Qing Yang - US grants

One of the most essential issues in a shared memory multiprocessor system isthe so called memory latency that includes both the main memory access timeand the interconnection network (IN) contention. The introduction of cachememory into parallel computers has been widely accepted as an efficient way ofimproving the system performance. However, multicaches create a cachecoherence problem which must be solved in order for a computer to workcorrectly. Traditional solution techniques for the cache coherence problem inshared memory multiprocessor systems have a potential bottleneck problem whichsignificantly restricted the scalability of multiprocessor systems. This pasresearch project investigates new distributed cache coherence protocols basedon high bandwidth, packet.switched multiple.bus INs. Two configurations ofthe multiple.bus will be considered in this research: hierarchicalstructures of buses and homogeneous single.level multiple buses. It alsoinvestigates analytical models for the cache protocols based on the twostructures to show the performance improvement of the new protocols. Theresearch thus involves (a) study of the correctness and efficiency of the newprotocols; (b) development of analytical models for the systems; (c)performance comparison of the two structures based on different applicationsoftwares; (d) examination of the effects of system parameters on the systemperformance including number of buses, cache block size, number of levels in ahierarchy, and number of modules connected to a bus at each level.

Yang This is a supplement to provide a Research Opportunity Award (ROA) for a faculty member at an urban undergraduate minority institution. The visiting faculty member is studying the caching behavior of an interval Newton method for solving large systems of nonlinear equations. Both analytical and experimental techniques are being used to determine the caching behavior of several algorithms for solving systems of equations. Caching behavior is being studied on both existing vector computers and on the new cache organization that has been designed by the principal investigator.

The main objective of this research is to investigate the issures related to hardware designs and performance evaluations of a number of innovative cache design techniques. These design techniques are currently being developed for single-chip general purpose processors and multiprocessors. Specifically, the focus of the research includes: 1) minimization of area cost of on-chip caches by CAT--caching address tags: 2) maximization of cache hit ratios by evenly distributing data across cache sets with the help of a few tag bits that change frequently during program executions; 3) investigation of potential impacts of the new CAT cache designs for various cache configurations as well as multiprocessor caches; and 4) devising analytical models, and performing trace-driven simulations and execution-driven simulations for evaluating implementation costs, performances, and design trade-offs of various designs.

This research is investigating the issues related to designs and performance evaluations of several innovative disk I/O architectures. These architectures may improve the I/O response time by orders of magnitude over current state-of-the-art disk architectures for office/engineering workloads. The focus of the research includes: * Minimization of disk write times by means of disk caching disk (DCD). This is a technique of using a small disk to cache writes sequentially, as in a log-structured file system, then using the log to update the file system on the larger disks. * Using a combination of DRAM and DCD to replace non-volatile RAM caches that are currently used in RAID systems. * Using DRAM as a component of the file system to allow fast reads and writes. In this scheme, a log of files written to RAM will be kept in either NVRAM or on disk, but only the RAM access will be in the critical path. * Upgrading existing file servers to emulate DCD using multiple existing disks. These architectures will be investigated using simulation to compare them to existing architectures under a wide range of I/O workloads. New analytical models will also be devised and used in evaluation of architectures. Actual implementations of the DCD architectures will be carried out using disk partitions as caches on the computers in the department. Performance measurements will be performed on the DCD implementations and used to compare with existing disk architectures.

We are investigating the macromolecular recognition and inhibition between serine proteases and their inhibitors. In particular, we are studying the origin of the broad inhibitory specificity of ecotin, an E. coli protease inhibitor. The goal of this study is to understand the molecular basis of ecotin's inhibitor specificity and introduce novel specificity into ecotin to target disease-related serine proteases. Based on the high resolution crystal structure of trypsin-ecotin complex, we have designed a series of ecotin variants at the two loops of ecotin and tested these mutants against different proteases. The result suggests a unique mode of inhibition of ecotin through a secondary binding site that does not occur in other serine protease inhibitors. Using phage display techniques, we have isolated several potent inhibitors against a serine protease, urokinase-type plasminogen activator, that plays an important role in cancer. This experiment validates a novel approach to design specific protease inhibitors. The molecular modeling of ecotin loop mutants is accomplished w ith MidasPlus software available in the Computer Graphics Laboratory.

9729839 Uht, Augustus K. University of Rhode Island Research Instrumentation: Equipment for Diverse Computer Architecture Research This research instrumentation grant enables three distinct research projects: - Realizing Instruction Level Parallelism in the 10's - A New Spectrum of Hierarchical Storage Architectures for High Performance Disk I/Os - Ultra-Dependable and High-Performance Distributed Computing while seeking to improve computer performance in many dimensions: uni- or micro-processors, I/O (Input/Output) systems, and multiple-computer systems. Further, the reliability of multiple-computer systems is to be enhanced. These objectives are to be achieved by using the new instrumentation equipment (multiprocessor and multiple-disk RAID storage system) in several different ways. First, the Levo research prototype uniprocessor is to be designed and simulated on the new multiprocessor. The latter will serve to verify Levo's implementation and measure its performance, as well as to allow Levo's construction, establishing that uniprocessors can be designed to execute an order-of-magnitude more instructions per cycle than current uniprocessors. Second, since I/O constraints can limit overall performance, the novel Disk Caching Disk (DCD) I/O system is to be realized on the new multiprocessor and RAID equipment, demonstrating factors of 2 to orders-of-magnitude performance improvements of the DCD architecture. Lastly, the Active Nodal Task Seeking (ANTS) distributed computing system achieves high performance and ultra-dependability in a coarse-grain environment. The new ANTS kernel is to be realized on the new multiprocessor, showing that the same can be achieved on fine-grain applications.

