PLL Performance for Signals in the Presence of Thermal noise, Phase noise, and Ionospheric Scintillation

Performance of several carrier tracking loops are evaluated by the wideband data that mimic the real ionospheric scintillation. To accommodate the deficiency of the receiver’s oscillator, six-state dynamic model driven by Gaussian-distributed random samples is used to simulate the phase noise inherited by the imperfect crystal unit. Ionospheric scintillation is associated with the deep power fades and drastic phase variation. The scintillation model proposed by MITRE is employed in this paper to replicate the plasma irregularity in the ionosphere on a digital computer. The evaluation targets to enable a robust carrier tracking by optimizing the configurations for the tracking strategy. Constant-bandwidth and variable bandwidth loops are theoretically examined based on the model, and the results are then validated by using the simulated wideband samples. Discriminators critically influence the carrier tracking capability, due to the severe bias caused by the sine characteristic or the nonlinear operation on the noise. Four discriminators, usually used in a GPS carrier tracking unit, are investigated from a statistical sense. The outcome can explain the worse performance of the ATAN loop, compared with the decision-directed loop, in presence of the deep power fades. Kalman filter based tracking approach, compared with 15 Hz constantbandwidth loops, lowers the lock threshold by 7.5 dB in the presence of strong scintillation, for both amplitude and phase; thus revealing the attractive application in occurrence of scintillation. The improvement benefits from (1) the soft treatment on the ambiguity of the bit sign, and (2) optimal estimation by Kalman filter in minimizing the mean square error. During weak scintillation or absence from scintillation, the tracking strategy on Kalman filter demonstrates a 4.5 dB improvement in aspect of tracking sensitivity and immunity to the cycle slip.