Research Terms
This power management tool makes data centers more environmentally friendly by ensuring that available power and computing loads are in alignment. Data centers are temperature-controlled facilities that house computer systems and components, including servers, power supplies, backup power equipment, chillers, cables, fire and water detection systems, and security controls. All major companies, universities, banks and government institutions have data centers that store information and maintain their computer networks. A single large data center's energy usage can exceed 30 megawatts (i.e. 30 million joules per second), comparable to the energy usage of a small town. It is estimated that carbon dioxide emissions from computing systems will reach 1.54 metric gigatons by 2020, which would make information technology companies the largest contributors to greenhouse emissions. Available power management schemes adapt computer loads to the time-varying power budget, causing slow job turnaround time and poor service availability. Researchers at the University of Florida have developed a power-management tool that employs a power demand shaping (PDS) technique using load following (online power generation for tracking changes in customer loads) to meet time-variable power demands in distributed generation (DG) systems.
A power management tool that controls available power and power consumption in distributed generation (DG) systems, restricting power or providing burst of energy as needed for energy-efficient data centers
This power management tool, developed by UF researchers, employs load following and distributed generation (DG) to achieve power demand shaping (PDS), which helps meet power demands that vary over time. Load following refers to the use of online power generation equipment to track changes in customer's needs for computing power. By leveraging the load following capability of onsite renewable generation, this device lowers carbon emissions while maintaining sufficient performance. The device can be implemented as a cross-layer power management module between a front-end distributed generation system and a back-end computing facility to provide coordinated tuning between supply and load. The technology has applications in DG-powered data centers, DG-powered cloud computing infrastructures and other green/sustainable facilities.
This data structure system measures network traffic and stores network contact information in a small memory space. Every day, businesses and individuals are bombarded with everything from spam e-mail to corporate hackers. As networks have gotten faster and information transfer has greatly increased, measuring network traffic has become progressively more important for allocating network resources and ensuring security. In order to measure and track what is entering and leaving your network, it is necessary to contain vast amounts of information in a compact memory space. Unfortunately, today's traffic far exceeds the capabilities of any system currently available. Researchers at the University of Florida have developed a new spread estimator that delivers excellent performance in a tight memory space where all existing estimators no longer work. Not only does it achieve space compactness, but it also operates more efficiently than existing systems.
Network traffic management systems and network security systems
A spread estimator is a software/hardware module on a router that inspects the arrival packets and estimates the spread of each source. The spread is defined as the number of distinct internal hosts that an external host (called a source) has contacted during a measurement period. It has important applications in detecting port scans and distributed denial of service attacks, measuring the infection rate of a worm, assisting resource allocation in a server farm, determining popular web contents for caching, and much more. The main technical challenge is to fit a spread estimator in a fast but small and expensive cache memory in order to operate it at the line speed in a high-speed network. In this invention, researchers at the University of Florida have designed a new spread estimator that delivers good performance in tight memory space where all existing estimators no longer work. The estimator effectively achieves space compactness and also operates more efficiently than anything available on the market today. Its accuracy and efficiency come from a new structure for data storage, called virtual vectors, which allow for the measurement and removal of errors in spread estimation.