The objective of this project is to investigate the design, implementation, and performance evaluation of a new architecture for storage area networks. The architecture under investigation, Distributed RAID and Locatin Independent Caching (DRALIC) will implement distributed RAID and global caching at the device level, using the embedded controllers in storage devices to perform complex functions. A preliminary study has shown that this architecture has significant performance and reliability advantages in web server applications. The present project includes detailed simulation studies and a proof-of-concept implementation on a network of general-purpose computers.

DESCRIPTION (provided by applicant):The heterotrimeric guanine nucleotide binding protein (G-protein) coupled receptors (GPCR) contain seven hydrophobic membrane-spanning regions. Thus, their intracellular domains consist of three loops and one carboxyl terminus. It is known that the intracellular domains, especially the short fragments near the amino- and carboxyl termini of the third intracellular loop (i3 loop) are critical for interaction with and activate of G-proteins. Recent evidence indicates that some proteins other than G-proteins, so called receptor interaction partners (RIPs), can interact with the carboxyl termini and the i3 loops of some GPCRs. Although only a minute percentage of the RIPs for some GPCRs have been examined in detail, careful studies might yield important clues to the roles of intracellular domains of GPCRs. The long-term objective of my laboratory is to identify novel signaling mechanisms of the angiotensin II subtype 1 receptor (AT1R), and eventually to develop novel therapeutics that target the interface between the receptor and its interacting proteins. The short-term objective of this K award proposal is to define novel signaling protein complexes that are associated with the AT1R intracellular domains, and to explore their functional significance. The initial work will focus on the identification of novel RIPs in rat aortic vascular smooth muscle cells (RASM), which interact with the carboxyl terminus (AT1aR-CT) and/or i3 loop (AT1aR-i3L) of the AT1aR. Our strategy will use glutathione S-transferase (GST) fusion proteins, "pull-down" assays and mass spectrometry. We will approach our short-term goals with two specific aims.1. To identify novel receptor interaction partners (RIPs) in RASM, which interact with AT1aR-CT and/or AT1aR-i3L.2. To identify the domains of AT1aR-CT and/or AT1aR-i3L that interact with the novel RIPsThis project will provide me with the opportunity to blend my previous experience in studying GPCR protein-protein interactions with new opportunities to use mass spectrometric methods to study GPCRs. Because of the emerging importance of the proteome, successful completion of this project will provide me with a solid foundation for developing an independent and self-sustaining research program.

Yang


HELP: A Profiling Tool for Disk I/O and Networked Storage Systems

Abstract


The major goal of this proposed research is to design and implement a hardware environment for low-overhead profiling/optimization, named HELP. HELP is a framework consisting of a hardware embedded system board and a set of easy-to-use APIs to allow system architects to develop their own efficient profiling and optimization tools for storage operations. A HELP board can be directly plugged into a server or storage system to speed up storage operations. Unlike most existing profiling and optimization techniques, our approach minimizes profiling overheads and data skews resulting in more accurate analysis of disk I/O behavior. By offloading profiling/optimization functions from the host, HELP makes it possible to do runtime monitoring, collecting, analyzing, and optimizing disk I/O and data storage operations in a production system. A system architect is able to explore more opportunities than ever in designing tools for profiling and optimization, studying behavior of storage accesses, and inventing new storage architectures using HELP.

Successful completion of the proposed research will result in an effective tool that will have great impact on computer architecture research as well as broader IT community. Researchers can benefit greatly from our HELP system in guiding their search for new architectures and new algorithms for data storage. IT industries can also benefit greatly from our tool in making design choices for storage systems and solutions. It will also have significant impact on computer architecture education by facilitating better understanding of detailed system behavior.

