ASG vs LIN

LIN exists because CAN was too expensive for a door mirror. It is deliberately minimal: one wire, one master, a fixed poll schedule, 20 kbit/s.

The cost of that minimalism shows up at vehicle scale, as dozens of small LIN clusters, each with its own master ECU and its own wiring, hanging off the CAN backbone.

At a glance

LIN

The cheapest way to network a handful of slow devices: one master polls each slave in turn over a single wire. Mirrors, window switches, rain sensors, small motors.

ASG

One ring for the whole vehicle. A window motor and a headlight sit on the same channel, in the same 1 ms window, with no sub-bus of their own.

Topology

LIN network

ASG single ring

LIN: one master polls each slave in turn over a single wire. ASG: every body node is a zone on the one ring, read every millisecond.

How they differ

Network structure

LIN
A separate LIN cluster per function group, each with a master node that also bridges to CAN.

ASG
No sub-buses. Every body node is a zone on the one ring.
Media access

LIN
The master polls slaves on a fixed schedule; a slave cannot speak until it is asked.

ASG
Every zone has its own carrier and is read every millisecond.
Latency

LIN
Best case is one poll period; worst case is a full schedule cycle, and the bus is only 20 kbit/s.

ASG
1 ms to the shared commit, the same for a latch as for a lamp.
Wiring and parts

LIN
Cheap per run, but every cluster adds a master, a connector set and a wire run.

ASG
One ring, and control blocks that are the same part wherever they sit.
Diagnostics

LIN
Each master reports its own slaves; whole-vehicle body state is stitched together in software.

ASG
Body state is part of the committed vehicle state, timestamped to the millisecond.

Common ground

Both are low-cost per node and tolerate simple, inexpensive slave devices.
Both are schedule-driven rather than a free-for-all: LIN by its poll table, ASG by its 1 ms grid.
Both are aimed at functions where a few milliseconds of latency is perfectly acceptable.

The matrix

LIN and ASG, side by side.
Dimension	LIN	ASG
What it is	A low-cost communication sub-bus for slow body functions.	A control architecture: protocol, real-time OS and ring channel as one system.
Topology	One master with several slaves on a single shared wire.	Single ring routed through the vehicle volume; zones attach within a wire-length budget.
Medium	A single dedicated wire referenced to ground. Power routed separately.	Power and data on the same conductors (PLC over 12 / 24 / 48 V), or a twisted pair.
Media access	The master polls each slave in turn. No arbitration.	Frequency-division. Each zone has its own carrier (a diad); no contention for the channel.
Timing model	Schedule-table driven by the master; slot period typically a few ms.	One global commit every 1 ms. Every zone updates in the same instant.
Determinism under load	Bounded by the schedule, but latency is a whole poll cycle and the bus is slow.	Fixed. Every event lands on the 1 ms grid regardless of how many zones or events.
Raw data rate †	Up to 20 kbit/s.	~2.7 Mbps on the shared channel; capacity added by carriers, not by frame time.
Vehicle-wide state	None. The master holds a local view of its own slaves only.	Native. Every module holds the whole committed state vector each millisecond.
Compute model	Sequential, master-driven.	Distributed and time-parallel: every module computes on the same state in the same window.
Harness impact †	Cheap per run, but many separate LIN runs add up across the body.	One ring carrying power and data. ~200-300 m of wiring per vehicle in ASG's own figures.
Fault handling	Minimal. The master detects a missing or malformed slave response.	10 us local reaction in each block acts as an electronic fuse; no relays needed.
Standardisation & maturity	ISO 17987. Mature and ubiquitous for body electronics.	ISO-compliant seven-layer protocol; pre-series. Standard number and assessments still to be cited.
Tooling & ecosystem	Broad and inexpensive; part of every automotive toolchain.	Early. A public simulator; silicon, stacks and test tools are in development.

† derivation still to be confirmed. Legacy figures are the nominal values from the published standards; ASG figures are set out on the methodology page.

Where LIN is the better choice today

Absolute cost floor
For one or two slow signals, a LIN slave is about as cheap as a networked device gets: a transceiver of a few cents and a single wire.
Simplicity
A LIN cluster is trivial to specify and to debug. There is very little to get wrong.
Off-the-shelf parts
Every body-electronics supplier ships LIN slave devices today, ready to drop in.
Independence
A LIN cluster can be developed and validated in isolation. ASG brings the node into the shared vehicle model instead.

Questions

Is LIN’s slow speed the problem ASG fixes?

For its job, 20 kbit/s is usually enough. What ASG removes is the separate network and the master node, not just the speed limit.

Would ASG be overkill for a single door module?

On ASG a door module is just another zone on the ring it is already connected to, so there is no extra network to add. The comparison is against a whole vehicle’s worth of LIN clusters, not one.

Can existing LIN devices attach to ASG?

Yes. The zone module runs the LIN schedule itself, polling the existing slaves and folding their responses into the shared vehicle state, then replying in LIN where a slave expects it. No separate LIN master or gateway box is added, the same as CAN.

Open the simulator Full comparison matrix