B. Earl Wells

Three‐dimensional structure of electron holes driven by an electron beam

Using 3-D particle-in-cell (PIC) simulations we studied the structure and temporal behavior of electron holes (e-holes) in a magnetized plasma driven by an electron beam. When e-holes are fully evolved from high-frequency waves in a time of about a few tens of electron plasma periods, most of the wave energy in the plasma resides in them. Parallel to the ambient magnetic field Bo, the potential distribution of an e-hole is approximately a Gaussian, and the scalelength lz is only a few Debye lengths when determined by the effective temperature of the beam-modified electron distribution function. Transverse to Bo, the potential distribution tends to have a flat top, which makes it difficult to fit a Gaussian distribution, but the scalelengths at which the potential decays in the transverse directions (lx and lz) are found to be only slightly longer than lz. The passages of electron holes monitored at several points in the simulation volume has the signature of bipolar parallel electric field and unipolar perpendicular electric-field pulses as measured from FAST and POLAR. The eventual decay of e-holes is accompanied by the generation of lower hybrid (lh) waves.

Creating an adaptive embedded system by applying multi-agent techniques to reconfigurable hardware

This paper describes how the multi-agent paradigm that is so prevalent in todays distributed reactive software systems can be extended to operate in a dynamically adaptable embedded system environment. In this model, both reconfigurable hardware and the embedded system software design space can be partitioned into autonomous units of execution that are called agents, where each agent has the capacity to interact with the environment and other agents in an intelligent manner. The incorporation of agents that reside in reconfigurable hardware makes it possible to create an intelligent system that exploits the speed advantages associated with reconfigurable hardware within the framework of a unified system model that spans the entire hardware/software continuum. In addition to dynamic adaptability, this approach has the potential to greatly facilitate the flexibility, efficiency, fault tolerance, scalability, and reliability of embedded systems. This paper highlights agent-based reconfigurable embedded systems, and illustrates how they can he applied to a real-time sensor fusion type application using a conventional hardware description language.

Task Scheduling in a Finite-Resource, Reconfigurable Hardware/Software Codesign Environment

This paper describes how static task-scheduling methods using stochastic search techniques can be applied to digital systems that are composed entirely of reconfigurable hardware. Such reconfigurable hardware is characterized by the property that its low-level logical functionality is not determined at the time of manufacture, but rather that this functionality becomes set only shortly before or during the invocation of the targeted application. The purpose of this paper is to introduce the basic framework for applying static scheduling theory to arbitrarily-structured task systems targeted to be implemented in prefabricated resource-constrained reconfigurable hardware. The paper also describes initial solutions to this heterogeneous task-scheduling problem that combine a distributed list-based scheduling technique with three probabilistic search strategies. The techniques introduced produce detailed task scheduling and high-level hardware configuration information that can serve as inputs to the software and reconfigurable hardware design automation tools used in the hardware/software codesign process.

Doppler effect on target tracking in wireless sensor networks

This paper presents a new detection algorithm and high speed/accuracy tracker for tracking ground targets in acoustic wireless sensor networks (WSNs). Our detection algorithm naturally accounts for the Doppler effect which is an important consideration for tracking higher-speed targets. This algorithm employs Kalman filtering (KF) with the weighted sensor position centroid being used as the target position measurement. The weighted centroid makes the tracker to be independent of the detection model and changes the tracker to be near optimal, at least within the detection parameters used in this study. Our approach contrasts with previous approaches that employ more sophisticated tracking algorithms with higher computational complexity and use a power law detection model. The power law detection model is valid only for low speed targets and is susceptible to mismatch with detection by the sensors in the field. Our tracking model also enables us to uniquely study various environmental effects on track accuracy, such as the Doppler effect, signal collision, signal delay, and different sampling time. Our WSN tracking model is shown to be highly accurate for a moving target in both linear and accelerated motions. The computing speed is estimated to be 50-100 times faster than the previous more sophisticated methods and track accuracy compares very favorably.

Three‐dimensional kinetic simulation of the nonlinear evolution of lower hybrid pump waves

Nonlinear evolution of lower hybrid (LH) waves is studied by means of a fully three-dimensional parallel particle-in-cell (PPIC) code. The plasma is driven by a monochromatic LH pump wave, which drives secondary LH and low-frequency waves having a broad frequency spectrum from , to ω∼ωo >ωlh, where Ωi, ωo and ωlh are the ion cyclotron, pump and LH resonance frequencies, respectively. The temporal variations in the electric field components show both amplitude and phase modulations. In a plasma with equal electron and ion temperatures the dominant amplitude modulation occurs at the ion cyclotron timescale τci. The pondermotive force associated with the vector nonlinearity arising from theE ×B drift of electrons is seen to generate both density depletions and enhancements depending on the time-varying phase difference between the orthogonal electric field componentsEx andEy transverse to the ambient magnetic fieldBo in thez direction. Despite the use of quite strong pump wave amplitude, wave collapse in density cavities alone is not seen; instead, equally strong density cavities (cavitons) and enhancements (pilons) occur quasi-periodically both in time and space. The phase difference betweenEx andEy and its evolution yield a rotating transverse electric field vector with hodograms ofEx andEy changing with time. The temporal evolution of the parallel acceleration of electrons and transverse heating of ions are discussed. For relatively slow pumps the electron acceleration is predominantly unidirectional parallel to the pump phase velocityVp‖o. On the other hand, for sufficiently large pump phase velocities the acceleration becomes bidirectional. The parallel electron acceleration up toV‖max ∼ 2Vp‖o is common, and the transverse ion acceleration occurs up toV⊥max ≅(m/M)1/2Vp‖o ≅Vp⊥o, wherem andM are the electron and ion mass, respectively. The relevance of the above results to the observations on LH waves and their role in electron and ion accelerations is discussed.