Abstract Not Provided.

bioimaging /biomedical imaging

DESCRIPTION (provided by applicant):The heterotrimeric guanine nucleotide binding protein (G-protein) coupled receptors (GPCR) contain seven hydrophobic membrane-spanning regions. Thus, their intracellular domains consist of three loops and one carboxyl terminus. It is known that the intracellular domains, especially the short fragments near the amino- and carboxyl termini of the third intracellular loop (i3 loop) are critical for interaction with and activate of G-proteins. Recent evidence indicates that some proteins other than G-proteins, so called receptor interaction partners (RIPs), can interact with the carboxyl termini and the i3 loops of some GPCRs. Although only a minute percentage of the RIPs for some GPCRs have been examined in detail, careful studies might yield important clues to the roles of intracellular domains of GPCRs. The long-term objective of my laboratory is to identify novel signaling mechanisms of the angiotensin II subtype 1 receptor (AT1R), and eventually to develop novel therapeutics that target the interface between the receptor and its interacting proteins. The short-term objective of this K award proposal is to define novel signaling protein complexes that are associated with the AT1R intracellular domains, and to explore their functional significance. The initial work will focus on the identification of novel RIPs in rat aortic vascular smooth muscle cells (RASM), which interact with the carboxyl terminus (AT1aR-CT) and/or i3 loop (AT1aR-i3L) of the AT1aR. Our strategy will use glutathione S-transferase (GST) fusion proteins, "pull-down" assays and mass spectrometry. We will approach our short-term goals with two specific aims.1. To identify novel receptor interaction partners (RIPs) in RASM, which interact with AT1aR-CT and/or AT1aR-i3L.2. To identify the domains of AT1aR-CT and/or AT1aR-i3L that interact with the novel RIPsThis project will provide me with the opportunity to blend my previous experience in studying GPCR protein-protein interactions with new opportunities to use mass spectrometric methods to study GPCRs. Because of the emerging importance of the proteome, successful completion of this project will provide me with a solid foundation for developing an independent and self-sustaining research program.

Project Summary

The PI proposes to validate and evaluate a new technology for mirroring data from a primary storage to mirror storage across a LAN/WAN network for the purpose of data protection. The idea of the new technology is to mirror only the parity resulting from a write operation instead of the data itself. By
leveraging the parity computation that exists in RAID storage systems, we mirror only
the new parity to the remote site whenever a disk write operation is performed.

Intellectual Merit

Mirroring-only-parity (MOP) results in substantial data reduction making it possible to
do real-time data mirroring over limited bandwidth network without negatively impacting
application performance. Data can be computed back using the received parity and
previously existing data at mirror storage. The new method trades off inexpensive and
fast computations outside of the critical data path for high cost and slow communication.
Preliminary analyses show that an order of magnitude data reduction is possible
without noticeable overheads that are generally required by existing technologies such as
compression algorithms and TCP/IP optimizations.

Broader Impact

The impact of the proposed research is expected to be dramatic and broad. Industries such as
financial, health care, information technology, manufacturing, governmental, research/education
all rely on digital data that needs to be protected from any unexpected disasters.If validated, our new technology makes it possible for any organization to inexpensively deploy remote data mirroring in
real time, which has not been possible because of the prohibitively high cost and intolerable delays of traditional technologies. The proposed research will have educational impact because the proposed experimental activities will be mainly carried out by graduate and undergraduate students.

Research Objectives and Approaches: The objective of this research is to develop ultra-sensitive and compact magnetometers, featuring the direct on-chip integration of magnetoelectric sensors, with advanced microelectronics, for non-invasive medical imaging. The approach is based on the potential of picoTesla magnetic sensitivity using integrated magnetoelectric composites. This research aims to further improve the sensitivity beyond those based on the direct stress mediated magnetoelectric coupling by improving the mechanical impedance match between the constituents, by exploring totally new magnetoelectric coupling, and by incorporating MEMs-based structures as resonant moving gates of MOSFETs for further amplification.

Intellectual Merit: The demonstration of a heterogeneous integration strategy at the devices and circuits level that seamlessly merges microelectronics with sensors holds great promise of enabling transformational shifts in the semiconductor and sensor industry. Further, the proposed research will provide fundamental understanding on the magnetic energy transfer efficiency between the constituents of the composites. The research will also provide insight into design and fabrication of resonant moving gate transistors that markedly improve the sensor performance.

Broader Impact: The successful development of chip scale magnetic sensors will revolutionize the field of biomagnetic field detection and imaging. Moreover, the availability of low cost diagnostic tools will greatly facilitate early disease detection. On the education and outreach font, this research will foster an interdisciplinary educational program that will inspire students to understand fundamental materials science and device physics and incorporate them into solid-state device fabrication to address complex detection and diagnosis problems in the life science and biomedical imaging area.

Data protection and recovery have become increasing important as business, education, and government depend more and more on digital information.
Failure events do occur such as virus attacks, user errors, defective software/firmware, hardware faults, and site failures etc that cause data damage. To ensure business continuity and minimize loss, data storage systems need data protection and recovery techniques. However, existing technologies have severe limitations and unable to recover data in many situations. This project aims at studying and understanding how data recovery is done in existing data storage systems, and designing new architectures that will overcome the limitations of existing technologies.

