Data Link Layer Error Control

Applying Error Concepts Specifically at This Layer

The general error detection concepts covered earlier in this project (parity, checksums, CRC) are put into practice specifically at the Data Link layer, checking each individual frame as it moves across one single physical link.

CRC in the Frame Trailer

As covered in the previous framing note, a frame's trailer typically includes a CRC value covering the frame's contents. The receiving device recalculates this value upon arrival and compares it - a mismatch means the frame was corrupted somewhere during that specific hop, and it's simply discarded.

Why the Data Link Layer Typically Only Detects, Not Corrects

Rather than attempting to correct a corrupted frame in place, most Data Link layer implementations simply discard it entirely and leave any needed retransmission to a higher layer (like TCP at the Transport layer) to handle. This division of responsibility keeps individual layers focused on their own specific, narrow jobs.

Data Link layer:  detects a corrupted frame -> discards it silently
Transport layer:  notices missing data (e.g. via TCP's sequence numbers) -> requests retransmission

Why This Check Only Covers One Single Hop

A frame's CRC is recalculated and verified fresh at every single hop along a path - a packet's IP header (from the Network layer) survives unchanged across the entire journey, but its enclosing frame is entirely rebuilt at every router it passes through, since each hop is really its own separate Data Link layer transmission.

Why This Layered Separation of Concerns Is Valuable

Because each individual link independently checks for its own transmission errors, a single unreliable link somewhere along a path doesn't need to be "trusted" by every other layer above it - problems are caught and handled close to where they actually occur, one hop at a time.