Three‐dimensional numerical simulation of ion and electron accelerations by parametric decay of fast lower hybrid waves

The transverse acceleration of ions by fast lower hybrid waves in the topside auroral ionosphere has remained elusive. The linear and nonlinear propagation and evolution of lower hybrid waves are essentially a three-dimensional problem. Analytical treatment of this problem, including the essential kinetic effects for particle acceleration, is a formidable task if not impossible. We have treated this problem by a fully three-dimensional particle-in-cell simulation. It is found that a fast lower hybrid pump wave undergoes a parametric decay generating secondary waves over a broad frequency range from the ion cyclotron frequency Ω i to just above the lower hybrid frequency ω lh . These secondary waves are instrumental in the accelerations of both cold ions and electrons perpendicular and parallel to the ambient magnetic field B, respectively.

A Multicomputer Software Interface for Parallel Dynamic System Simulation

This paper describes the creation of an intelligent software tool that performs automatic parallelization of dynamic system simulations for excecution on arbitrary message-passing multi- computing configurations. The tool greatly decreases the amount of time needed to obtain a viable multiple processor implementation of large-scale simulations and can be expanded to include the latest methodologies that exploit functional and data parallelism present within the simulation model and algorithm. Such a unifying tool is designed to increase the likelihood that a parallel realization of a simulation problem can be obtained which has an acceptable level of performance. To illustrate the feasibility of developing such a tool, this paper describes the prototype implementation of the Automated Partitioning and Mapping Engine (APME) and demonstrates its effectiveness when applied to a large-scale simulation executed on a number of multicomputer systems and topological configurations.

Two parallel implementations for one dimension FFT on symmetric multiprocessors

In this paper, an empirical comparison is made between two parallel implementations of a one-dimensional Fast Fourier transform (FFT) that is targeted for a symmetric multiprocessor (SMP). The paper compares the run time characteristics and overhead (time complexity) associated with the two algorithms with that of previous research. The scalability of the two algorithms is also accessed using the isoefficiency function and the effect of caches on performance is presented. The isoefficiency function is defined as the rate at which the data should be increased with the number of processors to maintain constant efficiency. The two implementations are based on a tree and transpose, respectively. In the tree algorithm, the speedup does not increase linearly with the number of processors, but rather super linear speedup can be achieved for the two processor case. The transpose algorithm obtained (approximately) linearly speedup with respect to the number of processors with only moderate increase in the data size. Additional performance can be obtained by overlapping computation with communication and by efficient use of caches.

Noise Mitigation for Multiple Target Tracking in Acoustic Wireless Sensor Networks

The wireless acoustic sensors are not only subject to Doppler effect when being used to tracking high speed ground vehicles but also are prone to be disturbed by environmental noises generated locally by, for example, small animal passing or foliage blowing. These noises cannot be expressed using common analytic forms such as white or Gaussian noises as they impact portions of the sensor field sporadically. These noises interfere target sound detection and so disrupt multiple target tracking by the sensors, especially when those sensors have no capability to discern the sound sources and to communicate between sensor nodes due to their limited power. In this work, we propose an environmental noise model that is based on the noise data gathered from the fields and develop a heuristic noise mitigation algorithm for tracking multiple targets by the acoustic sensors with the limited capability. After filtering the noises with the noise mitigation algorithm, we track the multiple targets by using the rule based multiple target tracking algorithm developed in our previous study. Our results show that the tracker with the noise mitigation algorithm tracks multiple targets reliably with low computational complexity in the noisy acoustic wireless sensor network environments as long as the target sensing ranges are larger than the sensor separation.

Optoelectronic design of the simultaneous optical multiprocessor exchange bus (SOME-Bus)

Low latency, high bandwidth interconnecting networks that directly link arbitrary pairs of processing elements without contention are very desirable for parallel computers. The simultaneous optical multiprocessor exchange bus (SOME-Bus) based on a fiber optic interconnect is such a network. The SOME-Bus provides a dedicated channel for each processor for data output and thus eliminates global arbitration. Each processor can receive data simultaneously from all other processors in the system using an array of receivers. The architecture allow for simultaneous multicast and broadcast messages using several processors with zero setup time and no global scheduling. In this paper, we discuss the design of a possible opto-electronic implementation of the SOME-Bus along with an optical power budget analysis. Slant Bragg fiber grains arranged to couple light out of a fiber ribbon cable into an array of amorphous silicon detectors vertically integrated on a silicon are presented as a low cost novel means of interconnecting 10 to 120 processors.