In order to study and understand the existing storage architectures, a new mathematical formulation will be developed to model and analyze capabilities and limitations of the storage architectures. This mathematical model provides a rigorous tool for researchers and practitioners to investigate and understand storage system architectures. Based on the new mathematical model, a class of new data storage system architectures will be designed that will have the maximum data recoverability. The new storage architectures make it possible for organizations of different sizes to have a cost-effective data storage that provides high data availability and allows quick data recovery upon failures. In addition to the theoretical study, experimental prototypes will be developed and implemented to demonstrate the feasibility, performance, reliability, and data recoverability of newly designed storage architectures. Furthermore, the project includes an education component that advocates a shift of emphasis from CPU-centric computer engineering (CE) curriculum to data-centric CE curriculum. The new curriculum provides CE students with in depth knowledge of data processing, data communication, and data storage.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).

The objective of this research is to develop a trustworthy and high-performance neural-machine interface (NMI) that accurately determines a user?s locomotion mode in real-time for neural-controlled artificial legs. The proposed approach is to design the NMI by integrating a new pattern recognition strategy with a high-performance computing embedded system.

This project tackles the challenges of accurate interpretation of information from the neuromuscular system, a physical system, using appropriate computation in a cyber system to process the information in real-time. The neural-machine interface consists of multiple sensors that reliably monitor the neural and mechanical information and a set of new algorithms that can fuse and coordinate the highly dynamic information for accurate identification of user intent. The algorithm is to be implemented on a high-performance graphic processing unit (GPU) to meet real-time requirements.

This project has the potential to enable the design of neural-controlled artificial legs and may initiate a new direction for research in and the design of prosthetic leg systems. Innovations in this domain have the potential to improve the quality of life of leg amputees, including soldiers with limb amputations. The proposed approaches seek to permit cyber systems to cope with physical uncertainty and dynamics, a common challenge in cyber-physical systems, and to pave a way for applying high-performance computing in biomedical engineering. Besides providing comprehensive training to undergraduate and graduate students, the investigators plan to introduce community college students to cyber-physical systems concepts in an interactive and engaging manner.

While storage capacity and CPU processing power have experienced rapid growth in the past, improvement in data bandwidth and access times of hard disk drives (HDD) has not kept pace. As a result, the speed gap between CPUs and disk I/O is widening. Disk arrays can improve overall I/O throughput but random access latency is still very large because of mechanical operations involved. Large buffers and deep cache hierarchy can improve latency, but the access time reduction has been very limited so far because of poor data locality at the disk storage level.

This proposal aims at rethinking the fundamental architecture of storage systems and makes an attempt at a paradigm shift of disk based storage architectures. The approach is to build a new storage architecture that exploits the two emerging semi-conductor technologies: flash memory SSD (solid state disks) and GPU (graphic processing unit). The new disk I/O architecture is referred to as I-CASH: Intelligently Coupled Array of SSDs and HDDs. The SSD is used to store mostly read "reference data blocks" to make best use of its high-speed random read performance. The HDD is used to store compressed delta between a current I/O block and its corresponding reference block in the SSD so that random writes are not performed on SSD during online I/O operations. The SSD and HDD are controlled by a high speed GPU that performs similarity detection, delta derivations, combining delta with reference blocks, and other necessary functions for interfacing the storage to the host OS. The idea is to leverage fast read performance of SSDs and the high speed computation of GPUs to replace and substitute, to a great extent, the mechanical operations of HDD to achieve I/O performance that is orders of magnitude better than traditional disk storage systems. Instead of working on HDD to catch up with processors' performance, which has been proven difficult if not impossible, the proposed approach lets storage systems ride the wave of the rapid advancement of multicore processors and be part of such success by trading high speed computation for low access latency.

It is anticipated that the proposed project will have significantly broad and transformative impact. 1) Servers at data centers run tens and hundreds of virtual machines that generate large amount of I/Os that can take full advantage of our new storage architecture with potentially orders of magnitude performance improvement. 2) The research will engage both graduate and undergraduate students so that they are ready for the real world need. 3) The success of this research will help the economic development of the state of Rhode Island and the nation.

This project studies a new computing and communication architecture, reflex-tree, with massive parallel sensing, data processing, and control functions designed to meet the challenges imposed by future smart cities. The central feature of this novel reflex-tree architecture is inspired by a fundamental element of the human nervous system -- reflex arcs, or neuro-muscular reactions and instinctive motions in response to urgent situations that do not require the direct intervention of the brain. The scientific foundation and engineering framework built by this project will pave the way for enhanced monitoring and management of critical smart city infrastructure, from gas/oil pipelines, water management, communication networks, and power grids, to public transportation and healthcare. The interdisciplinary and collaborative nature of the project will inspire broader participation in related areas of research.

Within the human body, a neural reflex arc is able to cause an individual to immediately react to a source of discomfort without the need for direct control from the brain. The reflex-tree architecture mimics such human neural circuits, using massive numbers of intermediate computing nodes, edge devices, and sensors to gather, process, and, most importantly, to react to data concerning critical infrastructure elements. Key innovations of the proposed reflex-tree architecture include: 1) A novel, 4-level, large scale, and application-specific hierarchical computing and communication structure capable of carrying out sensor-based decision-making processes. The required computation and nodal computing power increases at each successive stage in the hierarchy, with the level-1 cloud performing the most complex tasks. 2) A densely distributed fiber-optic sensing network and parallel machine learning algorithms will be developed targeting smart city applications. 3) Novel, complementary machine intelligence algorithms will be developed, providing accurate control decisions via multi-layer adaptive learning, spatial-temporal association, and complex system behavior analysis. 4) New parallel algorithms and software run-time environments will be proposed and developed that are specifically tailored to the novel reflex-tree system architecture.

