Quantum kinetic theory is founded upon the action of the conservation laws within systems that may be both strongly driven and subject to strong interparticle couplings. For any open mesoscopic conductor, conservation must act globally as well as microscopically. In maintaining global conservation, the explicit interplay of the mesoscopic device and its bounding leads is paramount. Within standard quantum kinetics, this device-lead interaction imposes very strong constraints on the possible behavior of the noise spectral density. That is so over the whole range of driving currents. We review a fully quantum kinetic theory of mesoscopic conduction and discuss the experimental consequences of its conserving constraints, with special reference to the experiment of Reznikov et al., Phys. Rev. Lett. 75, pp. 3340 - 3343, 1995.

This work brings into use the well known circuit theory concept that states that the short-circuit noise current is related to the open-circuit noise voltage through a small-signal admittance (or small-signal impedance). This concept is perfectly illustrated by the famous Nyquist theorem for nonlinear equilibrium systems. The original results that are described in this report are as follows: (1) the equilibrium fluctuation-dissipation theorem of the 3rd order that is consistent with the cumulant spectrum invariance concept holds only for the bispectrum with equal frequencies, i.e. for the one-frequency bispectrum. (2) the one-frequency bispectrum of the equilibrium electrical fluctuations is proportional to magnitude of the small-signal quadratic rectification at the equilibrium state.

Current noise measurements allow to enlighten the transport mechanisms in mesoscopic samples in a complementary way to conductance measurements. The noise gives directly access to the charge of the current carriers and is modified by interactions. This is particularly instructive in the context of hybrid superconductor-normal metal structures where charge pairs generated by Andreev reflection compete with quasiparticles in the current transport. We present investigations of current noise in various hybrid geometries at temperatures down to 50mK using a SQUID as fluctuations detector. The first two types of structures involve one superconductor in contact with a diffusive normal metal reservoir. To obtain a nearly perfect contact between a superconductor and the normal metal (high transmittive contact), we used niobium as a superconductor and copper as normal metal. Then, the current transport is mediated by charge pairs induced from the superconductor (proximity effect) and the noise is doubled S=2/3x2eI, compared to the case where the two electrodes are normal S=1/3x2eI (where I is the mean bias current and e the electron charge). The 1/3 reduction from the full Poisson noise (S=2eI), observed e.g. in a tunnel junction between two normal reservoirs, is due to the diffusive character of the normal metal. The noise doubling is observed as long as the voltage drop at the junction is smaller than the superconducting gap. Above the gap the shot noise has the same behavior as in the normal case. In the second type of structures under study, where a superconductor (here TiN) is in contact with a strongly disordered metal (heavily doped silicon), the contact is degraded because of the Schottky barrier that forms at the interface (low transmittive junctions). The noise in such junctions is also doubled, but is twice the full Poisson noise S=2x2eI since the noise is mainly generated at the interface barrier. This confirms that the enhancement of the conductance observed at low bias voltage is due to coherent backscattering of quasiparticles towards the interface by the strong disorder in the silicon (reflectionless tunneling). When this excess of subgap conductance vanishes at finite voltage, the noise slope crosses over to the Poisson value 2e indicating a large quasiparticle contribution to the current. The third type of structures investigated is a long diffusive S/N/S junction made of aluminum and copper. The noise is enhanced compared to the N/N/N-case due to the confinement of the electron gas between the two superconducting reservoirs and the current transport involve Incoherent Multiple Andreev Reflections. Inelastic processes are important in our samples because the lengths of the junctions (4, 10, and 60 μm) are of the same order of magnitude as the inelastic scattering length. We analyze the results quantitatively with recent semi-classical theory taking into account electron-electron interaction and heat transfer through the SN interfaces in the context of S/N/S junctions. For the longer junctions, we also considered electron-phonon-interaction as a possible cooling mechanism. Finally we show that the energy dependence of the re-entrance of the resistance, observed at low voltage, is essentially due to the increasing effective temperature of the quasiparticles in the normal metal.

