Matched filtering in improper complex noise and applications to DS-CDMA

Based on the criterion of maximizing signal to noise ratio (SNR), the optimum matched filter is, in general, the solution to a real integral equation. With regards to the detection of narrowband, bandpass signals, we present the full form of the integral equation based upon complex envelopes. The correct form of the integral equation includes a complementary or pseudo-autocorrelation function which can be ignored for most matched filtering applications. However, in the presence of improper complex noise, overlooking the function leads to a "non-optimal" matched filter. To illustrate the concept, we apply the theory to a simple direct-sequence code-division multiple-access (DS-CDMA) system.

[1]  A. J. Jerri Introduction to Integral Equations With Applications , 1985 .

[2]  James L. Massey,et al.  Proper complex random processes with applications to information theory , 1993, IEEE Trans. Inf. Theory.

[3]  J. M. Holtzman,et al.  DS/CDMA successive interference cancellation , 1994, Proceedings of IEEE 3rd International Symposium on Spread Spectrum Techniques and Applications (ISSSTA'94).

[4]  H. Urkowitz Signal Theory and Random Processes , 1983 .

[5]  Robert N. McDonough,et al.  Detection of signals in noise , 1971 .

[6]  D. W. Tufts,et al.  Statistical Communication Theory , 1958 .

[7]  Laurence B. Milstein,et al.  A noise-whitening approach to multiple access noise rejection .I. Theory and background , 1994, IEEE J. Sel. Areas Commun..

[8]  Harry Leib,et al.  Matched filters for locked and unlocked interferers in DS-CDMA , 1995, IEEE Pacific Rim Conference on Communications, Computers, and Signal Processing. Proceedings.

[9]  C. Helstrom,et al.  Statistical theory of signal detection , 1968 .