To demonstrate the feasibility and performance of the reflex-tree architecture, a proof-of-concept prototype will be constructed utilizing a miniaturized, laboratory-scale municipal gas pipeline system. The prototype will incorporate a complete 4-level reflex-tree--a distributed fiber-optic sensing network deployed alongside pipelines, edge devices performing data classification using parallel SVM, intermediate nodes performing massively-parallel spatial and temporal machine learning, and the cloud as the root node running sophisticated parallel behavioral analysis and decision making tasks. The resulting system is a cross layer, high performance, and massively parallel computing platform, providing a foundational sensing and computer architecture for future smart cities.

? DESCRIPTION (provided by applicant): Magnetic resonance (MR) imaging (MRI) and in vivo MR spectroscopy (MRS) techniques have become indispensable tools for imaging brain structure, function, connectivity, neurochemistry and neuroenergetics, and for investigating neurological disorders. However, it remains a challenge to achieve superior MRI/MRS detection sensitivity, spatial and temporal imaging resolutions adequate for addressing fundamental and challenging neuroscience questions even with the most advanced technology. The prevailing paradigms for improving MRI/MRS performance largely invoke increasing the magnetic field strength, which may have reached practically achievable limits for human studies due to many technological and safety (i.e., high specific absorption rate (SAR)) concerns, and increasing the receiver channel count which is also ultimately limited due to noise characteristics of coils of decreasing size. To alleviate these major limitations, this R24 proposal relies on the interdisciplinary research efforts and expertise of leading experts across two institutions to pioneer an entirely innovative engineering solution that uses the ultra-high dielectric constant (uHDC) material incorporated with ultrahigh-field MRI/MRS techniques for synergistically increase signal-to-noise ratio and concurrently reduce RF power demand, and for achieving unprecedented improvements in spatial/temporal resolution over the current state-of-the-art MR technologies. We will develop and optimize prototypes of uHDC material for human brain studies using 7 Tesla (T) and 10.5T whole-body human scanners. Moreover, we will exploit and assess the new utility and capability of the innovative uHDC-MR technology for cutting-edge neuroscience research. One pilot study is the functional mapping of neural circuits and resting-state connectivity at the level of columns and cortical layers in the human visual cortex with ultrahigh spatial resolution 1H MRI at 7T, complemented with anatomical connectivity derived from diffusion weighted images for tractography. The other one is to combine the uHDC technique with newly developed in vivo 31P and 17O MRS techniques for noninvasively and reliably imaging the cerebral metabolic rates of oxygen consumption and ATP, cerebral blood flow, oxygen extraction fraction and nicotinamide adenine dinucleotide (NAD) redox state in the human brain at resting and activated states. The proposed research will shift the current paradigm of neuroimaging development towards an efficient, cost-effective engineering solution that will attain multiplicative gains from uHDC and ultrahigh fields, and lead to next generation o MRI/MRS technology and instrument. Such advancement will accelerate human brain imaging and neuroscience research beyond what can be achieved through existing technology, promote new research directions, and transform our understanding regarding the human brain function and dysfunction.

Big data applications demand high speed, reliable, and energy
efficient data storage systems. Traditional storage architectures have
fundamental limitations because of legacy systems that have centered
on spinning hard disk drives. With rapid advances in nonvolatile
memory technologies such as NAND-gate flash, phase change memory, Memristor, and
magnetic RAM, a great opportunity arises for revolutionizing storage
architectures. The objective of this research is to start a paradigm
shift in storage architecture to meet the increasing demand of big
data applications. It is envisioned that future storage systems will
have machine intelligence that learns, analyzes, predicts, and
controls the system at runtime dynamically. A novel accelerator
architecture is introduced with machine intelligence to enable high
speed processing of storage data operations that are critical to high
performance computing in general and big data computing in particular.

The newly introduced I/O accelerator, residing either in a
many-core CPU chip or on a storage controller board, enables
sufficiently accurate predictions for effective optimization of
storage I/Os. With new architecture features, the proposed I/O
accelerator can carry out complicated I/O tasks in the speed
comparable to the emerging nonvolatile memories, which is critical to
I/O performance because it no longer operates in milliseconds as
spinning disks do. The project will explore and implement the I/O
accelerator that can effectively deal with the complexity and high
dimensionality of factors related to diverse storage technologies, a
large variation of application workloads, different
reliability/availability requirements, and power consumptions of
various storage components. The result is a new heterogeneous storage
architecture that is optimized for future computing infrastructure.
With the accelerator as an enabler, comprehensive methodology will be
investigated that proactively learns system behavior to anticipate
long-term trends and to respond quickly to fast changing I/O events.
The new architecture is believed to be the first of the kind providing
dynamic optimizations by means of 1) intelligent data placements and
replacements across heterogeneous devices, 2) optimal resource
allocation and provisioning to applications' workloads, 3) effective
data deduplication based on content locality, and 4) smart policy
decision on data protection and recovery adaptive to different data
types. Furthermore, the new accelerator enables fast in-situ data
analytics in active storage systems.

