Applications of signal processing techniques in direct-sequence spread spectrum communication systems

by Lee, Bong-Woon

Abstract (Summary)
This work investigates direct-sequence(DS) spread spectrum communication systems in conjunction with additional signal processing techniques to improve performance, in an environment where either strong narrowband interference or intentional interceptors exist. The objective of the signal processing techniques is to effectively reject any interference having a power level greater than the processing gain while concealing the DS signal so that it cannot be easily intercepted by eavesdroppers. The techniques which were used include scalar filtering in the WHT domain, and complementary filtering using an additionally generated reference signal in both the TDL and the lattice filters, either adaptively or non-adaptively. To further improve a capability to suppress interference, a reference signal generated from a proposed scheme of a reference generating loop is employed which consists of a pair of WHT matrices, a diagonal filter matrix, and a hard limiter. Although the reference loop using a scalar filtering method appears to be relatively more complicated than ones using the averaging method, it does not introduce a time-delayed reference signal due to the structure, and obtains a mean-square error (MSE) performance close to that of the optimal filter. Simulation results show that the reference signal contained an intermittent scattering of errors, instead of a burst of errors, during the data transition and its average chip error probability is lower than that of the reference signal generated from lowpass filter or chip decision. Such properties of the reference signal make it possible to use complementary filters to further suppress the interference. Since the reference signal is employed as a reference input to the complementary filters, any distortion of the desired signal due to chip-pulse dispertion is not introduced and weight cycling does not occur in the steady state. As a result, it was shown that complementary filters have better performance regardless of the interference type when the chip error probability (P c) of the reference signal is less than 0.25 for SNR=20dB, SIR=-20dB, P c=0, and filter order of 4. The SINR improvement of the complementary filters was better by at least 10dB for single-tone interference and 3.9dB for narrowband interference than that of the other suppression filters reported in the literature. The one sided complementary filter is implemented in the latttice structure. Two types of complementary lattice filters are proposed: forward and backward complementary lattice filters. Simulation results show that the theoretical performance of the complementary filters was the same in both the TDL and the lattice forms. However, when adaptive algorithms such as the LMS and GAL were applied to the complementary filters, it is shown that the forward complementary lattice filter had the lowest average MSE, the fastest convergence rate, and was the most stable in the steady state. A FIR-FIR method in connenction with the reference loop was employed to enhance the AI performance of the DS system and its detectability was quantitatively analyzed for radiometer detection. While the FIR-FIR method sufficiently suppressed the PSD of the DS signal and gave the PN sequence a multi-level random amplitude, it did not cause any loss in performance due to inverse filtering. The simulation results showed that this method somewhat lessened the DC component due to an unbalanced code which not only caused a deterioration in the performance of the DS system but also provided a possibility of detection by intentional interceptors. In addition, it was shown that the detection probability was sufficiently reduced relative to the increase in the degree of power suppression and the energy-to-noise spectral density ratio.
Bibliographical Information:


School:Ohio University

School Location:USA - Ohio

Source Type:Master's Thesis

Keywords:signal processing techniques direct sequence spread spectrum communication systems fir method


Date of Publication:01/01/1990

© 2009 All Rights Reserved.