Technical detail ยท Practice: ISA100.11a Security
What each layer defends against
What Each Layer Defends Against
The two layers stop different attacks. DLL hop-by-hop security means every relay checks the integrity of each frame it receives before forwarding, so an attacker cannot inject or alter frames on a single radio link without detection. Hop-by-hop alone would still let a compromised router read and rewrite the payload it forwards, because it necessarily holds the link key.
TL end-to-end security, with integrity and confidentiality enabled, closes that gap. The payload is protected with a session key shared only by the true source and destination, so an intermediate router forwards ciphertext it cannot decrypt and cannot forge. Put together, DLL stops link-level tampering and TL stops a malicious relay from reading or altering the data. A compromised middle node cannot read or forge that protected payload, although it can still disrupt forwarding.
The attack model separates four kinds of evidence.
| Scenario | Expected defensive evidence |
|---|---|
| Outsider injects a frame near the pipe rack | The receiving neighbor rejects it because the hop integrity check fails. |
| Routing device is allowed to forward but later becomes suspect | The device may still pass frames, but it lacks the end-to-end session key needed to read or alter the payload. |
| Gateway translation is wrong | Neither DLL nor TL proves the application value is correct; the lab needs gateway diagnostics and timestamp evidence. |
| Retired device tries to rejoin | Admission policy and revocation evidence reject it before transport security becomes relevant. |
A routed, tunneled example exposes why diagnostics matter. Suppose a maintenance request produces a two-fragment tunneled response across field device, router, and gateway. The wireless path might report per-hop retries, the tunnel might report a translation error, and the plant application might show only a stale value. Security review should require all three views to line up: frame integrity or retry counters at the wireless layer, tunnel and gateway status at the boundary, and stale-value handling at the application.
The conservative conclusion is that cryptography gives the deployment a defendable boundary, but the lab must prove the boundary is operated correctly. A release-ready ISA100.11a security lab shows accepted joins, denied joins, role assignment, key renewal, revocation, routed-path diagnostics, gateway translation evidence, and retest triggers.