The rich variety of noise properties that make the field of mesoscopic transport so fascinating is going to be shared with "common" VLSI devices. Typical MOSFETs used of present-day VLSI circuits and systems already have feature sizes smaller than what we usually consider mesoscopic devices. In this talk, we focus on shot noise of the drain and gate currents in nanoscale MOSFETs. The subject is of interest from the point of view of applications, since adequate models of noise in such MOSFETs are required, especially for high-frequency analog and mixed-signal applications, and from the point of view of the understanding of the underlying physics, since effects typical of mesoscopic devices can now be observed at room temperature and in silicon.

The bispectrum of the 1/f noise is investigated in the present work. For the Gaussian noise it equals zero. LEDs on self-organized InAs/GaAs quantum dots and laser diodes on In_0.2Ga_0.8As/GaAs/InGaP quantum wells made in Russia were tested. The voltage noise was analyzed in a wide interval of currents through the diodes. Estimates of the probability density function and semi-invariants of the noise have not revealed any appreciable deviations from the Gauss law. Noise spectra S_v(f)in the range 1 Hz - 20 kHz were analyzed. In most cases the frequency exponent γ_s of the spectrum is close to one, the Hooge’s parameter α_H has magnitude of the order 10^-4. The bispectrum B_v(f₁,f₂of the noise is a complex function of frequencies f₁ and f₂. Its absolute value is decreasing while moving from the beginning of the frequency plane Of₁f₂. The decrease along the bisector (f₁ = f₂ = f) follows the power law characterized by the frequency exponent γ_B ≈ 1.5 γ_s. The dependence of the "height" of |B_v(f,f)| on the current through the diode is qualitatively similar to this one for the spectrum. The power law describes these dependences, however the exponents are essentially different.

This paper describes the fundamental concepts that characterize the phase modulation (PM) noise and amplitude modulation (AM) noise of electronic devices in the frequency-domain, and their relationship to tradition time-domain measures of frequency stability. The statistical confidence of the data is discussed. Using the fundamental concepts, the affects of frequency multiplication, division, and mixing on PM noise are explored. Also covered is the relationship between noise figure and PM/AM noise in an amplifier. The affect of summing a large number of similar sources or amplifiers on the resulting PM noise is briefly mentioned. Common techniques used to measure PM noise in oscillators such as single channel two-oscillator, dual channel two-oscillator, three-cornered-hat with cross corrrelation, delay line discriminator, and carrier suppression are described. It is shown how these techniques can be extended to the measurement of PM noise added by other electronic devices such as amplifiers and frequency multipliers/dividers. Common errors and the strengths and weaknesses of the various measurement and calibration techniques are also described.

We review recent experimental and theoretical work towards the realization of a fast and ultrasensitive, nanomechanical displacement detector based on the radio frequency single electron transistor (rf-SET).

We have performed an experiment in order to measure the displacement of a mirror at the attometer level. The mirror is coated on a high-Q millimeter-sized mechanical resonator made of fused silica. Using a high-finesse optical cavity, a highly stabilized Ti:Sa laser source and a low-noise detection system allows us to reach a shot-noise limited sensitivity of 2.8 10^-19 m/√Hz at the resonator’s fundamental acoustic resonance frequency, about 2 MHz. We have also implemented a feedback scheme, where the information about the mirror's motion is used in a feedback loop to control the intensity of a radiation pressure force applied onto the mirror. This allows to reduce the thermal noise and to cool down the mirror well below its initial temperature. The effect of this cold damping mechanism is visualized both on the temporal evolution of the mirror displacement and on its distribution in phase-space, with a sensitivity of a few attometers. We have also observed the thermal noise squeezing in the case of a parametrically-modulated feedback force, observing both the 50% theoretical limit of squeezing below the mirror’s parametric oscillation threshold, and the oscillation above threshold. Enhancement of the experimental setup, with the use of an optical cavity with a higher finesse inserted in a liquid helium cryostat, in order to observe quantum effects of radiation pressure such as the Standard Quantum Limit, will also be discussed.

