Cholesky decomposition is a widely used algorithm to solve linear equations with symmetric and positive definite coefficient matrix. With large matrices, this often will be performed on high performance supercomputers with a large number of processors. Assuming a constant failure rate per processor, the probability of a failure occurring during the execution increases linearly with additional processors. Fault tolerant methods attempt to reduce the expected execution time by allowing recovery from failure. This paper presents an analysis and implementation of a fault tolerant Cholesky factorization algorithm that does not require checkpointing for recovery from fail-stop failures. Rather, this algorithm uses redundant data added in an additional set of processes. This differs from previous works with algorithmic methods as it addresses fail-stop failures rather than fail-continue cases. The proposed fault tolerance scheme is incorporated into ScaLAPACK and validated on the supercomputer Kraken. Experimental results demonstrate that this method has decreasing overhead in relation to overall runtime as the matrix size increases, and thus shows promise to reduce the expected runtime for Cholesky factorizations on very large matrices.

Extreme scale systems available before the end of this decade are expected to have 100 million to 1 billion CPU cores. The probability that a failure occurs during an application execution is expected to be much higher than today's systems. Counteracting this higher failure rate may require a combination of disk-based checkpointing, diskless checkpointing, and algorithmic fault tolerance. Diskless checkpointing is an efficient technique to tolerate a small number of process failures in large parallel and distributed systems. In the literature, a simultaneous failure of no more than N processes is often tolerated by using a one-level Reed-Solomon checkpointing scheme for N simultaneous process failures, whose overhead often increases quickly as N increases. We introduce an N-level diskless checkpointing scheme that reduces the overhead for tolerating a simultaneous failure of up to N processes. Each level is a diskless checkpointing scheme for a simultaneous failure of i processes, where i = 1, 2,..., N. Simulation results indicate the proposed N-level diskless checkpointing scheme achieves lower fault tolerance overhead than the one-level Reed-Solomon checkpointing scheme for N simultaneous processor failures.

Distributed transmit beamforming is a wireless communication technique in which multiple independent antennas transmit a common message while controlling their transmitted carrier phases to constructively interfere the message at a destination. Depending on the transmission goals and choice of network parameters, certain noteworthy benefits arise from beamforming. These benefits may include improved received signal to noise ratios, increased transmission ranges, power consumption reductions, or physical layer security against eavesdropping. We present a distributed cross-layer cooperative beamforming protocol implemented into the IEEE 802.11 wireless communication stack: Distributed Decode and Forward Beamforming. In this paper we conduct a brief literature review of current beamforming research, and present a detailed discussion on the mathematical foundation of the single- and multi-destination decode and forward beamforming technique. We then describe our distributed multi-destination decode and forward beamforming protocol, and present results from our in-depth evaluation of our protocol in a realistic simulation environment. In our evaluation, we investigate five aspects of our distributed beamforming protocol: beamforming event frequency, queue length, latency, participation level, and throughput.

Covariance Matrix Adaptation - Evolutionary Strategy (CMA-ES) is a black-box optimization method useful for applications where no direct inversion is possible. We present the development of a parallel CMA-ES algorithm that reduces the runtime for a specific geophysical data analysis, dipole localization. We compare our parallel algorithm against several other parallel CMA-ES variants on a sample dataset for dipole localization. We improve the performance of CMA-ES for the problem of finding dipoles in a subsurface environment as part of a closed-loop near-real-time wireless bioremediation system, REACTS (near-REal-time Autonomous bioremediation of ConTamination in the Subsurface). The goal of the performance improvement is to enable near-real-time analysis of geophysical data. For this application, our algorithm shows significant performance improvement over the other variants.

Prior research indicates that today's students, especially women, are attracted to careers in which they recognize the direct benefit of the field for serving societal needs. Traditional college level computer science courses rarely illustrate the potential benefits of computer science to the broader community. This paper describes a curricula development effort designed to embed humanitarian projects into undergraduate computer science courses. The impact of this program was measured through student self-report instruments. Through this investigation, it was found that students generally preferred projects that they perceived as "fun" over the projects that were social in nature. This may, in part, be due to the fact that it was difficult to reduce socially relevant problems to a level that beginning students could easily comprehend. This made it difficult to capitalize on the appeal of socially relevant problems in the early computer science courses.

Modeling and analysis of large scale scientific systems often use linear least squares regression, frequently employing Cholesky factorization to solve the resulting set of linear equations. With large matrices, this often will be performed in high performance clusters containing many processors. Assuming a constant failure rate per processor, the probability of a failure occurring during the execution increases linearly with additional processors. Fault tolerant methods attempt to reduce the expected execution time by allowing recovery from failure. This paper presents an analysis and implementation of a fault tolerant Cholesky factorization algorithm that does not require checkpointing for recovery from fail-stop failures. Rather, this algorithm uses redundant data added in an additional set of processors. This differs from previous works with algorithmic methods as it addresses fail-stop failures rather than fail-continue cases. The implementation and experimentation using ScaLAPACK demonstrates that this method has decreasing overhead in relation to overall runtime as the matrix size increases, and thus shows promise to reduce the expected runtime for Cholesky factorizations on very large matrices.

Participation in elementary school classroom settings is dependent on students' abilities to process questions quickly. Pedagogical studies theorize that some students are able to answer a question before other students are able to process the question, creating a division in classroom participation. This paper presents LEVEL (Listening to Everyone's Voice Enhances Learning), a wireless sensor network application to balance, diversify, and encourage in-class participation. LEVEL offers two operating modes: a centralized interface and a distributed adaptive personal interface. The centralized interface features a traffic light to compel students to wait before responding to a question; we LEVEL the participation in the class by visually informing students when a question can be answered. The distributed interface adaptively sets a participation time delay on a per student basis predicated upon previous participation. LEVEL brings a wireless sensor network application to primary school classrooms. LEVEL connects student learning with technology while providing an enhanced in-class experience for all students.

Diskless checkpointing is an efficient technique to tolerate a small number of processor failures in large parallel and distributed systems. In literature, a simultaneous failure of no more than N processors is often tolerated by using a one-level Reed-Solomon checkpointing scheme for N simultaneous processor failures, whose overhead often increases quickly as N increases. In this paper, we study an N-level diskless checkpointing scheme to reduce the overhead for tolerating a simultaneous failure of no more than N processors by layering the schemes for a simultaneous failure of i processors, where i = 1, 2, . . . ,N. Simulation results indicate the proposed N-level diskless checkpointing scheme achieves lower fault tolerance overhead than the one-level Reed-Solomon checkpointing scheme for N simultaneous processor failures.

Within the field of event driven data reporting from wireless sensor networks, reducing energy consumption is an ongoing problem. Using an applicationpsilas tolerance toward data imprecision (or, data quality) allows energy savings, via fewer messages sent, through the selection of how and when to send sensor readings. This paper is an early work that examines pushing or pulling data depending on which approach is expected to send fewer messages, based upon the recent history of application requests for a sensor and the changes of the sensor values predicted by application specific models. The simulation results indicate that our method is more efficient relative to the push only or pull only methods for situations where application request frequency or data change rate is variable or unknown.