Heterogeneous computing is one effective method of high performance computing with many advantages. Task scheduling is a critical issue in heterogeneous environments as well as in homogeneous environments. A number of...Heterogeneous computing is one effective method of high performance computing with many advantages. Task scheduling is a critical issue in heterogeneous environments as well as in homogeneous environments. A number of task scheduling algorithms for homogeneous environments have been proposed, whereas, a few for heterogeneous environments can be found in the literature. A novel task scheduling algorithm for heterogeneous environments, called the heterogeneous critical task (HCT) scheduling algorithm is presented. By means of the directed acyclic graph and the gantt graph, the HCT algorithm defines the critical task and the idle time slot. After determining the critical tasks of a given task, the HCT algorithm tentatively duplicates the critical tasks onto the processor that has the given task in the idle time slot, to reduce the start time of the given task. To compare the performance of the HCT algorithm with several recently proposed algorithms, a large set of randomly generated applications and the Gaussian elimination application are randomly generated. The experimental result has shown that the HCT algorithm outperforms the other algorithm.展开更多
Agricultural science and technology parks(ASTPs) represent an important growth pole in China’s agricultural modernization.Clarifying their diffusion laws can optimize the technological diffusion process and improve i...Agricultural science and technology parks(ASTPs) represent an important growth pole in China’s agricultural modernization.Clarifying their diffusion laws can optimize the technological diffusion process and improve its efficiency. Our study uses disaggregated spatial information in its model to analyze ASTP technology diffusion in a heterogeneous space. We constructed a comprehensive index system to evaluate the diffusion environmental quality and introduced the heterogeneous diffusion equation to calculate the technological diffusion probability. We applied this framework to a real-world scenario: the apple planting technology diffusion of the Yangling ASTP in the Loess Plateau, China. The results indicated: (1)the technological diffusion environment of the Loess Plateau advantageous apple producing area showed strong spatial heterogeneity caused by climate, topography, and external transportation links.(2) Under the combined effects of distance and spatial heterogeneity, the spatial diffusion pattern of the Yangling ASTP apple technology was expansion diffusion supplemented by hierarchical diffusion and banded diffusion, and (3) ASTP technology diffusion showed a strong distance attenuation effect, and the frictional effect of distance can be decreased by improving the diffusion environmental quality. These laws can promote regional balanced ASTP-driven development.展开更多
Federated learning(FL)is a novel technique in deep learning that enables clients to collaboratively train a shared model while retaining their decentralized data.However,researchers working on FL face several unique c...Federated learning(FL)is a novel technique in deep learning that enables clients to collaboratively train a shared model while retaining their decentralized data.However,researchers working on FL face several unique challenges,especially in the context of heterogeneity.Heterogeneity in data distributions,computational capabilities,and scenarios among clients necessitates the development of customized models and objectives in FL.Unfortunately,existing works such as FedAvg may not effectively accommodate the specific needs of each client.To address the challenges arising from heterogeneity in FL,we provide an overview of the heterogeneities in data,model,and objective(DMO).Furthermore,we propose a novel framework called federated mutual learning(FML),which enables each client to train a personalized model that accounts for the data heterogeneity(DH).A“meme model”serves as an intermediary between the personalized and global models to address model heterogeneity(MH).We introduce a knowledge distillation technique called deep mutual learning(DML)to transfer knowledge between these two models on local data.To overcome objective heterogeneity(OH),we design a shared global model that includes only certain parts,and the personalized model is task-specific and enhanced through mutual learning with the meme model.We evaluate the performance of FML in addressing DMO heterogeneities through experiments and compare it with other commonly used FL methods in similar scenarios.The results demonstrate that FML outperforms other methods and effectively addresses the DMO challenges encountered in the FL setting.展开更多
