The Experts below are selected from a list of 34899 Experts worldwide ranked by ideXlab platform
Andrés J. García - One of the best experts on this subject based on the ideXlab platform.
-
A centrifugation cell adhesion assay for high-throughput screening of biomaterial surfaces.
Journal of biomedical materials research. Part A, 2003Co-Authors: Catherine D. Reyes, Andrés J. GarcíaAbstract:A quantitative analysis of cell adhesion is essential in understanding physiological phenomena and designing biomaterials, implant surfaces, and tissue-engineering scaffolds. The most common cell adhesion assays used to evaluate biomaterial surfaces lack sensitivity and reproducibility and/or require specialized equipment and skill-Intensive Operation. We describe a modified centrifugation cell adhesion assay that uses simple and convenient techniques with standard laboratory equipment and provides reliable, quantitative measurements of cell adhesion. This centrifugation assay applies controlled and uniform detachment forces to a large population of adherent cells, providing robust statistics for quantifying cell adhesion. The applicability of this system to the design and characterization of biomaterial surfaces is shown by evaluating cell adhesion on substrates using different coating proteins, cell types, seeding times, and relative centrifugal forces (RCF). Results verify that this centrifugation cell adhesion assay represents a simple, convenient, and standard method for high-throughput characterization of a variety of biomaterial surfaces and conditions.
-
a centrifugation cell adhesion assay for high throughput screening of biomaterial surfaces
Journal of Biomedical Materials Research Part A, 2003Co-Authors: Catherine D. Reyes, Andrés J. GarcíaAbstract:A quantitative analysis of cell adhesion is essential in understanding physiological phenomena and designing biomaterials, implant surfaces, and tissue-engineering scaffolds. The most common cell adhesion assays used to evaluate biomaterial surfaces lack sensitivity and reproducibility and/or require specialized equipment and skill-Intensive Operation. We describe a modified centrifugation cell adhesion assay that uses simple and convenient techniques with standard laboratory equipment and provides reliable, quantitative measurements of cell adhesion. This centrifugation assay applies controlled and uniform detachment forces to a large population of adherent cells, providing robust statistics for quantifying cell adhesion. The applicability of this system to the design and characterization of biomaterial surfaces is shown by evaluating cell adhesion on substrates using different coating proteins, cell types, seeding times, and relative centrifugal forces (RCF). Results verify that this centrifugation cell adhesion assay represents a simple, convenient, and standard method for high-throughput characterization of a variety of biomaterial surfaces and conditions. © 2003 Wiley Periodicals, Inc. J Biomed Mater Res 67A: 328–333, 2003
Catherine D. Reyes - One of the best experts on this subject based on the ideXlab platform.
-
A centrifugation cell adhesion assay for high-throughput screening of biomaterial surfaces.
Journal of biomedical materials research. Part A, 2003Co-Authors: Catherine D. Reyes, Andrés J. GarcíaAbstract:A quantitative analysis of cell adhesion is essential in understanding physiological phenomena and designing biomaterials, implant surfaces, and tissue-engineering scaffolds. The most common cell adhesion assays used to evaluate biomaterial surfaces lack sensitivity and reproducibility and/or require specialized equipment and skill-Intensive Operation. We describe a modified centrifugation cell adhesion assay that uses simple and convenient techniques with standard laboratory equipment and provides reliable, quantitative measurements of cell adhesion. This centrifugation assay applies controlled and uniform detachment forces to a large population of adherent cells, providing robust statistics for quantifying cell adhesion. The applicability of this system to the design and characterization of biomaterial surfaces is shown by evaluating cell adhesion on substrates using different coating proteins, cell types, seeding times, and relative centrifugal forces (RCF). Results verify that this centrifugation cell adhesion assay represents a simple, convenient, and standard method for high-throughput characterization of a variety of biomaterial surfaces and conditions.