The stochastic component of chemical sensor signal contains valuable information that can be visualized not only by spectral analysis but also by using nonlinear characteristic components. The analysis of nonlinear stochastic components enables the extraction of physically interesting and useful features and may lead to significant improvements in selectivity and sensitivity. Various measures of nonlinearity are presented and estimated for sample sensor data obtained from commercial chemical sensors. Particular attention was paid to the bispectrum function that detects nonlinear and non-stationary components in the analyzed noise. The results suggest that bispectrum measurements provide valuable information about the nature of noise generation in chemical sensors. Moreover, we have found, by analyzing skewness and kurtosis distributions, that the measured time series were stationary.

Carbon nanotubes may constitute the ultimate conducting wires for nano-electronics, with their diameters as small as a few tens of atoms and their length of order one micrometer. Because of the particular band structure of graphite, nanotubes have at most two conducting channels, which makes them a one dimensional conductor with very exotic properties. Experimental investigations have indeed shown non conventional features, such as non-ohmic behavior, superconductivity and an ability to carry a huge current density. We have carried out shot noise measurements on nanotubes which are suspended between metallic electrodes. One consequence of the suspended geometry is a very low 1/f noise, thereby enabling the extraction of shot noise. In bundles of nanotubes, we find a reduction of shot noise by more than a factor 100 compared to the full noise 2.e.I expected for uncorrelated electrons. A low noise is also found in an isolated single wall nanotube. In a simple non-interacting-electron picture, such a low shot noise implies that the electrical conduction through a bundle of nanotubes is concentrated in a few ballistic tubes. Another interpretation however would be that a substantial fraction of the tubes conduct with a strong reduction of the effective charge (more than a factor 50) due to electron-electron interaction.

We survey a theoretical investigation of shot-noise in single and multiple barrier diodes. Several mechanisms responsible of the suppression of shot noise are reviewed together with the conditions for obtaining shot noise enhancement. The coherent versus sequential tunneling model for the double barrier resonant diode is discussed in the light of existing experiments.

A numerical approach for the evaluation of conductance and shot noise suppression in mesoscopic structures is presented and applied to a few relevant cases. Details are provided both of the technique based on a recursive Green's function procedure that is used for calculations in the absence of a magnetic field and of the recursive scattering matrix method that is applied to simulations with nonzero magnetic field. Shot noise suppression in cascaded chaotic cavities is studied and discussed in comparison with the suppression obtained for cascaded potential barriers. It is observed that the Fano factor for multiple cascaded cavities is the same as that for a single cavity, as long as its apertures are small compared to its width. Finally, a particular structure, consisting in a cavity with a central potential barrier, is studied and from its noise behavior conclusions are drawn about the very different role played by a constriction or by a potential barrier in the presence of edge states.

We show that shot noise is an informative tool for studies of hopping and resonant tunnelling between localized electron states. In hopping via several states, shot noise is seen to be always suppressed compared with its classical Poisson value S_I = 2eI (I is the average current). The degree of suppression depends on the geometry of current paths and the distribution of the barriers between the localized states. In resonant tunnelling through a single impurity an unusual enhancement of shot noise is observed. It has been established, both experimentally and theoretically, that a considerable increase of noise occurs in the case of two interacting resonant tunnelling channels.

A lower bound on the amount of energy needed to carry out an elementary logical operation on a qubit system, with a given accuracy and in a given time, has been recently postulated. This paper is an attempt to formalize this bound and explore the conditions under which it may be expected to hold. For a specific, important case (namely, when the control system is a quantized electromagnetic field) it is shown how one can extend this result to a generally stronger constraint on the minimum energy density required, per pulse.

In a recent study (Physics Letters A 305 (2002) 144-149) the forthcoming end of Moore's Law was predicted. The study was based on the assumption that the power dissipation was dominated by the energy dissipated during charging/discharging the CMOS gate capacitance during bit flips. In the present paper the fundamental lower limit of power dissipation during the operation of a CMOS gate is obtained. The results indicate that the energy efficiency of today’s microprocessors is extremely low. Thus, a significant improvement of the energy efficiency of microprocessors may be able to prolong the lifetime of Moore's Law. We compare the results with published data on the lower limit of power dissipation of quantum gates. Interestingly, "Classical" beats "Quantum," if we give the same chance to them. Finally, we evaluate the energy cost of Shannon-information transfer. This measure, which cannot be improved by error correcting algorithms, is the ultimate one, the real characteristic of performance versus power dissipation. In the most ideal case, the CMOS gate performs by at least an order of magnitude better than the quantum gate. It is shown that, at the same complexity of hardware, same speed and same temperature, quantum computing means more noise, less information channel capacity and greater power dissipation than the same measures in classical computers.

