How is the IP header checksum computed?

• A. Take 1's complement of each 16-bit field and then add each together
• B. XOR all 16-bit words
• C. Sum 8-bit words and invert
• D. Use CRC-32 over the header

Answer: (A)

The IP header checksum is computed by treating the header as a sequence of 16-bit words and combining them using 1's complement arithmetic, then inverting the final sum to form the 16-bit checksum. During calculation, the checksum field is zeroed, and after adding all 16-bit words (with end-around carry when needed), the result is inverted to produce the checksum stored in the header. This method detects errors in the header by ensuring that the total of all 16-bit words, including the checksum, sums to all ones.

The option describing taking the 1's complement of each 16-bit field and then adding them aligns with using 1's complement arithmetic to combine the header words, which is why it is the best choice. The other options don’t describe this 16-bit, ones-complement summation process (XOR, 8-bit summing with inversion, or CRC-32).