-
a centrifugation cell adhesion assay for high throughput screening of biomaterial surfaces
Journal of Biomedical Materials Research Part A, 2003Co-Authors: Catherine D. Reyes, Andrés J. GarcíaAbstract:A quantitative analysis of cell adhesion is essential in understanding physiological phenomena and designing biomaterials, implant surfaces, and tissue-engineering scaffolds. The most common cell adhesion assays used to evaluate biomaterial surfaces lack sensitivity and reproducibility and/or require specialized equipment and skill-Intensive Operation. We describe a modified centrifugation cell adhesion assay that uses simple and convenient techniques with standard laboratory equipment and provides reliable, quantitative measurements of cell adhesion. This centrifugation assay applies controlled and uniform detachment forces to a large population of adherent cells, providing robust statistics for quantifying cell adhesion. The applicability of this system to the design and characterization of biomaterial surfaces is shown by evaluating cell adhesion on substrates using different coating proteins, cell types, seeding times, and relative centrifugal forces (RCF). Results verify that this centrifugation cell adhesion assay represents a simple, convenient, and standard method for high-throughput characterization of a variety of biomaterial surfaces and conditions. © 2003 Wiley Periodicals, Inc. J Biomed Mater Res 67A: 328–333, 2003
Silva, Tércio Cassiano - One of the best experts on this subject based on the ideXlab platform.
-
Estudo de viabilidade de sistemas de detecção e intrusão em redes baseados em GPUs
2013Co-Authors: Ribeirinho, Lucas Polonio, Silva, Tércio CassianoAbstract:Sistemas de Detecção e Prevenção de Intrusão em Redes (IDPS) são importantes ferramentas para a análise do tráfego que circula por redes de computadores. O Snort é um dos IDPS mais utilizados mundialmente e considerado uma das melhores ferramentas disponíveis. A performance é um dos pontos principais de um IDPS, e no caso do Snort, a operação de maior custo computacional é a comparação dos pacotes da rede com milhares de patterns presentes em suas regras, existentes para detectar padrões conhecidos de ataques a computadores. Visto que cada pacote é comparado a milhares de patterns, uma análise cuidadosa mostrou que este problema poderia ser adaptado para execução em GPUs, em grande parte por sua arquitetura SIMD (Single Instruction Multiple Data). Para minimizar o overhead de transferência dos pacotes da CPU para a GPU, foi necessário acumular pacotes e enviar um buffer de pacotes para a GPU. O Snort, porém, é uma plataforma que não foi pensado para execução paralela. Para resolver estes problemas, e com a premissa da não modificação da arquitetura interna do Snort, foi criada uma arquitetura para execução de vários processos de Snort simultâneos, acúmulo de pacotes em um buffer e uso de GPU para o pattern matching. Os resultados obtidos com esta arquitetura não foram melhores do que a execução de um Snort não-modificado. Por outro lado, sugere-se ser possível excelentes ganhos de performance, de mais de 433%, caso a arquitetura do Snort seja modifcada ou em outro IDPS que seja favorável ao acúmulo de um buffer de pacotes para matching. _________________________________________________________________________ ABSTRACTIntrusion Prevention and Detection Systems (IDPS) are very important tools for the analysis of network traffic across computers. Snort is one of the most famous and useful IDPS worldwide and is considered to be one of the best ones available. Performance is the key factor of an IDPS and, especially in Snort, the most computationally Intensive Operation is the matching of all the network packets against thousands of patterns present in its rules, responsible for detecting known patterns of malicious traffic. Based on the fact that every packet is compared to thousands of patterns, it was realized that this problem could be adapted for GPU execution, due to GPU's SIMD (Single Instruction Multiple Data) architecture. To minimize the overhead caused by transferring packets from CPU's to GPU's memory, it was needed to store packets and send whole buffers to GPU in only one transfer. However, Snort's arquitecture was not planned for parallel execution, so that it could acquire this buffer. To solve these problems, and considering we would not alter Snort's core architecture, it was created an architecture for executing multiple Snort processes simultaneously, for capturing packets ready for matching in a buffer and, finally, for using the GPU to perform the pattern matching. The results, achieved using this architecture, were not better than the standard Snort. Still, it was showed there could be a possibility of excellent performance boosts (over 433%) if Snort's core architecture was modified or whether it was used another IDPS whose architecture suits better the creation of the packet buffer for matching
-
Estudo de viabilidade de sistemas de detecção e intrusão em redes baseados em GPUs
