The proposed model uses speech signals specifically as the carrier that carries the message bits through the network. The process is done using wavelet analysis of the speech signals. The input signal is sampled and decomposed into several wavelets which are encoded in the form of different echos introduced to it through displacement of the original wavelet with an echo signal obtained by a filter according to the message bit taken, i.e. 1 or 0. The message is considered to a binary message. Any other form is converted to binary first and then applied to the encoder block. The addition of echo is done on a pseudorandom basis which is obtained from the pseudorandom function generator which provides the specific points of adding the echo to the speech signal. The cycle is completed when all message bits are added to the speech and the decomposed wavelets are taken together to form the sampled output wave, which is reconstructed to form the output stego speech signal carrying the encoded message bits.

A similar process takes place on the decoder side of the network, where the carrier signal is decomposed back into wavelets and using the stego key, which in this case is the pseudorandom sequence, generated using the same function unique to this particular encoder-decoder pair. The specific positions of the wavelet coefficients are analyzed using the filters as used in the encoder and tested for the particular binary bit encrypted in that position. The cycle is repeated for all positions to obtain the total binary sequence of the message, which is converted to its original form as and where required.

The proposed design is dynamic and boasts of increased robustness, immunity to channel noise, high value of signal to noise ratio in comparison with other related designs and increased information hiding capacity and security.

The proposed design finds its application in the communication industry especially in the defence networking and communication sector where information security is the primary concern of any encoder-decoder system.

We claim:-
1. A wavelet based echo hiding encoder architecture consisting of the following:-
Architecture of a multimedia or text message to binary sequence converter for encoding.
Architecture of a sampler for sampling of speech signals at specific frequency for use as carrier of message signals.
Design of a predefined function based pseudorandom sequence generator, configured according to the speech signal input as carrier.
Architecture of encryption block for encryption of message signals in the speech input through equivalently fabricated echo filters for the wavelets.
Architecture of Reconstruction block for producing encoded stego output signal, from wavelet decomposed sampled wave.

2. The design covering Claim 1, where in architecture of the block for converting multimedia and text to binary message sequence.
3. The design covering Claim 1, where in architecture of the Sampler block is designed to sample the input speech signal at 44.2 KHz.
4. The design covering Claim 1, where in a predefined function based pseudorandom sequence generator is fabricated to produce a unique sequence depending on the length of the input speech signal.
5. The architecture of Claim 1, where in, the Encryption block is designed, covering the outputs of the architecture made in Claims 2, 3 and 4, to encode the messages.
6. The architecture of Claim 1, where in, the reconstruction block is designed specifically for the reconstruction of the encoded wavelet decomposed sampled wave at 44.2 KHz.

7. A decoder block architecture comprising of the following:-
A similar sampler block design as covered in Claim 1 and Claim 3, for sampling the input stego speech signal.
A similar pseudorandom generator block architecture as covered in Claim 1 and Claim 4 for reproducing the unique sequence for the decoder.
Architecture of decryption block for extracting binary message bits from the wavelet decomposed encoded speech signal.
Architecture of binary to multimedia converter for converting all forms of binary to the original message signal.

8. The design of Claim 7, where in, the architecture of the decryption is provided to decode the encoded speech signal from its sampled wavelet decomposed form.
9. The design of Claim 7, where in, the architecture of the binary to multimedia or text converter for producing the actual message signal encoded in the speech signal.
10. The architecture of the echo filter used for encoding the message signal bits for increasing robustness and noise immunity. , Description:The architecture proposed herein comprises of both, an encoder and a decoder block for transmission of message signals through speech signals as carrier. The encoder-decoder pair provided is unique as per the key to the encryption and decryption algorithm used herein.