Step 1: The input is converted to binary, padded to the next 16-bit boundary.
Step 2: Each chunk is summed as a 16-bit integer.
Step 3: If the sum exceeds 16 bits, it wraps by adding the overflow back.
Step 4: The final sum is inverted to produce the checksum.