2012Co-Authors: Ribeirinho, Lucas Polonio, Silva, Tércio CassianoAbstract:Monografia (graduação)—Universidade de Brasília, Instituto de Ciências Exatas, Departamento de Ciência da Computação, 2012.Sistemas de Detecção e Prevenção de Intrusão em Redes (IDPS) são importantes ferramentas para a análise do tráfego que circula por redes de computadores. O Snort é um dos IDPS mais utilizados mundialmente e considerado uma das melhores ferramentas disponíveis. A performance é um dos pontos principais de um IDPS, e no caso do Snort, a operação de maior custo computacional é a comparação dos pacotes da rede com milhares de patterns presentes em suas regras, existentes para detectar padrões conhecidos de ataques a computadores. Visto que cada pacote é comparado a milhares de patterns, uma análise cuidadosa mostrou que este problema poderia ser adaptado para execução em GPUs, em grande parte por sua arquitetura SIMD (Single Instruction Multiple Data). Para minimizar o overhead de transferência dos pacotes da CPU para a GPU, foi necessário acumular pacotes e enviar um buffer de pacotes para a GPU. O Snort, porém, é uma plataforma que não foi pensado para execução paralela. Para resolver estes problemas, e com a premissa da não modificação da arquitetura interna do Snort, foi criada uma arquitetura para execução de vários processos de Snort simultâneos, acúmulo de pacotes em um buffer e uso de GPU para o pattern matching. Os resultados obtidos com esta arquitetura não foram melhores do que a execução de um Snort não-modificado. Por outro lado, sugere-se ser possível excelentes ganhos de performance, de mais de 433%, caso a arquitetura do Snort seja modifcada ou em outro IDPS que seja favorável ao acúmulo de um buffer de pacotes para matching. _________________________________________________________________________ ABSTRACTIntrusion Prevention and Detection Systems (IDPS) are very important tools for the analysis of network traffic across computers. Snort is one of the most famous and useful IDPS worldwide and is considered to be one of the best ones available. Performance is the key factor of an IDPS and, especially in Snort, the most computationally Intensive Operation is the matching of all the network packets against thousands of patterns present in its rules, responsible for detecting known patterns of malicious traffic. Based on the fact that every packet is compared to thousands of patterns, it was realized that this problem could be adapted for GPU execution, due to GPU's SIMD (Single Instruction Multiple Data) architecture. To minimize the overhead caused by transferring packets from CPU's to GPU's memory, it was needed to store packets and send whole buffers to GPU in only one transfer. However, Snort's arquitecture was not planned for parallel execution, so that it could acquire this buffer. To solve these problems, and considering we would not alter Snort's core architecture, it was created an architecture for executing multiple Snort processes simultaneously, for capturing packets ready for matching in a buffer and, finally, for using the GPU to perform the pattern matching. The results, achieved using this architecture, were not better than the standard Snort. Still, it was showed there could be a possibility of excellent performance boosts (over 433%) if Snort's core architecture was modified or whether it was used another IDPS whose architecture suits better the creation of the packet buffer for matching
Reinhard Madlener - One of the best experts on this subject based on the ideXlab platform.
-
switching from fossil fuel to renewables in residential heating systems an empirical study of homeowners decisions in germany
Energy Policy, 2016Co-Authors: Carl Christian Michelsen, Reinhard MadlenerAbstract:Abstract The replacement of outdated and inefficient fossil fuel residential heating systems (RHS) by more efficient and less CO 2 -Intensive appliances primarily based on renewable energy sources is an important pillar for the transition to a cleaner and more sustainable energy system. This paper empirically investigates drivers and barriers behind homeowners' decisions to switch from a fossil fuel to a renewable RHS in Germany. For this purpose, we draw on data from a 2010 questionnaire survey among owners of existing single-family and duplex houses in Germany that had received a financial grant to install an RHS (i.e. condensing boiler with solar thermal support, heat pump or wood pellet boiler). We show that environmental protection, a lower dependency on fossil fuels, and a higher degree of RHS-related knowledge are key drivers. In contrast, the perceived difficulty of getting used to the system and a misunderstanding of its principal functioning are obstacles for the heat pump. For the wood pellet boiler, perceived barriers include the low usability, the labor-Intensive Operation, and the systems' fault liability. Hence, a higher replacement rate requires the willingness to relinquish old habits and perceptions of how an RHS works and operates.