Nowadays,application migration becomes more and more attractive.For example,it can make computation closer to data sources or make service closer to end-users,which may significantly decrease latency in edge computing...Nowadays,application migration becomes more and more attractive.For example,it can make computation closer to data sources or make service closer to end-users,which may significantly decrease latency in edge computing.Yet,migrating applications among servers that are controlled by different platform owners raises security issues.We leverage hardware-secured trusted execution environment(TEE,aka.,enclave)technologies,such as Intel SGX,AMD SEV,and ARM TrustZone,for protecting critical computations on untrusted servers.However,these hardware TEEs propose non-uniform programming abstractions and are based on heterogeneous architectures,which not only forces programmers to develop secure applications targeting some specific abstraction but also hinders the migration of protected applications.Therefore,we propose UniTEE which gives a unified enclave programming abstraction across the above three hardware TEEs by using a microkernel-based design and enables the secure enclave migration by integrating heterogeneous migration techniques.We have implemented the prototype on real machines.The evaluation results show the migration support incurs nearly-zero runtime overhead and the migration procedure is also efficient.展开更多
Aims Foraging behavior was concerned less in plants especially for clonal aquatic species.The goal of this study was to test the foraging behav-ior and the function of phenotypic plasticity of a clonal submerged plant...Aims Foraging behavior was concerned less in plants especially for clonal aquatic species.The goal of this study was to test the foraging behav-ior and the function of phenotypic plasticity of a clonal submerged plant,Potamogeton maackianus A.Benn in variable manipulative environments.Methods To explore the foraging behavior of a clonal submerged plant,clones of P.maackianus were cultivated in variable environments of different substrate types and light conditions.We measured bio-mass,specific root area(SRA),specific root length(SRL),branch-ing angle of the ramet,chlorophyll content of apical leaves(CCAL),Fv/Fm(chlorophyll fluorescence parameter that indicates maximal photosystem II(PS II)efficiency)and biomass distribution in soil and light contrast treatments after harvest.Important Findings We found that the atypical non-clonal-like foraging behavior homogenized the performance of the stoloniferous clonal spe-cies,P.maackianus,in the measurement of biomass in our study.Substrate type had significant effects on most traits of P.maackianus,except on total biomass,shoot biomass distribution and CCAL.Patchy substrate mediated more root distribution into sand patch rather than into clay patch.Increase of SRA,SRL and Fv/Fm might be adaptive to substrate heterogeneity.Moreover,substrate hetero-geneity induced‘sidewalk’stolon lineage on the edge of sand patch along the patch border.Light condition only had significant effects on branching angle,shoot biomass distribution and Fv/Fm.Shade led to more shoot assignment in light-rich zone than in light-poor zone.Potamogeton maackianus adjusted the branching angle of ramet rather than clonal exploitation in heterogeneous substrates with patchy shade.Moreover,shade led to higher Fv/Fm.A com-promising foraging strategy promoted P.maackianus to escape unfa-vorable environments.Phenotypic plasticity,which maximized the performance of P.maackianus in the measurement of biomass in variable environments,may explain its domination in some shallow lakes of the middle and lower reaches of Yangtze River.展开更多
To reduce the running time of network simulation in heterogeneous computing environment,a network simulation task partition method,named LBPHCE,is put forward.In this method,the network simulation task is partitioned ...To reduce the running time of network simulation in heterogeneous computing environment,a network simulation task partition method,named LBPHCE,is put forward.In this method,the network simulation task is partitioned in comprehensive consideration of the load balance of both routing computing simulation and packet forwarding simulation.First,through benchmark experiments,the computation ability and routing simulation ability of each simulation machine are measured in the heterogeneous computing environment.Second,based on the computation ability of each simulation machine,the network simulation task is initially partitioned to meet the load balance of packet forwarding simulation in the heterogeneous computing environment,and then according to the routing computation ability,the scale of each partition is fine-tuned to satisfy the balance of the routing computing simulation,meanwhile the load balance of packet forwarding simulation is guaranteed.Experiments based on PDNS indicate that,compared to traditional uniform partition method,the LBPHCE method can reduce the total simulation running time by 26.3%in average,and compared to the liner partition method,it can reduce the running time by 18.3%in average.展开更多
In this paper,we investigate a reaction-diffusion-advection model with expanding fronts,which models the spatial transmission of West Nile virus(WNv)in a heterogeneous environment.A free boundary problem is formulated...In this paper,we investigate a reaction-diffusion-advection model with expanding fronts,which models the spatial transmission of West Nile virus(WNv)in a heterogeneous environment.A free boundary problem is formulated and the global existence and uniqueness of the solution is presented.In addition to a classical basic reproduction number,the spatial-temporal basic reproduction number for the model with null Dirichlet boundary condition is introduced and the risk index associated with the virus in spatial setting is defined,and their properties are discussed.Sufficient conditions for the WNv to vanish or spread are given,and the asymptotic behavior of the solution to the free boundary problem when the spreading occurs is established.Our results show that the initial number of infected populations and the expanding capability of the expanding fronts exhibit important impacts on the extinction or persistence of the virus.展开更多
