Supression of Narrow-band and Wide-band Interference in Direct-seqyence Spread Spectrum Systems
It is well known that the narrow - band (NB) interference rejection capability of a direct - sequence (DS) spread spectrum system can be enhanced by passing received signal through a linear prediction error filter prior to correlating it with PN correlator. Recently, nonlinear filters have also been proposed to solve the problem. It has been shown that the nonlinear filters can have better performance compared to the linear ones. However, the channel effect, which always exists in the communication channel, was not taken into account. If we consider the channel effect, the received PN sequence will no longer be white and binary. Instead, the PN sequence will approach a colored Gaussian process. In this case, the performances of the linear / nonlinear approaches will all be degraded. In the first part of this thesis, we propose a new method improving the NB interference suppression capability. Using a filter bank, we first split the received DS signal into subbands. Then, we model the channel and interference in each band as autoregressive (AR) processes and a Kalman filter is used to estimate the interference signal. Finally, the interference suppressed subband signals are combined to form the fullband signal. In the second part of this thesis, we consider the wide - band interference, we suggest a new receiver structure. This distinct feature of the structure is that it employs a median fileter sand can suppress pulse - tone as well as impulse interferences. Simulations show that the proposed new algorithms can perform significantly better than existing ones.
|Appears in Collections:||Thesis|