before topology is frozen.

Decision criteria (domain vs. zonal)

Keep a central domain gateway when most ECUs are legacy CAN/CAN FD with stable signal databases and end-to-end latency budgets are already validated; move to zonal + Ethernet backbone when you have several high-bandwidth sources (cameras, radar, central compute), plan frequent OTA feature delivery under UNECE R156, and can validate TSN scheduling (IEEE 802.1Q traffic shaping) and gPTP time sync (IEEE 802.1AS) on the target silicon.

Pre-implementation checklist

Collect the full signal/PDU inventory with cycle times and ASIL levels, the worst-case bus load per CAN segment, the list of safety-relevant routes needing E2E protection (AUTOSAR E2E profiles), the freshness/key-management model for SecOC, DoIP entity and logical addressing per ISO 13400-2, the TARA outputs required by ISO/SAE 21434, and the AUTOSAR release (Classic and Adaptive) each supplier actually supports, so routing tables and service interfaces are generated from one source of truth.

Illustrative scenario and recovery (analytical example, not measured data)

A platform with 6 CAN FD segments at 2 Mbit/s data phase and 4 cameras migrates to 4 zonal controllers on a 1 Gbit/s Ethernet backbone. The rollout keeps a fallback signal-gateway path for powertrain and braking routes during the first program phase, adds zones one at a time behind feature flags, and defines a safe-state rule: if gPTP sync is lost or SecOC verification fails beyond a configured threshold, the zone falls back to the last validated static routing table, logs the event for the vehicle SOC, and blocks OTA activation until diagnostics confirm recovery.

Technical FAQ: How do we verify that a zonal gateway migration keeps legacy CAN behavior while adding SOME/IP services and security?

Treat the legacy signal database as the reference contract: generate the zonal routing and SOME/IP service mappings from it, then run back-to-back tests that compare signal timing, cycle and content on the legacy bus against the translated path under worst-case bus load. Check that E2E protection and SecOC freshness hold across the CAN-to-Ethernet boundary, that gPTP sync and TSN schedules meet the latency budget on the real hardware, and that UDS over DoIP reaches every ECU through the new topology. Recovery has to be designed in: keep a validated static fallback routing table, define safe states per ASIL route, and record the evidence in the ISO/SAE 21434 cybersecurity case and the UNECE R156 software update management records before you enable OTA.

Related Architecture Specifications (AUTOMOTIVEGATEWAY): gatewaycan.com · automotivecompute.com · vehiclegateway.com

More Technical Articles on automotivegateway.com

Why Technical Domains Decide Authority in the AI Era →CAN Gateway Filtering, Routing and Diagnostics →Software-Defined Vehicle Reference Stack →ISAC and V2X Convergence for Cooperative Mobility →Centralized E/E Architecture: Consolidating the Vehicle →Zonal Architecture: Wiring, Power and the Ethernet Backbone →Automotive High-Performance Computing (HPC) →Automotive Ethernet and TSN: The Deterministic In-Vehicle Backbone →