France – University of Franche-ComtÃ© PhD Scholarship
Ph.D. Training Opportunity & Stipend – starting fall 2008
Spectral analysis of phase noise and frequency stability of oscillators
proposed byEnrico Rubiola
FEMTO-ST Institute, Dept. of Time and Frequency
CNRS et University of Franche Comte
Time and frequency have a privileged role in physics and applications because they are the most precisely measured physical quantities. The wrist watch, for example, is the only artifact accurate within 1E-5 – 1E-6 at a cost affordable to all consumers. Atomic clocks exhibit the amazing accuracy of 1E-15, and a stability better by a factor 10. Though the accuracy of 1E-15 relates only to fundamental physics and metrology, short term stability is a major concern in telecommunications, space applications and radars.
Traditionally, frequency stability is measured in the time domain and described using wavelet variances known as the Allan variance sigma_y^2(tau) and its modified versions. The variable tau is the measurement time. The stability depends on the measurement time in the same way of a balance, which is less “precise” when the mass to be measured is very different from 1 kg.
The estimation of sigma_y^2(tau) and of its confidence is well documented in the literature available in the past 20 years. Conversely, the spectral analysis of oscillators is still an empirical domain, in which results are usually given as a raw power spectral density of the phase noise, i.e., Sphi(f) ou L (f). Analysis seldom goes beyond the identification of the interference from the mains power supply (50 or 60 Hz et harmonics) and the identification of two-three fundamental types of noise. This approach is no longer suitable to the needs of emerging technologies.
This is a new domain. Hence, the minimum target is reasonably low: to adapt established spectral-analysis methods to the oscillator and to identify the oscillator’s stochastic phenomena. Nonetheless, this is a great opportunity for a smart student, as he can innovate in the interpretation of the oscillator physics and in the measurement methods and achieve outstanding experimental results. A side branch is the improvement of the measurement accuracy, which is currently of 2 dB in primary laboratories. The bulk of the expected work is about the statistical analysis of phase noise spectra. Besides, the student will achieve a deep understanding of oscillators and of experimental techniques.
University and laboratory
Ph.D. scholarship is managed by the University of Franche Comt’e, Besançon, France. The work site is the Time and Frequency Dept. of the FEMTO-ST Institute, affiliated to the University of Franche Comt’e, Besançon. This department, merging the laboratory of chronometry (LCEP) and the laboratory of physic and metrology of oscillators (LPMO), issues from the Laboratoire de l’Horloge Atomique, funded by the Nobel prize Alfred Kastler. The T&F Dept. is the world leader in the domain of the measurement of the oscillator noise. Besides, this department is the european leader in the domain of ultra-stable oscillators and chip-scale atomic clocks.
The best student for this subject is fond to computer programming and calculus, and eager to learn sophisticated experimental methods of electronics. He is motivated by the will of learning and by the investment on his future life. In the longer term, he would like to do applied research in academic institutions or industry, or high-tech engineering, preferably in international environment.
During his Ph.D., the student will be advised by E. Rubiola (2/3-3/4), full professor, and by F. Vernotte (1/4-1/3), full professor and head of the Observatory of Besançon.
The student will receive a stipend from the French ministry of education for three years, as ruled by the French government. The candidate must fill two conditions
age 29 years max,
university degree “Master 2″ (5 year university degree) or equivalent, obtained in 2008. The equivalence is managed on site.
Candidates unaware of the French education system should know that: tuition are very small or zero, and the stipend is generally sufficient for a modest yet quite reasonable life standard.
No comments yet.