-
switching from fossil fuel to renewables in residential heating systems an empirical study of homeowners decisions in germany
2013Co-Authors: Carl Christian Michelsen, Reinhard MadlenerAbstract:This paper investigates key drivers and barriers behind homeowners’ decisions to switch from a fossil fuel (i.e. oil or gas boiler) to a renewable residential heating system (RHS). For this purpose, we draw on data from a 2010 questionnaire survey among owners of existing 1-family or 2-family homes in Germany that had received a financial grant for installing an (at least partly) renewable RHS (i.e. heat pump or wood pellet boiler). We analyze the data by means of logistic regression techniques. First, our results show that the motivation to deal with external threats (i.e. environmental protection and reduction of dependence on fossil fuels) and a higher degree of RHS-related knowledge are key drivers for switching to a renewable RHS. Second, we identify the different barriers that prevent homeowners from adopting a specific RHS. In particular, for the adoption of fossil fuel RHS, the perceived strong reliance on pricey oil or natural gas is found to be a major hurdle. For the heat pump, we find that the perceived difficulty of getting used to the system and a misunderstanding of its principal functioning are important obstacles. Finally, for the wood pellet boiler, our results imply that non-adopters perceive the low usability, the labor-Intensive Operation, and the systems’ high fault liability to be important barriers. We conclude that homeowners often fear major changes to their current status quo (e.g. replacement of the existing heating system infrastructure) and, thus, tend to opt for minor and thus quick adjustments to their RHS (e.g. replacing only the boiler). Likewise, a higher replacement rate of fossil fuel by renewable RHS requires the homeowners’ willingness to relinquish old habits and perceptions of how an RHS works and operates.
Ribeirinho, Lucas Polonio - One of the best experts on this subject based on the ideXlab platform.
-
Estudo de viabilidade de sistemas de detecção e intrusão em redes baseados em GPUs
2013Co-Authors: Ribeirinho, Lucas Polonio, Silva, Tércio CassianoAbstract:Sistemas de Detecção e Prevenção de Intrusão em Redes (IDPS) são importantes ferramentas para a análise do tráfego que circula por redes de computadores. O Snort é um dos IDPS mais utilizados mundialmente e considerado uma das melhores ferramentas disponíveis. A performance é um dos pontos principais de um IDPS, e no caso do Snort, a operação de maior custo computacional é a comparação dos pacotes da rede com milhares de patterns presentes em suas regras, existentes para detectar padrões conhecidos de ataques a computadores. Visto que cada pacote é comparado a milhares de patterns, uma análise cuidadosa mostrou que este problema poderia ser adaptado para execução em GPUs, em grande parte por sua arquitetura SIMD (Single Instruction Multiple Data). Para minimizar o overhead de transferência dos pacotes da CPU para a GPU, foi necessário acumular pacotes e enviar um buffer de pacotes para a GPU. O Snort, porém, é uma plataforma que não foi pensado para execução paralela. Para resolver estes problemas, e com a premissa da não modificação da arquitetura interna do Snort, foi criada uma arquitetura para execução de vários processos de Snort simultâneos, acúmulo de pacotes em um buffer e uso de GPU para o pattern matching. Os resultados obtidos com esta arquitetura não foram melhores do que a execução de um Snort não-modificado. Por outro lado, sugere-se ser possível excelentes ganhos de performance, de mais de 433%, caso a arquitetura do Snort seja modifcada ou em outro IDPS que seja favorável ao acúmulo de um buffer de pacotes para matching. _________________________________________________________________________ ABSTRACTIntrusion Prevention and Detection Systems (IDPS) are very important tools for the analysis of network traffic across computers. Snort is one of the most famous and useful IDPS worldwide and is considered to be one of the best ones available. Performance is the key factor of an IDPS and, especially in Snort, the most computationally Intensive Operation is the matching of all the network packets against thousands of patterns present in its rules, responsible for detecting known patterns of malicious traffic. Based on the fact that every packet is compared to thousands of patterns, it was realized that this problem could be adapted for GPU