The tunneling rate of single electronic tunneling (SET) transistor at nonzero temperature has been analyzed. A stability condition, which allows one bit flip error/year, is used. The aspects of dissipation and performance are studied, for a single SET and for a microprocessor built of SETs, versus the size.

We report on studies aimed at understanding and improving the intrinsic noise of high-performance sensors using a 2D electron gas channel confined by a quantum well in the pseudomorphic AlGaAs/InGaAs/GaAs heterostructure. MIS gated and ungated Hall sensors shaped as a Greek cross with dimensions ranging from 100 μm down to submicrometer range have been investigated. At room temperature the predominant low frequency Hall voltage noise originates from the ensemble of trapping/detrapping events occurring within the continuum of GaAs surface states. Its power spectral density can be deduced from independent measurements of the interface trap density-of-states by applying Shockley-Read-Hall dynamics and the Fluctuation-Dissipation Theorem. In fact, theoretical spectra calculated without any adjustable fitting parameter coincide closely with the experimentally measured ones. At cryogenic temperature this interface traps noise freezes out, thus revealing a much weaker intrinsic background noise with 1/f spectrum. For small sensors the intrinsic 1/f noise converts into one or a few lorentzians due to the action of individual random telegraph signals (RTS). For Hall crosses with an intersection of 4x4μm², we find statistically less than 1 fluctuator per each decade of time constant at 77 K. Due to the random distribution of the elementary fluctuators, some of these small Hall crosses may show less low-frequency noise than much larger 60x60μm² sensors.

Hall sensors are used in a very wide range of applications. A very demanding one is electrical current measurement for metering purposes. In addition to high precision and stability, a sufficiently low noise level is required. Cost reduction through sensor integration with low-voltage/low-power electronics is also desirable. The purpose of this work is to investigate the possible use of SOI (Silicon On Insulator) technology for this integration. We have fabricated SOI Hall devices exploring a wide range of silicon layer thickness and doping level. We show that noise is influenced by the presence of LOCOS and p-n depletion zones near the edges of the active zones of the devices. A proper choice of SOI technological parameters and process flow leads to up to 18 dB reduction in Hall sensor noise level. This result can be extended to many categories of devices fabricated using SOI technology.

Noise in phototransistor due to packaging has been investigated via theoretical models and measurements. A new design of package for phototransistor to reduce noise is proposed and tested for commercial photoelectric sensors. Generally, there are two kinds of materials, namely, metal and plastic used for the package, and a metal case made of aluminum or copper is the more popular. However, a metal case could pick up noise from electromagnetic fields, such as radio frequency noise caused by micro controllers. The noise could go into a phototransistor because of the metal case connected to the collector of the phototransistor, and amplified further, and could turn on a photoelectric sensor. Surface current induced by electromagnetic waves on a metal case due to skin effect is analyzed using theory of electrodynamics. Comparison of phototransistor packaged using the new design in photoelectric sensor and one in commercial is carried out. The latter is locked on under radio frequency noise environment, but, the former is not.

The excess noise in commercial semiconductor gas sensors, during exposure to different gases at various temperatures, was studied. The measurements have been carried out using different concentrations of ethanol vapor, CO, NO_x, H₂ and SO₂ gases in synthetic air. The normalized noise spectra of the sensors stabilize after the first day of the burning-in process in synthetic air. The sensors do not show significant deviation from the ohmic behavior over the full range of applied voltage. The induced excess noise is proportional to the square of the applied voltage, so resistance fluctuations explain the spectra. Each gas induces a characteristic noise spectrum. The analysis of resistance fluctuations in gas sensors, during exposure to gases, has the potential of a strongly enhanced sensitivity and selectivity.

