Xiaojun Lu

On the Probabilistic Characterization of Nano-Based Circuits

The paper presents a novel probabilistic logical model to describe the nanodevice states. It describes the probability distribution of outputs. The model is based on observations on statistical physics and Markov random field. Different from previous model [Bahar (2004), Nano, Quantum and Molecular Computing: Implications to High Level Design and Validation, S. Shukla and R. I. Bahar, Eds. Norwell, MA: Kluwer], it uses probability density function to describe the probability behavior of the nanoscale circuits, which is more reasonable and flexible.

Probabilistic diagnosis of clustered faults for shared structures

The probabilistic diagnosis model is useful in many fields such as distributed network, digital system level testing and wafer fault testing. Some topologies and continuous defect units distributions are studied in our previous work. In this paper, we extend the model to arbitrary topology structure with share nodes and to the discrete defect distributions, such as Poission distribution and Binomial distribution. The results show high identification percentage of the nodes.

A Probabilistic Logic for Nanoscale Devices

A logic with probabilistic characterization is suitable for expressing the states of nanoscale devices. This paper describes a novel method of calculating probability distribution of nano gate states. It is based on the Markov random field theory with new features, such as clique potential, probability density of initial nodes. We demonstrate the effectiveness of the method by basic gates and circuits. The analysis shows that the device probability distribution highly depends on the system structures and temperature parameters.

A new device for the study of electron-ion interactions

Abstract The conceptual design of a new electron beam ion trap primarily intended for the study of electron–ion interactions is outlined along with some preliminary predictions regarding its capabilities.

Magnetic and electrode structure design for a new EBIT

Electron beam trajectory simulations have been performed to design a new electron beam ion trap. The design of the magnet and electrode structures was optimized based on the results of the simulations.

A new device for the study of electron-ion interactions: The Belfast EBIT

An electron beam ion trap (EBIT) has been designed and is currently under construction for use in atomic physics experiments at the Queens University, Belfast. In contrast to traditional EBITs where pairs of superconducting magnets are used, a pair of permanent magnets will be used to compress the electron beam. The permanent magnets have been designed in conjunction with bespoke vacuum ports to give unprecedented access for photon detection. Furthermore, the bespoke vacuum ports facillitate a versatile, reconfigurable trap structure able to accommodate various in-situ detectors and in-line charged particle analysers. Although the machine will have somewhat lower specifications than many existing EBITs in terms of beam current density, it is hoped that the unique features will facilitate a number of hitherto impossible studies involving interactions between electrons and highly charged ions. In this article the new machines design is outlined along with some suggestions of the type of process to be studied once the construction is completed.

Numerical simulation of the charge balance in an EBIT

Abstract The electron beam ions traps (EBITs) are widely used to study highly charged ions (HCIs). In an EBIT, a high energy electron beam collides with atoms and ions to generate HCIs in the trap region. It is important to study the physics in the trap. The atomic processes, such as electron impact ionisation (EI), radiative recombination (RR), dielectronic recombination (DR) and charge exchange (CX), occur in the trap and numerical simulation can give some parameters for design, predict the composition and describe charge state evolution in an EBIT [Phys. Rev. A 43 (1991) 4861]. We are presently developing a new code, which additionally includes a description of the overlaps between the ion clouds of the various charge-states. It has been written so that it can simulate experiments where various machine parameters (e.g. beam energy and current) can vary throughout the simulation and will be able to use cross-sections either based on scaling laws or derived from atomic structure calculations. An object-oriented method is used in developing the new software, which is an efficient way to organize and write code.

Probabilistic modelling and simulation of nanoscale circuits

In the emerging nano-scale technology, models for simulation play a key role for circuit design. This article presents a probabilistic logic model for nano-scale circuits. On the basis of probability distributions, the failure probability can be characterised. Models for primary gates are derived and extensive numerical experiments are performed based on model simulation.

The probability logics for nanoscale inverters cascade

Device failure is an important consideration in nano-scale design. This paper presents a probabilistic logic model to compute the probability distribution of the nano gate states. The characterization is based on Markov random field and statistical physics. The basic logic gates are probabilistically characterized. The effectiveness of the method is demonstrated by an inverter and the inverter cascade. Our analysis shows that the device probability distribution highly depends on the system structures and other performance parameters.

Probabilistic modeling of nanoscale adder

This paper presents the probabilistic logic model to compute the probability distribution of the nano gate states. The characterization is based on the Markov random field and statistic physics. The primary logic gates are probabilistically characterized. The effectiveness of the method is demonstrated by a full adder and an 8-bit adder. The analysis shows that the device probability distribution highly depends on the system structures and other performance parameters.