Le détecteur à jonction Josephson d’un signal sinusoïdal masqué par un bruit

Abstract

Signals detection is as far as other applications of Josephson junctions. It is shown that for the junctions to be useful for practical purposes, the arrays of junctions are efficient. This situation motivates in this thesis to model the Josephson junctions sinusoidal signals detector. The device is made up with the series array of Josephson junctions coupled through the RLC circuit. The array is defined in two different ways : identical array for which all the junctions have the same parameters and the disordered array in which some parameters vary from one junction to another. We have set two models of signal detection associated to the junctions of the array namely, the model of the non constant bias that is linearly ramping bias current at the speed v in one hand and in the other hand, the model of colored noise characterized by the correlation time. For identical series array, the deterministic dynamics reveals that the device exhibits birhythmic properties characterized by the coexistence of two stable attractors, each displaying a diffrent frequency of oscillation. In the presence of noise, the attractors are deformed and still well separated. After a certain time, one observes a transition from one attractor to another, leading to the problem of stability which is studied through the use of the effective energy. The deterministic dynamics also reveals that the disordered array missed the birhythmic properties for large values of disorder parameter. For low values of disorder parameter, the deterministic dynamics is similar to the one of identical array. The effect of noise is also similar but the switch from one attractor to another is less neat and therefore it is difficult to locate the separatrix, hinting to a more complex effective energy. The detection properties are studied though the analysis of the escape times sample collected by the use of two strategies namely, the coherent strategy and the incoherent strategy. The bias ramping speed does not influence the performance of the detector at low speed but improves the performance of the detector when the speed becomes high. For the case of the detector with colored noise, it is shown that the distribution of the escape times, the probability density function depends up on the noise correlation time. A careful analysis reveals that two effects are preeminent : the increase of stability ( as measured by the average escape time from the metastable state) and the increase of the extreme events in the tail of the escape time distribution.