6,610 Works

Physicalnet: A Framework for Programming Mobility-Aware, Cross-Network, Concurrent Applications for Sensor and Actuator Networks

Pascal Vicaire & John Stankovic
This paper describes the design, implementation, and eval- uation of Physicalnet, a framework allowing the programming and concurrent execution of applications that span multiple sensor and actuator networks. Physicalnet uses a service ori- ented architecture to make node services globally accessible and interoperable. Within this architecture, dedicated pro- cesses list the services available at a given geographical site, allow the concurrent access to these services, and enforce spec- ified user access rights. On top of...

A Discussion on the Thermal Benefit of Multicore Floorplanning at the Microarchitectural Level

Karthik Sankaranarayanan, Mircea Stan & Kevin Skadron
This paper presents research to address the temperature challenge in multicore processors through the lever of thermally-aware floorplanning. Specifically, it examines the thermal benefit in a variety of placement choices available in a multicore processor including alternative core orientation and insertion of L2 cache banks between cores as cooling buffers. In comparison with an idealized scheme that scatters the functional blocks of a multicore across the entire chip area to maximize uniformity, a combination of...

Toward an Architectural Treatment of Parameter Variations

Eric Numenay, Wei Huang, Mircea Stan & Kevin Skadron
This paper develops a new model of parameter varia- tins for use in early-stage, pre-RTL architecture studies. It improves over prior models by extending the FMAX model to more faithfully model various microarchitecture struc- tures, especially SRAM, which is dominant in contempo- rary superscalar processors. It also incorporates optical phenomena, which show strong spatial correlation but nev- ertheless cannot be ignored for large dies. Finally, it incor- porates IR Vdd drop and temperature, and closes...

Query Formulations for Information Retrieval by Intelliginer Analysts

Ziangyu Jin, James French & Jonathan Michel
There have been many claims to the effect that advanced information retrieval technologies such as the vector space model (VSM) significantly outperform Boolean search engines in retrieval effectiveness. Most of these claims are anecdotal and rely on arguments such as: queries are diffi- cult to construct or output size is hard to control. In spite of these claims, most operational re- trieval systems in use today are Boolean. Some of the popularity derives from the...

Braid: Integrating Task and Data Parallelism

Emily West & Andrew Grimshaw
Archetype data parallel or task parallel applications are well served by contemporary languages. However, for applications containing a balance of task and data parallelism the choice of language is less clear. While there are languages that enable both forms of parallelism, e.g., one can write data parallel programs using a task parallel language, there are few languages which support both. We present a set of data parallel extensions to the Mentat Programming Language (MPL) which...

On a Constrained Bin-packing Problem

Yingfeng Oh & Sang Son
We study a bin-packing problem which is one-dimensional and is constrained in the manner items are placed into bins. The problem is motivated by a practical real-time scheduling problem, where redundant periodic tasks need to be assigned to a multiprocessor system. The problem is stated in the traditional light: to use as few bins as possible to pack a given list of items, and it is a generalization of the classical bin-packing problem. We first...

Distributed Error Recovery for Continuous Media Data in Wide-Area Multicast

Matthew Lucas, Bert Dempsey & Alfred Weaver
This paper proposes new mechanisms to improve the quality of wide-area disseminations of time-sensitive streams, such as packet voice and video, from a single source to multiple receivers. Most current packet-switched networks provide no end-to-end quality-of-service guarantees for packet delay or loss, which can reduce the playback quality at receivers. Buffering at the multicast receivers can be used to minimize the playback degradation due to network delay variations. Addressing packet losses is more problematic. Solutions...

An Isotach Network Simulator

Lance Hopenwasser
We have developed a simulator of Isotach systems which run on a Myrinet network. The simulator allows us to study the performance characteristics of the system and provides a means for experimenting with alternative implementations of Isotach networks. The rate at which tokens flow through the network determines the rate at which Isotach logical time progresses; therefore, token behavior is of keen interest. The simulator gathers performance data that is used to examine the behavior...

