dsss and correlation

Question

Generally, a DSSS (Direct Sequence Spread Spectrum) communication is used to secure a communication. However, because of the repetition of the spreading sequence, a cyclostationary analysis makes it possible to locate the communication. Moreover, there are also methods which make it possible to find the spreading sequence.

To prevent finding the content of the communication, can another spreading sequence which is known only to the sender and the receiver be used in addition to the spreading sequence used for the communication and which is not used for transmission? This would make communication completely secure. In fact, on reception, it would suffice to use the 2 spreading sequences to find the message.

In mathematical terms: $c_1(t)$ is the spreading sequence of the message and $c_2(t)$ is the sequence which serves as a key for the sender and the receiver. The transmitted signal is given by: $s(t)=\sum_{k=1}^{N_s}d_kc_1(t-kT_s)$ with $N_s$ the number of symbols and $T_s$ the duration of the sequence of spreading. At the reception, we use a matched filter with the 2 sequences: $r(t) \ast c_1^{\ast}(-t)$ and $r(t) \ast c_2^{\ast}(-t)$ to find the binary data.

I would like your opinion on this question.

Answer 1

The GPS Y-code is an example of this as the combination of the P-code with a W encryption code. However the main point is the use of cryptographically secure codes: If the spreading sequence is cryptographically secure and known only to the sender and receiver of the message, then a second sequence is not needed and the link will be secure with DSSS alone. The issue or concern with DSSS occurs when codes are generated with linear feedback in which case the spreading sequence can be easily determined with only a short and error free portion of the sequence, as detailed further in this link.