G correction of memory errors when fetching data from

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: = 101, then we should correct the message bit with index 5, i.e., d2 . This corrective action is exactly the one described in the earlier table we built by inspection. SECTION 6.5. PROTECTING LONGER MESSAGES WITH SEC CODES 15 Figure 6-7: Dividing a long message into multiple SEC-protected blocks of k bits each, adding parity bits to each constituent block. The red vertical rectangles refer are bit errors. The Hamming code’s syndrome calculation and subsequent corrective action can be efficiently implemented using digital logic and so these codes are widely used in contexts where single error correction needs to be fast, e.g., correction of memory errors when fetching data from DRAM. ￿ 6.5 Protecting Longer Messages with SEC Codes SEC codes are a good building block, but they correct at most one error. As messages get longer, they are unlikely to provide much correction if we use the entire message as a single block of k bits. The solution, of course, is to break up a longer message into smaller blocks of k bits each, and to protect each one with its own SEC code. T...
View Full Document

Ask a homework question - tutors are online