We present a model for detection of the states of a coupled quantum dots (qubit) by a quantum point contact. Most proposals for measurements of states of quantum systems are idealized. However in a real laboratory the measurements cannot be perfect due to practical devices and circuits. The models using ideal devices are not sufficient for describing the detection information of the states of the quantum systems. Our model therefore includes the extension to a non-ideal measurement device case using an equivalent circuit. We derive a quantum trajectory that describes the stochastic evolution of the state of the system of the qubit and the measuring device. We calculate the noise power spectrum of tunnelling events in an ideal and a non-ideal quantum point contact measurement respectively. We found that, for the strong coupling case it is difficult to obtain information of the quantum processes in the qubit by measurements using a non-ideal quantum point contact. The noise spectra can also be used to estimate the limits of applicability of the ideal model.

The impact of control field fluctuations on the optimal manipulation of quantum dynamics phenomena is investigated. The presence of significant field fluctuations is shown to break down the evolution into a sequence of partially coherent robust steps. Robustness occurs because the optimization process reduces sensitivity to noise-driven quantum system fluctuations. This process takes advantage of the observable expectation value being bilinear in the evolution operator and its adjoint. The consequences of this inherent robustness bodes well for the future success of closed loop quantum optimal control experiments.

We present experimental work that investigates whether quantum information carried by light can be stored via reversible mapping of the quantum state of such light onto a collective atomic coherence. Such a quantum memory could be utilized to allow quantum communication over long, lossy channels. Current efforts concentrate on writing a photon-number-squeezed state of light onto a collective coherence between the ground-state hyperfine levels of ⁸⁷Rb atoms in a warm vapor cell, and subsequent on-demand retrieval of this light. In this approach, intensity squeezing between the written and retrieved photon fields provides evidence for storage of a photon-number-squeezed state of light. The scheme is based on spontaneous Raman transitions that create the atomic coherence, and at the same time convert control fields into signal fields that propagate under conditions of electromagnetically induced transparency. We present experimental results demonstrating the storage and retrieval of light using this method, and discuss techniques for measuring intensity squeezing between these photon fields.

A model of parallel optical computer for Fourier transform tasks is built. The computer operates in both quantum and classical modes. The quantum computer (QC) works with entangled photons and the input gates are n-qubit sensitive, while the classical computer (CC) operates without entanglement. The information capacity (I) for both CC and QC is given as a function of number of qubits and N_p (number of photons per input pixel). It is shown that the information capacity significantly increases for the quantum computer compared to the classical one as the number of qubits is increased, assuming that N_p is constant for both the QC and CC, i.e. same amount of energy input is assumed for the computers. However, it is also pointed out that the complexity of the QC significantly increases, too. To quantify this, we introduce a new physical quantity called physical complexity (noted as Q). We define the physical complexity (Q) of a computer as ∑_i,klog₂n^k_ai where k runs over all the gates/elements in the computer; n_ai is the number of distinguishable states what a gate can set for a_i. For QC a_i is defined as the i^th coefficient of the wave function. For CC it is the i^th independent component of the Fourier transform of the energy flow what a particular gate is able to control. By using this definition we show that, for both the quantum and classical computers built in this work, I ≤ Q, in other words the information capacity is less or equal than the physical complexity. Intuitively we suggest that this is a general law that is valid for every computer irrespectively of type, quantum or classic, and architecture.

Decoherence is the main obstacle to the realization of quantum computers. Until recently it was thought that quantum error correcting codes are the only complete solution to the decoherence problem. Here we present an alternative that is based on a combination of a decoherence-free subspace encoding and the application of strong and fast pulses: "encoded recoupling and decoupling" (ERD). This alternative has the advantage of lower encoding overhead (as few as two physical qubits per logical qubit suffice), and direct application to a number of promising proposals for the experimental realization of quantum computers.

The concept of quantum computing has arisen as a methodology by which very rapid computations can be achieved. There also has been considerable discussion about physical implementations of the qubit. This has led, in recent years, to a situation in which quantum computing and quantum information theory are being rapidly developed. In general, the specific advantages offered by quantum computing have been somewhat nebulous. On the one hand, faster computing was promised, but we now know that no speedup of most algorithms exists relative the speed that can be obtained with massive parallel processing. Then, we are promised that the use of entanglement will make quantum computing possible with a much smaller use of resources. Yet, entanglement must be viewed as a hidden variable, which is not accessible in experiment. How does this provide the speedup? We have suggested that analog processing may provide a suitable alternative, and may be the basis which provides the speedup in quantum computing, but this is a controversial assertion. In this talk, we will discuss these particular viewpoints, along with several approaches to a wave basis for (quantum) computing.