Automatic Counterflow Pipeline Synthesis

Bruce Childers & Jack Davidson
The Counterflow Pipeline (CFP) organization may be a good target for synthesis of application-specific microprocessors because it has a regular and simple structure. This paper describes early work using CFP's to improve overall application performance by tailoring a CFP to the kernel loop of an application. A CFP is customized for an application using the kernel loop's data dependency graph to determine processor functionality and interconnection. Our technique builds the design space for a given...

Partial Security Policies to Support Timeliness in Secure Real-Time Databases

Sang Son, Craig Chaney & Norris Thomlinson
Conflicts in database systems with both real-time and security requirements can be unresolvable. We address this issue by allowing a database system to provide partial security in order to improve real-time performance when necessary. Systems that are partially secure allow potential security violations such as covert channel use at certain situations. We present the idea of requirement specification that enables the system designer to specify important properties of the database at an appropriate level. To...

Rapid Prototyping of Application-Specific Counterflow Pipelines

Bruce Childers & Jack Davidson
Application-specific processor (ASIP) design is a promising approach for meeting the performance and cost goals of an embedded system. We have developed a new microarchitecture for automatically constructing ASIP's. This new architecture, called a wide counterflow pipeline (WCFP), is based on the counterflow pipeline organization proposed by Sproull, Sutherland, and Molnar. Our ASIP synthesis technique uses software pipelining and design space exploration to generate a custom WCFP and instruction set for an embedded application. This...

Statistical Multiplexing Gain of Link Scheduling Algorithms in QoS Networks (Short Version)

Boorstyn Robert, Burchard Almut, Liebeherr Jorg & Oottamakorn Chaiwat
This report is an abbreviated version of CS-99-21. A statistical network service which allows a certain fraction of traffic to not meet its QoS guarantees can extract additional capacity from a network by exploiting statistical properties of traffic. Here we consider a statistical service which assumes statistical independence of flows, but does not make any assumptions on the statistics of traffic sources, other than that they are regulated, e.g., by a leaky bucket. Under these...

MetaSystems: An Approach Combining Parallel Processing and Heterogeneous Distributed Computing Systems

Andrew Grimshaw, Jon Weissman, Emily West & Jr Loyot
Large metasystems comprised of a variety of interconnected high-performance architectures are becoming available to researchers. To fully exploit these new systems, software must be provided that is easy to use, supports large degrees of parallelism in applications code, and manages the complexity of the underlying physical architecture for the user. This paper describes our approach to constructing and exploiting metasystems. Our approach inherits features of earlier work on parallel processing systems and heterogeneous distributed computing...

reFresh SSDs: Enabling High Endurance, Low Cost Flash in Datacenters

Vidyabhushan Mohan, Sriram Sankar & Sudhanva Gurumurthi
Storage performance and power are critical issues in modern datacenters. Solid State Drives (SSDs) offer both performance and power advantages over hard disk drives. With the advent of MLC flash, the cost-per-Gigabyte of Flash has dropped significantly enough to make it attractive for use in large- scale storage in datacenters. However, flash suffers from limited endurance and wears away after a certain number of writes and erases, thereby entailing higher replacement and service costs and...

TMMAC: ATDMA Based Multi-Channel MAC Protocol Using a Single Radio Transreceiver for Mobile Ad Hoc Networks

Jingbin Zhang, Gang Zhou, Chengdu Huang, Ting Yan, Sang Son & John Stankovic
In this paper, we propose amedium access control (MAC) protocol called TMMAC for Mobile Ad Hoc Networks (MANETs). TMMAC requires only a single half-duplex radio transceiver on each node. In addition to explicit frequency negotiation which is adopted by the conventional multi-channel MAC protocols, TMMAC introduces lightweight explicit time negotiation. This two-dimension negotiation enables TMMAC to exploit the advantage of both frequency diversity and TDMA. Performance evaluation results show that TMMAC achieves up to 62%...