The unidirectional impacts of soil conservation on dryland farming and local livelihood have been well studied.However,there are gaps in literature regarding the interactions among factors,such as labor force and fina...The unidirectional impacts of soil conservation on dryland farming and local livelihood have been well studied.However,there are gaps in literature regarding the interactions among factors,such as labor force and financial capital,in response to soil conservation,as well as the influence of such interactions on the relationships between environmental projects and agricultural development.By combining the Chinese Soil Loss Equation(CSLE)model and statistical analyses,this study investigated the spatiotemporal characteristics of soil erosion while also identifying the controlling factors that affect crop and livestock farming.The aim was to reveal how these factors interacted to influence agriculture in two case study areas between 2000 and 2017.The results showed that soil erosion substantially declined as a consequence of soil conservation,while the output from crop farming increased,indicating that there was a synergy between soil conservation and dryland farming in both the districts,over time;and factors facilitating crop farming,such as labor force,agricultural machinery,and irrigation,appeared to exert an adverse effect on livestock farming in both areas,implying a tradeoff between livestock and crop farming in the context of programs such as the Grain for Green Program(GGP)in the Loess Hills.Such a tradeoff is a result of the competition for cropland and labor that exists due to grazing prohibition and rural-urban migration.Therefore,policy-related resolutions are required to address the land use conflicts and migration-related rural labor losing due to implementation of the GGP in the Loess Hills.展开更多
This paper presents the design and implementation of a message-based distributed operating system kernel NDOS. The main purpose of the kernel is to support a distributed data processing system and a distributed DBMS. ...This paper presents the design and implementation of a message-based distributed operating system kernel NDOS. The main purpose of the kernel is to support a distributed data processing system and a distributed DBMS. It uses the abstraction of communication between processes as basic mechanism. In NDOS, services and facilities such as message passing and process synchronization, which are related to IPC and may cause the change of the state of a process, are integrated into a single concept, an event. The initial version of NDOS ker- nel has been implemented on a full heterogeneous environment of different machines, LANs, and OSs with the original high-layered systems and applications are still provided.展开更多
In this paper,a reaction-diffusion SEI epidemic model with nonlinear incidence rate is proposed.The well-posedness of solutions is studied,including the existence of positive and unique classical solution and the exis...In this paper,a reaction-diffusion SEI epidemic model with nonlinear incidence rate is proposed.The well-posedness of solutions is studied,including the existence of positive and unique classical solution and the existence and the ultimate boundedness of global solutions.The basic reproduction numbers are given in both heterogeneous and homogeneous environments.For spatially heterogeneous environment,by the comparison principle of the diffusion system,the infection-free steady state is proved to be globally asymptotically stable if R_(0)<1,if R_(0)>1,the system will be persistent and admit at least one positive steady state.For spatially homogenous environment,by constructing a Lyapunov function,the infect ion-free steady state is proved to be globally asymptotically stable if,R_(0)<1,and then the unique positive steady state is achieved and is proved to be globally asymptotically stable if R_(0)>1.Finally,two examples are given via numerical simulations,and then some control strategies are also presented by the sensitive analysis.展开更多
文摘Heterogeneous computing is one effective method of high performance computing with many advantages. Task scheduling is a critical issue in heterogeneous environments as well as in homogeneous environments. A number of task scheduling algorithms for homogeneous environments have been proposed, whereas, a few for heterogeneous environments can be found in the literature. A novel task scheduling algorithm for heterogeneous environments, called the heterogeneous critical task (HCT) scheduling algorithm is presented. By means of the directed acyclic graph and the gantt graph, the HCT algorithm defines the critical task and the idle time slot. After determining the critical tasks of a given task, the HCT algorithm tentatively duplicates the critical tasks onto the processor that has the given task in the idle time slot, to reduce the start time of the given task. To compare the performance of the HCT algorithm with several recently proposed algorithms, a large set of randomly generated applications and the Gaussian elimination application are randomly generated. The experimental result has shown that the HCT algorithm outperforms the other algorithm.