execution, due to GPU's SIMD (Single Instruction Multiple Data) architecture. To minimize the overhead caused by transferring packets from CPU's to GPU's memory, it was needed to store packets and send whole buffers to GPU in only one transfer. However, Snort's arquitecture was not planned for parallel execution, so that it could acquire this buffer. To solve these problems, and considering we would not alter Snort's core architecture, it was created an architecture for executing multiple Snort processes simultaneously, for capturing packets ready for matching in a buffer and, finally, for using the GPU to perform the pattern matching. The results, achieved using this architecture, were not better than the standard Snort. Still, it was showed there could be a possibility of excellent performance boosts (over 433%) if Snort's core architecture was modified or whether it was used another IDPS whose architecture suits better the creation of the packet buffer for matching
-
Estudo de viabilidade de sistemas de detecção e intrusão em redes baseados em GPUs
2012Co-Authors: Ribeirinho, Lucas Polonio, Silva, Tércio CassianoAbstract:Monografia (graduação)—Universidade de Brasília, Instituto de Ciências Exatas, Departamento de Ciência da Computação, 2012.Sistemas de Detecção e Prevenção de Intrusão em Redes (IDPS) são importantes ferramentas para a análise do tráfego que circula por redes de computadores. O Snort é um dos IDPS mais utilizados mundialmente e considerado uma das melhores ferramentas disponíveis. A performance é um dos pontos principais de um IDPS, e no caso do Snort, a operação de maior custo computacional é a comparação dos pacotes da rede com milhares de patterns presentes em suas regras, existentes para detectar padrões conhecidos de ataques a computadores. Visto que cada pacote é comparado a milhares de patterns, uma análise cuidadosa mostrou que este problema poderia ser adaptado para execução em GPUs, em grande parte por sua arquitetura SIMD (Single Instruction Multiple Data). Para minimizar o overhead de transferência dos pacotes da CPU para a GPU, foi necessário acumular pacotes e enviar um buffer de pacotes para a GPU. O Snort, porém, é uma plataforma que não foi pensado para execução paralela. Para resolver estes problemas, e com a premissa da não modificação da arquitetura interna do Snort, foi criada uma arquitetura para execução de vários processos de Snort simultâneos, acúmulo de pacotes em um buffer e uso de GPU para o pattern matching. Os resultados obtidos com esta arquitetura não foram melhores do que a execução de um Snort não-modificado. Por outro lado, sugere-se ser possível excelentes ganhos de performance, de mais de 433%, caso a arquitetura do Snort seja modifcada ou em outro IDPS que seja favorável ao acúmulo de um buffer de pacotes para matching. _________________________________________________________________________ ABSTRACTIntrusion Prevention and Detection Systems (IDPS) are very important tools for the analysis of network traffic across computers. Snort is one of the most famous and useful IDPS worldwide and is considered to be one of the best ones available. Performance is the key factor of an IDPS and, especially in Snort, the most computationally Intensive Operation is the matching of all the network packets against thousands of patterns present in its rules, responsible for detecting known patterns of malicious traffic. Based on the fact that every packet is compared to thousands of patterns, it was realized that this problem could be adapted for GPU execution, due to GPU's SIMD (Single Instruction Multiple Data) architecture. To minimize the overhead caused by transferring packets from CPU's to GPU's memory, it was needed to store packets and send whole buffers to GPU in only one transfer. However, Snort's arquitecture was not planned for parallel execution, so that it could acquire this buffer. To solve these problems, and considering we would not alter Snort's core architecture, it was created an architecture for executing multiple Snort processes simultaneously, for capturing packets ready for matching in a buffer and, finally, for using the GPU to perform the pattern matching. The results, achieved using this architecture, were not better than the standard Snort. Still, it was showed there could be a possibility of excellent performance boosts (over 433%) if Snort's core architecture was modified or whether it was used another IDPS whose architecture suits better the creation of the packet buffer for matching