This research project is expected to have the following broader impacts:
1) In today's cloud computing and big data
applications, servers generate a large amount of I/Os that can take
full advantage of the new storage architecture. 2) The new accelerator
can be incorporated into many core CPUs as a specialized core for
future heterogeneous processors. 3) The new storage architecture will
speed up the adoption of emerging storage class memories. 4) The new
methodology will stimulate more research in applying machine learning
to storage systems. 5) The new CPU-and-data centric Computer Engineering curriculum will train
both graduate and undergraduate students for real world needs. 6) The
outreach program will continue the success stories of prior NSF
projects to help the economic development of the state of Rhode Island and the
nation.

The University of Montana in Missoula, in partnership with Montana State University in Bozeman, Montana Tech (MTech) in Butte, and the Salish Kootenai Tribal College (SKC), will undertake a planning effort for a larger statewide initiative to introduce CS high school courses that engage and retain a diverse population of students, promote and advocate for these courses, and prepare high school teachers to offer and sustain them. The expanded curriculum will include material from the Joy and Beauty of Computing course recently created and piloted by one of the collaborating institutions of this proposal (MSU), and the new AP principles course currently being developed by NSF and the College Board. Dual credit opportunities will be available for students. In addition, the curriculum will be supported with activities proven to improve diversity (mentoring, pair programming, role models). High school teachers will participate via team-teaching, curriculum development, and professional development (PD) workshops.

This planning grant would enable the PIs to build on their curriculum within the context of the CS10K Project with initial research and development around diversity (in particular, strategies for engaging and retaining Native American students), curriculum (positioning their course offerings in the context of the CS10K effort), and teacher professional development.

Project summary Paclitaxel-induced peripheral neuropathy (PIPN) and associated neuropathic pain is the most common and serious adverse effect experienced by cancer patients accepted paclitaxel infusion, which adversely affects daily activities and thereby quality of life, and sometimes forces the suspension of treatment, negatively impacting survival. However, mechanisms underlying the pathogenesis of PIPN are uncertain, which hinders the development of effective therapies for this comorbidity. The proposed studies attempt to identify molecular mechanisms underlying PIPN, as well as a potential therapeutic target to prevent the development of PIPN and neuropathic pain. Excessive neuronal excitation is a primary source of PIPN. Our preliminary data indicate that the hyperexcitability of primary sensory neurons might result from paclitaxel-induced inhibition of KCNQ/Kv7 channels, which are abundant in sensory neurons and axons. Retigabine, an FDA-approved drug that opens KCNQ/Kv7 channels, could be a plausible treatment to reduce paclitaxel-induced pathology and symptoms. We hypothesize that paclitaxel induces peripheral neuropathy and chronic pain by inhibiting KCNQ/Kv7 channels and exciting primary sensory neurons, activating KCNQ/Kv7 channels during chemotherapeutic agent infusion may thus prevent the development of PIPN. In this project, we will produce PIPN in adult, tumor-free rats or mice, and utilize techniques of immunohistochemistry, electrophysiology, electron microscopy, and behavioral testing to test three important predictions: 1) Paclitaxel excites primary sensory neurons by inhibiting KCNQ/Kv7 channels. Paclitaxel-induced effects on KCNQ currents in CHO cell lines in which KCNQ2/3 are overexpressed, as well as its effects (KCNQ currents and membrane potential) on DRG neurons from naïve rats, Kcnq2fl/fl//Pax3-Cre, Kcnq3-/-, and their littermate control mice will be assessed electrophysiologically (in vitro recording); 2) Paclitaxel induces peripheral neuropathy and chronic pain by inhibiting KCNQ/Kv7 channels. XE-991, a selective KCNQ/Kv7 channel blocker, will be delivered to naïve rats to see whether XE-991 can simulate PIPN and chronic pain. The role of KCNQ/Kv7 channels in PIPN will then be evaluated by exposing of Kcnq2fl/fl//Pax3-Cre, Kcnq3-/-, and their littermate control mice to paclitaxel. Pain-related behavior, morphological alterations (gliosis in the spinal cord, IENF in the skin, ROS in DRG neurons, mitochondria and microtubules in nerve sections), and neuronal excitability will be assessed; 3) Combining retigabine with paclitaxel can prevent the development of peripheral neuropathy and neuropathic pain. Retigabine will be given to rats during the exposure to paclitaxel. The excitability of DRG neurons, gliosis in spinal cord, IENF in the epidermis, and pain-related behaviors will be measured. Finally, the chemosensitivity of a breast cancer tumor to paclitaxel will be assessed in the presence of retigabine. These studies may lead both to a better understanding of the causes of PIPN as well as delineate novel targets for therapeutic development.