基金Under the auspices of the National Natural Science Foundation of China(No.41771129)Social Science Foundation of Shaanxi(No.2015D055)Social Science Research Project on Major Theoretical and Practical Issues of Shaanxi(No.2020Z026)。
文摘Agricultural science and technology parks(ASTPs) represent an important growth pole in China’s agricultural modernization.Clarifying their diffusion laws can optimize the technological diffusion process and improve its efficiency. Our study uses disaggregated spatial information in its model to analyze ASTP technology diffusion in a heterogeneous space. We constructed a comprehensive index system to evaluate the diffusion environmental quality and introduced the heterogeneous diffusion equation to calculate the technological diffusion probability. We applied this framework to a real-world scenario: the apple planting technology diffusion of the Yangling ASTP in the Loess Plateau, China. The results indicated: (1)the technological diffusion environment of the Loess Plateau advantageous apple producing area showed strong spatial heterogeneity caused by climate, topography, and external transportation links.(2) Under the combined effects of distance and spatial heterogeneity, the spatial diffusion pattern of the Yangling ASTP apple technology was expansion diffusion supplemented by hierarchical diffusion and banded diffusion, and (3) ASTP technology diffusion showed a strong distance attenuation effect, and the frictional effect of distance can be decreased by improving the diffusion environmental quality. These laws can promote regional balanced ASTP-driven development.
基金supported by the National Natural Science Foundation of China(Nos.U20A20387,62006207,and 62037001)the Young Elite Scientists Sponsorship Program by China Association for Science and Technology(No.2021QNRC001)+4 种基金the Zhejiang Provincial Natural Science Foundation,China(No.LQ21F020020)the Project by Shanghai AI Laboratory,China(No.P22KS00111)the Program of Zhejiang Province Science and Technology(No.2022C01044)the StarryNight Science Fund of Zhejiang University Shanghai Institute for Advanced Study,China(No.SN-ZJU-SIAS-0010)the Fundamental Research Funds for the Central Universities,China(Nos.226-2022-00142 and 226-2022-00051)。
文摘Federated learning(FL)is a novel technique in deep learning that enables clients to collaboratively train a shared model while retaining their decentralized data.However,researchers working on FL face several unique challenges,especially in the context of heterogeneity.Heterogeneity in data distributions,computational capabilities,and scenarios among clients necessitates the development of customized models and objectives in FL.Unfortunately,existing works such as FedAvg may not effectively accommodate the specific needs of each client.To address the challenges arising from heterogeneity in FL,we provide an overview of the heterogeneities in data,model,and objective(DMO).Furthermore,we propose a novel framework called federated mutual learning(FML),which enables each client to train a personalized model that accounts for the data heterogeneity(DH).A“meme model”serves as an intermediary between the personalized and global models to address model heterogeneity(MH).We introduce a knowledge distillation technique called deep mutual learning(DML)to transfer knowledge between these two models on local data.To overcome objective heterogeneity(OH),we design a shared global model that includes only certain parts,and the personalized model is task-specific and enhanced through mutual learning with the meme model.We evaluate the performance of FML in addressing DMO heterogeneities through experiments and compare it with other commonly used FL methods in similar scenarios.The results demonstrate that FML outperforms other methods and effectively addresses the DMO challenges encountered in the FL setting.
基金supported in part by the National Key Research and Development Program of China under Grant No.2020AAA-0108502the National Natural Science Foundation of China under Grant Nos.61972244,U19A2060,and 61925206the HighTech Support Program from Shanghai Committee of Science and Technology under Grant No.19511121100.
文摘Nowadays,application migration becomes more and more attractive.For example,it can make computation closer to data sources or make service closer to end-users,which may significantly decrease latency in edge computing.Yet,migrating applications among servers that are controlled by different platform owners raises security issues.We leverage hardware-secured trusted execution environment(TEE,aka.,enclave)technologies,such as Intel SGX,AMD SEV,and ARM TrustZone,for protecting critical computations on untrusted servers.However,these hardware TEEs propose non-uniform programming abstractions and are based on heterogeneous architectures,which not only forces programmers to develop secure applications targeting some specific abstraction but also hinders the migration of protected applications.Therefore,we propose UniTEE which gives a unified enclave programming abstraction across the above three hardware TEEs by using a microkernel-based design and enables the secure enclave migration by integrating heterogeneous migration techniques.We have implemented the prototype on real machines.The evaluation results show the migration support incurs nearly-zero runtime overhead and the migration procedure is also efficient.
