Quasi-optimum digital phase-locked loops (DPLL) are derived utilizing nonlinear estimation theory. Nonlinear approximations are employed to yield realizable loop structures. Baseband equivalent loop gains are derived, which, under high signal-to-noise ratio conditions may be calculated off line. Additional simplifications are made that permit the application of the Kalman filter algorithms to determine the minimum mean-square error (MSE) loop filter. Consideration is given to sampling rate and implementation requirements. Performance is evaluated by a theoretical analysis and by simulation. Theoretical and simulated results are discussed and a comparison to analog results is made.
[1]
A. Jazwinski.
Stochastic Processes and Filtering Theory
,
1970
.
[2]
Someshwara C. Gupta,et al.
Discrete-time demodulation of continuous-time signals
,
1972,
IEEE Trans. Inf. Theory.
[3]
Someshwar C. Gupta,et al.
Fundamentals of automatic control
,
1970
.
[4]
Donald L. Snyder,et al.
The state-variable approach to continuous estimation
,
1966
.
[5]
S. Gupta,et al.
The Digital Phase-Locked Loop as a Near-Optimum FM Demodulator
,
1972,
IEEE Trans. Commun..
[6]
D. N. Wormley,et al.
Fundamentals of Automatic Controls
,
1971
.