A quantum computer is highly promising and widely studied since it enables a much higher calculation speed than current computers. Due to physical limitations, however, the current quantum computer can only solve much smaller scale problems than conventional computers. In order to process a large-scale problem at high speed, we have been studying a quantum computing emulator utilizing semiconductor memories. As a result, we have realized a dedicated processor that solves search problems, such as the satisfiability problem, at much higher speed than current computers. Consequently, without using "quantum process," the possibility of a quantum computer executing a large-scale problem is shown.

We show that, if time comes when quantum algorithms can be used for universal parallel computation, the "quantum-parallel" computer hardware will most probably be a classical physical system corresponding to a Hilbert space and the actual realization may be the combination of analog and digital circuits. We first point out the practical difficulties of universal quantum computing which may prohibit practical applications as universal computers. Then we show how to apply analog microelectronic circuits to realize the architecture, data processing and parallel computing abilities of quantum computing via Hilbert space computing with analog circuits. Such a Hilbert-space-analog (HSA) computer simulates the Hilbert space and its operators, and it is able to use and test quantum algorithms developed for the real quantum computers. Such a computer would be free of most of the practical difficulties of realizing and running a real quantum computer. This computer can be made universal. It is remarkable that by using the same numbers of transistors as in today's PCs, such a HSA computer can manipulate ~10⁷ analog numbers corresponding to ~22 qubits, simultaneously, by quantum-parallel processing.

The technique of projective measurements in linear optics can provide apparent, efficient nonlinear interaction between photons, which is technically problematic otherwise. We present an application of such a technique to prepare large photon-number path entanglement. Large photon-number path entanglement is an important resource for Heisenberg-limited optical interferometry, where the sensitivity of phase measurements can be improved beyond the usual shot-noise limit. A similar technique can also be applied to signal the presence of a single photon without destroying it. We further show how to build a quantum repeater for long-distance quantum communication.

A plethora of qubit devices is under consideration for possible development ranging from quantum telereporters to quantum computers. The primary obstacle to the success of these efforts is the phenomenon of quantum decoherence, the rapid vanishing of the off-diagonal components of the reduced density matrix representing the computational degrees of freedom of the device. Following a review of the physics of quantum decoherence, several instructive examples are exposited. Decoherence issues associated with the various possible approaches to quantum computing are addressed. Also, possible generic methods are reviewed for surmounting the decoherence obstacle.

We introduce a new approximation scheme for evaluation of onset of decoherence at low temperatures in quantum systems interacting with environment. The approximation is argued to apply at short and intermediate times. It provides an approach complementary to Markovian approximations and appropriate for evaluation of quantum computing schemes.

Electrical engineers and physicists are naturally very interested in noise in circuits, amplifiers and detectors. With the advent of quantum computation and other high frequency electronics operating at low temperatures, we have entered a regime where quantum noise and quantum-limited detectors are important. Here we describe the general concept of a two-level system as a quantum spectrum analyzer and apply it to a simple superconducting qubit, the Cooper-pair box. We then discuss the coupling of a Cooper-pair box to its electromagnetic environment, whose noise leads to a finite polarization and excited-state lifetime of the qubit. Finally, we describe a theoretical technique for treating a qubit coupled to a measurement system, which allows one to calculate the full quantum noise of the measurement device. We present results for such a calculation for the case of a normal SET.

The quantum evolution of an individual solid-state qubit during the process of its continuous measurement can be described by the recently developed Bayesian formalism. In contrast to the conventional ensemble-averaged formalism, it takes into account the measurement record (in a way similar to the standard Bayesian analysis) and therefore is able to consider individual realizations of the measurement process. The formalism provides testable experimental predictions and can be used for the analysis of a quantum feedback control of solid-state qubits. The Bayesian formalism can be also applied to the continuous measurement of entangled qubits; in particular, it shows how to create a fully entangled pair of qubits without their direct interaction, just by measuring them with an equally coupled detector.