A Delay Sensitive Error Control Scheme for Continuous Media Communications

Bert Dempsey, Jorg Liebeherr & Alfred Weaver
In this paper we present a new error control approach for continuous media communication protocols. Since continuous media data are inherently bursty, these protocols commonly use a control time for the first packet in a burst to ensure continuous playback in the presence of delay jitter. The retransmission scheme presented in this paper, which is called Slack Automatic Repeat Request (S-ARQ), is based on appropriately extending the control time for the first packet in a...

A Video Mail Distribution System for Networked Personal Computers

F Street & A Weaver
The Computer Networks Laboratory at the University of Virginia has used the Xpress TranSferProtocol (XTP) to construct the Xpress File System (XFS), a local area network peer-to-peer file sharing system for MS-DOS based IBM Personal Computers. XFS is usable for a variety of purposes, most notably the real time delivery of multimedia data streams. Using specialized hardware and software from Fluent Machines Inc., we have shown that XFS is capable of maintaining syffrcient file system...

Pattern Minefield Detection From Inexact Data

Gabe Robins & Brian Robinson

Retransmission-Based Error Control for Continuous Media Traffic in Packet-Switched Networks (PhD dissertation)

Bert Dempsey
Distribution of digital audio and video, continuous media, over packet - switched networks has become increasingly feasible due to technology trends leading to powerful desktop computers and high - speed networks. Unlike reliable data transfers, transmission of continuous media streams is sensitive to network delays and has some tolerance for limited data loss. End - to -end protocols for continuous media traffic are now emerging, and an area of active research is error control in...

Safety Kernel Enforcement of Software Safety Policies

Kevin Wika
Computing systems in which the consequences of failure are very serious are termed safety - critical. Many such systems exist in application areas such as aerospace, defense, transportation, power-generation, and medicine. The software in these systems is typically large and complex, critical to system safety, and difficult to implement and verify. Even when great effort is expended to develop the software, there is no assurance that the soft» ware will operate with the required level...

Exploiting Data-Flow for Fault-Tolerance in a Wide-Area Parallel System

Anh NguyenTuong, Andrew Grimshaw & Mark Hyett
Wide-area parallel processing systems will soon be available to researchers to solve a range of problems. In these systems, it is certain that host failures and other faults will be a common occurrence. Unfortunately, most parallel processing systems have not been designed with fault-tolerance in mind. Mental is a high-performance object-oriented parallel processing system that is based on an extension of the data-ffow model. The functional nature of data-flow enables both parallelism and faulttolerance. In...

Rapid Prototyping of CMP Floorplans: A Technical Report

Gregory Faust, Brett Meyer & Kevin Skadron
The Computer Architecture literature is now replete with papers concerned with the change in architectural direction from ever more complex single cores to single chip multi-core designs. Along with this opportunity come major challenges. Among them is the sheer size of the space of possible designs. The investigation of this space is far from complete. What is needed to aid in this task is an integrated suite of tools that provides support throughout the design...

Selective Notification: Combining Forms of Decoupled Addressing for Internet-Scale Command and Alert Dissemination

Jonathan Hill & John Knight
By using an information survivability control system, the survivability of critical networked information systems can be enhanced using a variety of fault-tolerance mechanisms. Essential to the effective implementation of such mechanisms is communication from the error detection component to the various application nodes in the network. In this paper, we introduce a technique called Selective Notification for the communication of commands and alerts in very large distributed systems. The technique combines intentional addressing, content addressing...

Tutorial on UVA SAFENET Lightweight Communications Architecture

Bert Dempsey, Jeffrey Michel & Alfred Weaver

SAFENET Internals

Bert Dempsey, John Fenton, Jeff Michel, Alex Waterman & Alfred Weaver

Registration Year

  • 2017

Resource Types

  • Report


  • UNSW Sydney
  • University of Virginia
  • Bonn-Rhein-Sieg University of Applied Sciences
  • University of Bonn