The University of Montana is building a collaborative effort involving three major state universities and a tribal college. Its two main goals are to prepare high school teachers to teach new computer science (CS) courses in a manner that is sustainable over the long-term, and to expand CS curriculum in Montana's high schools with a special focus on improving the diversity of students who engage in CS. This project represents the first of its kind in the state of Montana due to its scope, the partnerships involved, and the potential to expand the number of students introduced to a wide variety of CS topics at an earlier age. The end result promises to be a broader population of high school students who engage with CS, who go on to major in a computing field, and who ultimately land in successful, high-paying and fulfilling professions. This project can change the economy and the future of Montana, and prepare a more diverse group of young people to take over the next generation of computing in our country.

To expand the current offerings of CS courses in Montana high schools, this effort includes the Joy and Beauty of Computing (JBC) course developed and piloted at Montana State University and rolled out to Bozeman High School, and a second CS Principles course. Activities shown to improve diversity (mentoring, pair programming, role models, etc.) will be integrated into course offerings, dual credit opportunities will be available for students, and high school teachers will participate in team-teaching arrangements and on-going professional development (PD) workshops. The educational assessment portion of the project centers on the high-level question: Does the two-course CS curriculum broaden interest and participation in high school CS career pathways? A set of sub-questions that focus separately on student participation and teacher participation will guide the data collection and analysis portion of this study. Six critical measures will be used: pre/post surveys, interviews, CS course enrollment data, artifact reviews, classroom observations, and case study. Formative and summative summaries will be provided each year by Cedar Lake Research Group to inform the research.

Vehicular networks are becoming increasingly popular. To make them truly useful, irrelevant information exchanges among vehicles should to be eliminated to avoid unnecessary driver distraction. This project aims to tackle this fundamental problem, wherein what information is delivered to which vehicle(s) is intelligently determined. The project will study the closeness between vehicles based their interactions, in the form of information exchange, so a driver can determine whether a received message is relevant based on the closeness information. Because information is filtered by a vehicle's close 'friends', the amount of irrelevant information it receives will be reduced, and thus efficient information dissemination is achieved. The research will produce an efficient information dissemination system that complements and enhances existing intelligent transportation systems, connected vehicles, and vehicular telematics. The project will also include efforts to deploy the system to offer a better information provision service to drivers. Two PhD students and several undergraduate students will be trained in this project.

The researchers propose to use interactions between vehicles to estimate their closeness, and most importantly, to determine what data should be delivered to which vehicle(s) based on the closeness information. The key to their approach is constructing a vehicular social network (VSN) that enables drivers to integrate their social network with vehicular network. The list of points of interest (POIs) that a vehicle visited is considered its genome, and vehicles with similar genetic features are considered initially connected in a VSN. These connections are then cultivated by the interactions among vehicles. With positive, negative, and uncertain interactions, the closeness between two vehicles having direct interactions is modeled as a Dirichlet distribution. For vehicles that have no direct interactions, their closeness is inferred from the social network between them. The PIs will design a polynomial-time solution to addressing the massive closeness assessment problem, i.e., computing the closeness from a driver to all others in a VSN. The researchers also propose an efficient algorithm for the all-pair closeness assessment problem, i.e., computing the closeness of any pair of vehicles in a VSN. A cloud-hosted service is proposed to coordinate social connection construction, VSN maintenance, closeness assessment, and information dissemination.

By understanding what and how data are exchanged among autonomous vehicles, from a machine learning perspective, it is possible to realize precise cooperative perception on autonomous vehicles, enabling massive amounts of sensor information to be shared amongst vehicles. Such an advance can be extremely useful to extend the line of sight and field of view of autonomous vehicles, which otherwise suffers from blind spots and occlusions. The extended field of view on autonomous vehicles will be beneficial at times when there are occlusions preventing a complete perception of the environment. This increase in situational awareness promotes safe driving over a narrow scope and improves traffic flow efficiency over an extended scope. The proposed research work will not only change the way people think about the perception system on autonomous vehicles but could also open up opportunities to design novel systems that were previously inconceivable. This project offers a wide variety of research activities from data collection, algorithm design, system development, and in-the-field evaluation, which will be attractive to students with various backgrounds and goals. Undergraduate and graduate students will be involved directly in the research activities as assistants at different levels. The expected research outcomes from this project will also enhance the current curricula related to machine learning, Internet of things, and wireless communications.

