you're reading...
Deep Packet Analysis, Hacking / Counter Intelligence, Lawful Intercept

Audio Steganography PT3: Understanding MP3Stego’s Technique

MP3Stego’s Information-hiding method was the research result of the Fraunhofer-IIS Institute. The MP3 audio encoding process is shown in Figure 1.

The important parts that carry out the MP3 encoding are the two loops. The inner loop is a quantization and encoding loop. If the bits produced by quantization and encoding exceed the available max_length bits, the quantization step is increased to get a smaller quantization value. This operation is repeated continuously, until the Huffman coding bits are short enough. The out loop is a noise control loop. It controls the quantization noise according to the threshold value and adjusts the scale factor.

When the digital audio signal is quantized and encoded, quantization error appears. This error is an indetermination value, for example adopting different psycho-acoustic models results in Proceedings of the 10th WSEAS International Conference on COMPUTERS, Vouliagmeni, Athens, Greece, July 13-15, 2006 (pp368-371)different error. And this error can be adjusted in the process of quantization and encoding. The MP3Stego uses the block length after encoding to hide the secret information by adjusting the quantization error. Odd length means 1 and even length means 0. Thus the secret information is concealed in the MP3 bit stream. MP3Stego was carried out on the base of the 8hz-mp3 encoder. The hiding process could be explained by Figure 2.

static int inner_loop( … ) 
{ … 
  do { 
    quantizerStepSize += 1.0; 
    bits=quantize( ); 
switch (hiddenBit)  { 
 case 2: 
   embedRule = 0; break; 
 case 0: case 1: 
embedRule = ((bits + part2length) % 2) != hiddenBit; 
  ERROR("inner_loop: unexpected hidden 
  } while((bits>max_bits) | embedRule); 
  return bits; 

In the inner iteration loop of quantization and encoding, the hiding rule is added to the quantization rules. If the part23 length after being quantified doesn’t obey the hiding rule (odd represents 1 and even represents 0) or quantization rules, the quantization error is adjusted and the encoding process is repeated.


About Pythorian

Exploration and Production oriented security consultant for securing IT infrastructures relating to natural resources.


No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: