The spatial resolution of source data, the impact factor selection on the grid model and the size of the grid might be the main limitations of global land datasets applied on a regional scale. Quantitative studies of ...The spatial resolution of source data, the impact factor selection on the grid model and the size of the grid might be the main limitations of global land datasets applied on a regional scale. Quantitative studies of the impacts of rasterization on data accuracy can help improve data resolution and regional data accuracy. Through a case study of cropland data for Jiangsu and Anhui provinces in China, this research compared data accuracy with different data sources, rasterization methods, and grid sizes. First, we investigated the influence of different data sources on gridded data accuracy. The temporal trends of the History Database of the Global Environment （HYDE）, Chinese Historical Cropland Data （CHCD）, and Suwan Cropland Data （SWCD） datasets were more similar. However, differ- ent spatial resolutions of cropland source data in the CHCD and SWCD datasets revealed an average difference of 16.61% when provin- cial and county data were downscaled to a 10 x 10 km2 grid for comparison. Second, the influence of selection of the potential arable land reclamation rate and temperature factors, as well as the different processing methods for water factors, on accuracy of gridded datasets was investigated. Applying the reclamation rate of potential cropland to grid-processing increased the diversity of spatial distri- bution but resulted in only a slightly greater standard deviation, which increased by 4.05. Temperature factors only produced relative disparities within 10% and absolute disparities within 2 km2 over more than 90% of grid cells. For the different processing methods for water factors, the HYDE dataset distributed 70% more cropland in grid cells along riverbanks, at the abandoned Yellow River Estuary （located in Binhai County, Yancheng City, Jiangsu Province）, and around Hongze Lake, than did the SWCD dataset. Finally, we ex- plored the influence of different grid sizes. Absolute accuracy disparities by unit area for the year 2000 were within 0.1 km2 at a 1 km2 grid size, a 25% improvement over the 10 km2 grid size. Compared to the outcomes of other similar studies, this demonstrates that some model hypotheses and grid-processing methods in international land datasets are truly incongruent with actual land reclamation proc- esses, at least in China. Combining the model-based methods with historical empirical data may be a better way to improve the accuracy of regional scale datasets. Exploring methods for the above aspects improved the accuracy of historical crop/and gridded datasets for finer regional scales.展开更多
The packet generator （pktgen） is a fundamental module of the majority of soft- ware testers used to benchmark network pro- tocols and functions. The high performance of the pktgen is an important feature of Future I...The packet generator （pktgen） is a fundamental module of the majority of soft- ware testers used to benchmark network pro- tocols and functions. The high performance of the pktgen is an important feature of Future Internet Testbeds, and DPDK is a network packet accelerated platform, so we can use DPDK to improve performance. Meanwhile, green computing is advocated for in the fu- ture of the internet. Most existing efforts have contributed to improving either performance or accuracy. We, however, shifted the focus to energy-efficiency. We find that high per- formance comes at the cost of high energy consumption. Therefore, we started from a widely used high performance schema, deeply studying the multi-core platform, especially in terms of parallelism, core allocation, and fre- quency controlling. On this basis, we proposed an AFfinity-oriented Fine-grained CONtrolling （AFFCON） mechanism in order to improve energy efficiency with desirable performance. As clearly demonstrated through a series of evaluative experiments, our proposal can reduce CPU power consumption by up to 11% while maintaining throughput at the line rate.展开更多
基金Under the auspices of National Natural Science Foundation of China(No.41471156,41501207)the Strategic Priority Research Program of the Chinese Academy of Sciences(No.XDA05080102)Special Fund of National Science and Technology of China(No.2014FY130500)
文摘The spatial resolution of source data, the impact factor selection on the grid model and the size of the grid might be the main limitations of global land datasets applied on a regional scale. Quantitative studies of the impacts of rasterization on data accuracy can help improve data resolution and regional data accuracy. Through a case study of cropland data for Jiangsu and Anhui provinces in China, this research compared data accuracy with different data sources, rasterization methods, and grid sizes. First, we investigated the influence of different data sources on gridded data accuracy. The temporal trends of the History Database of the Global Environment （HYDE）, Chinese Historical Cropland Data （CHCD）, and Suwan Cropland Data （SWCD） datasets were more similar. However, differ- ent spatial resolutions of cropland source data in the CHCD and SWCD datasets revealed an average difference of 16.61% when provin- cial and county data were downscaled to a 10 x 10 km2 grid for comparison. Second, the influence of selection of the potential arable land reclamation rate and temperature factors, as well as the different processing methods for water factors, on accuracy of gridded datasets was investigated. Applying the reclamation rate of potential cropland to grid-processing increased the diversity of spatial distri- bution but resulted in only a slightly greater standard deviation, which increased by 4.05. Temperature factors only produced relative disparities within 10% and absolute disparities within 2 km2 over more than 90% of grid cells. For the different processing methods for water factors, the HYDE dataset distributed 70% more cropland in grid cells along riverbanks, at the abandoned Yellow River Estuary （located in Binhai County, Yancheng City, Jiangsu Province）, and around Hongze Lake, than did the SWCD dataset. Finally, we ex- plored the influence of different grid sizes. Absolute accuracy disparities by unit area for the year 2000 were within 0.1 km2 at a 1 km2 grid size, a 25% improvement over the 10 km2 grid size. Compared to the outcomes of other similar studies, this demonstrates that some model hypotheses and grid-processing methods in international land datasets are truly incongruent with actual land reclamation proc- esses, at least in China. Combining the model-based methods with historical empirical data may be a better way to improve the accuracy of regional scale datasets. Exploring methods for the above aspects improved the accuracy of historical crop/and gridded datasets for finer regional scales.
基金supported by the National Science Foundation of China (No. 61472130, Research on Graphic Processing Units-based High-performance Packet Processing)the China Postdoctoral Science Foundation funded project (No. 61702174)
文摘The packet generator （pktgen） is a fundamental module of the majority of soft- ware testers used to benchmark network pro- tocols and functions. The high performance of the pktgen is an important feature of Future Internet Testbeds, and DPDK is a network packet accelerated platform, so we can use DPDK to improve performance. Meanwhile, green computing is advocated for in the fu- ture of the internet. Most existing efforts have contributed to improving either performance or accuracy. We, however, shifted the focus to energy-efficiency. We find that high per- formance comes at the cost of high energy consumption. Therefore, we started from a widely used high performance schema, deeply studying the multi-core platform, especially in terms of parallelism, core allocation, and fre- quency controlling. On this basis, we proposed an AFfinity-oriented Fine-grained CONtrolling （AFFCON） mechanism in order to improve energy efficiency with desirable performance. As clearly demonstrated through a series of evaluative experiments, our proposal can reduce CPU power consumption by up to 11% while maintaining throughput at the line rate.