The main research objective of this project is to understand the sensing and communication challenges to achieving cooperative perception among autonomous vehicles, and to use the insights thus gained to guide the design of suitable data exchange format, data fusion algorithms, and efficient millimeter wave vehicular communications. Results from this project will include a machine learning based cooperative perception framework, which will shed light on effectively combining feature maps, derived from machine learning models on autonomous vehicles, in a distributed manner. The resulted feature map compression and feature map selection approaches will significantly reduce the amount of data exchanged among vehicles, enabling agile and precise cooperative perception on connected and autonomous vehicles. The proposed scalable feature map transmission mechanism jointly considers the application requirements, link and physical layer characteristics of millimeter wave links, enabling sensor data sharing on a massive scale among autonomous vehicles. The implemented system and evaluation platform will serve as a convincing proof-of-concept for the proposed solution, thus opening the door to widespread adoption of cooperative perception applications via millimeter wave communications in future vehicle networks. The collected dataset from this project will be made publicly available, serving as a catalyst for enabling innovative research on cooperative object detection, vehicular edge computing, and machine learning.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

It is now a national research priority to design and develop novel computer hardware for artificial intelligence as well as smart sensing with capabilities for high-performance and energy-efficient computing. This project proposes a new architecture, PARIS (Phased Array Radar with In-Sensor Computing), that simultaneously senses and processes 3D images in real time. The new architecture mimics the human visual system, which not only detects and senses 3-D visual images but also performs first-stage image processing before the more complex processing in the visual cortex of brain. It is the first in-sensor computing architecture leveraging a phased-array radar system for an energy-efficient, high-performance, low-cost, and compact sensing/computing platform. By combining phased-array radar imaging and neuromorphic computing, PARIS opens up a new avenue for research in in-sensor computing and intelligent image processing. The project has transformative impact on a wide range of industries including medical instruments, autonomous vehicles, machine vision, robotic control, IoT devices, smartphones, and consumer electronics. Research activities of the project are involving female and minority students, strengthening the PIs’ current K-12 outreach activities, and enhancing grand-challenge courses for all majors at the university.

It is well-known in artificial-intelligence systems that moving raw data from sensors to processing elements is very costly in terms of energy, performance, and hardware. The newly proposed architecture starts artificial neural-network computation concurrently while sensors are acquiring data, substantially reducing data movement between sensors and processing elements. Such simultaneous sensing and computation are made possible by several technology breakthroughs. 1) In-sensor dot-product computations with linearly tunable weights allowing in-sensor training and inference of artificial neural network; 2) a novel statistical signal acquisition technique, namely, Jitter-based Analog-to-Probability Conversion, allowing for direct use of digital pins of integrated circuits for high-bandwidth measurement with low-overhead; 3) a compact single-board system supporting very high radio frequency bandwidths and completely removing the radio frequency analog front-end required by conventional microwave imaging systems (e.g. analog-to-digital converter, filter, or amplifier) by leveraging the very short rising/falling edges of the digital waveforms used in today’s I/O interfaces. The theory and design of the in-sensor computing architecture is being established, and a 16-node prototype is being built. A thorough evaluation and comparison is being conducted to demonstrate its performance, energy efficiency, hardware cost, and potential applications to a wide range of industries. Research results are being published in professional conferences and journals and incorporated into undergraduate electrical/computer engineering curricula.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

While autonomous vehicle companies focus on manufacturing driverless cars, electrification and connectivity technologies can upgrade the industry by allowing vehicles to communicate with the driving circumstance to achieve better safety with less emission. A key obstacle to the development of the electric, connected and autonomous vehicle system is the shortage of synergy research from needed domains. The proposed Center for Electric, Connected and Autonomous Technologies for Mobility (eCAT) will concentrate on interdisciplinary research, aiming to initiate and accelerate the transformation of mobility methods from conventional vehicles to electric, connected and autonomous vehicles by creating innovative electric, connected and autonomous technologies.

A partnership between Wayne State University (WSU), University of North Texas (UNT), and Clarkson University (Clarkson), the center not only serves as an apparatus of academic researchers collaborating with industry on important problems, but also provides industry partners opportunities to access advanced synergic research produced from a diverse group of researchers. The UNT Site will particularly focus on developing novel perception algorithms and system architecture to support connected autonomous vehicles. The UNT’s eCAT will leverage the Connected Autonomous Vehicular System Lab, the Center for Integrated Intelligent Mobility Systems, and long-running relationships with industry to bring significant contributions to the center.

The Center will not only advance the science and technologies of electric, connected and autonomous vehicles, but also accelerate both knowledge and intellectual property transfer between academia and industry through collaborative partnerships. This will allow for the rapid transformation of the state-of-the-art mobility technologies from research labs to industry. Educational efforts will be devoted to (1) curriculum design for the BS/MS programs, (2) summer camp development for K-12 students, (3) broadening participation in computing and engineering, at WSU, UNT, and Clarkson. Research results will be disseminated through our annual conference - MetroCAD, publications, and the development of publicly-available open-source projects.

Data obtained from this project will include experimental, computational, text-based, and/or curricular data. All data products will be retained for a minimum of three years after conclusion of the award or three years after public release (publication), whichever is later. WSU, UNT, and Clarkson will have password-protected shared network drives to facilitate data sharing within the teams, with collaborators, and indeed anyone in the community with an interest in the NSF-funded data. A landing page for navigation to archived data will be created at http://ecat.center.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.