The influence of the built-in electric field (e.g. several potential barriers, impurity gradients, etc) at presence of external crossed electric and magnetic fields on the level of low-frequency noise is theoretically studied by the use of interrelated Langevin type Boltzmann transport equations for the systems of electrons and phonons for non-degenerate n-type semiconductors. At the first time it is shown that within the context of the proposed problem the built-in field causes origin of separate 1/f-noise component, which, besides the main parameters of semiconductor, depends also on the value of this field. The spectral density of this component by the form is similar to the Hooge's empirical formula. For the parameter analogous to the Hooge parameter α the comparison with experimental data for n-type Si is carried out. It is shown that for range of values of the built-in electric field from 50 to 600 V/m at low temperatures from 77K to 150K, the longitudinal component of the analog of the Hooge parameter varies from 10^-6 up to 10^-4, which in some cases may exceed corresponding values of the generally observed 1/f-noise. The transversal component of the Hooge parameter analog has square dependence on external magnetic field intensity; even for low temperature region, at values of magnetic field from 50 to 400 A/m it has very low values and varies from 10^-15 up to 10^-11. Basing on the calculations a physical model describing the origin of built-in component of 1/f-noise is proposed. It explains the mechanism of origin of the additional component and helps to shed light on the origin of the general 1/f-noise.

The two main causes of origin of the mobility fluctuation of the electrons in homogeneous, unlimited, and non-degenerated semiconductors are discussed. It is shown that the mobility fluctuation is conditioned by the symmetric component of the fluctuation of the distribution function, i.e. by the fluctuations of the conduction electrons energy. On the base of the developed quasi-classical model the spectrum of electrons lattice mobility fluctuations is calculated. In the frequency wide variation range it has 1/f form.

The current internal fluctuations appearing in the homogeneous, unlimited and non-degenerate semiconductors having parabolic band structure are investigated. At the considered case the external electric field is absent and the semiconductor is in thermal equilibrium state. For the definiteness only the behavior of the conduction electron system is considered. It is shown that equilibrium fluctuations of the electron current, conditioned by the fluctuations of the electron quasi-momentum, are describing by fluctuations of the asymmetric component of the electron distribution function. The following mechanism of these fluctuations is suggested. During the random phonon-phonon scattering the fluctuations of the quasi-momentum of acoustic phonons are coming into existence, which are transmitting to the electron system via electron-phonon interactions. On the base of this mechanism, the spectral density of the electron current equilibrium fluctuations S_I(ω) is calculated. It is shown that S_I ≈ const in the frequency range ω < ω₀. In the frequency range ω>₀ < ω < ω₁, frequency dependence of spectrum S_I(ω) described by 1/ω law, and in the range ω > ω₁ it is described by 1/ω² law. The characteristic frequencies ω₁ and ω₀ are determined by parameters of semiconductors being under investigation as well as by processes of scattering and diffusion of electrons and phonons in quasi-momentum space.

Both selectivity and sensitivity of chemical sensors can be significantly improved by exploiting the information contained in microfluctuations present in the sensor system. We call our collection of methods to extract information from these microfluctuations Fluctuation-Enhanced Chemical Sensing. In this review paper we summarize our recent experimental and theoretical results using commercial Taguchi sensors, Surface Acoustic Wave (SAW) Devices and MOS-FET based sensors.

Finite pulse width effects in level crossing detectors and similar systems, such as neurons, cause an output noise which is a monotonically increasing function of frequency in the low frequency limit. The effect is also relevant for shot noise phenomena with reduced strength.

Electrochemical corrosion processes can be investigated by observation of the charge flow between electrolyte and the corroding metal. Usually, the charge flow is observed as spontaneous current and voltage fluctuations (electrochemical noise) in a three-electrode setup. Different types of corrosion processes can be recognized by electrochemical noise analysis. Uniform corrosion rate can be evaluated by estimation of polarization resistance between metal and electrolyte. Local corrosion events (breakdowns of the passive layer) that produce characteristic transients observed in noise can be detected. The different methods of electrochemical noise analysis are presented. The limitations and advantages of the method for corrosion monitoring and research are underlined. The experimental results are also discussed.