基金Natural Science Foundation of China(31170339)the Special Foundation of National Science and Technology Basic Research(2013FY112300).
文摘Aims Foraging behavior was concerned less in plants especially for clonal aquatic species.The goal of this study was to test the foraging behav-ior and the function of phenotypic plasticity of a clonal submerged plant,Potamogeton maackianus A.Benn in variable manipulative environments.Methods To explore the foraging behavior of a clonal submerged plant,clones of P.maackianus were cultivated in variable environments of different substrate types and light conditions.We measured bio-mass,specific root area(SRA),specific root length(SRL),branch-ing angle of the ramet,chlorophyll content of apical leaves(CCAL),Fv/Fm(chlorophyll fluorescence parameter that indicates maximal photosystem II(PS II)efficiency)and biomass distribution in soil and light contrast treatments after harvest.Important Findings We found that the atypical non-clonal-like foraging behavior homogenized the performance of the stoloniferous clonal spe-cies,P.maackianus,in the measurement of biomass in our study.Substrate type had significant effects on most traits of P.maackianus,except on total biomass,shoot biomass distribution and CCAL.Patchy substrate mediated more root distribution into sand patch rather than into clay patch.Increase of SRA,SRL and Fv/Fm might be adaptive to substrate heterogeneity.Moreover,substrate hetero-geneity induced‘sidewalk’stolon lineage on the edge of sand patch along the patch border.Light condition only had significant effects on branching angle,shoot biomass distribution and Fv/Fm.Shade led to more shoot assignment in light-rich zone than in light-poor zone.Potamogeton maackianus adjusted the branching angle of ramet rather than clonal exploitation in heterogeneous substrates with patchy shade.Moreover,shade led to higher Fv/Fm.A com-promising foraging strategy promoted P.maackianus to escape unfa-vorable environments.Phenotypic plasticity,which maximized the performance of P.maackianus in the measurement of biomass in variable environments,may explain its domination in some shallow lakes of the middle and lower reaches of Yangtze River.
基金supported by the National Natural Science Foundation of China(Grant No.61103223)the Natural Science Foundation of Jiangsu Province(No.BK2011003).
文摘To reduce the running time of network simulation in heterogeneous computing environment,a network simulation task partition method,named LBPHCE,is put forward.In this method,the network simulation task is partitioned in comprehensive consideration of the load balance of both routing computing simulation and packet forwarding simulation.First,through benchmark experiments,the computation ability and routing simulation ability of each simulation machine are measured in the heterogeneous computing environment.Second,based on the computation ability of each simulation machine,the network simulation task is initially partitioned to meet the load balance of packet forwarding simulation in the heterogeneous computing environment,and then according to the routing computation ability,the scale of each partition is fine-tuned to satisfy the balance of the routing computing simulation,meanwhile the load balance of packet forwarding simulation is guaranteed.Experiments based on PDNS indicate that,compared to traditional uniform partition method,the LBPHCE method can reduce the total simulation running time by 26.3%in average,and compared to the liner partition method,it can reduce the running time by 18.3%in average.
基金supported by the Natural Science Foundation of China(11872189,11472116).
文摘In this paper,we investigate a reaction-diffusion-advection model with expanding fronts,which models the spatial transmission of West Nile virus(WNv)in a heterogeneous environment.A free boundary problem is formulated and the global existence and uniqueness of the solution is presented.In addition to a classical basic reproduction number,the spatial-temporal basic reproduction number for the model with null Dirichlet boundary condition is introduced and the risk index associated with the virus in spatial setting is defined,and their properties are discussed.Sufficient conditions for the WNv to vanish or spread are given,and the asymptotic behavior of the solution to the free boundary problem when the spreading occurs is established.Our results show that the initial number of infected populations and the expanding capability of the expanding fronts exhibit important impacts on the extinction or persistence of the virus.
基金the Beijing Municipal Commission of Planning and Natural Resources[No.GSJING(2022)0520]supported by the National Key Research and Development Program(2016YFC0501707)。
文摘The unidirectional impacts of soil conservation on dryland farming and local livelihood have been well studied.However,there are gaps in literature regarding the interactions among factors,such as labor force and financial capital,in response to soil conservation,as well as the influence of such interactions on the relationships between environmental projects and agricultural development.By combining the Chinese Soil Loss Equation(CSLE)model and statistical analyses,this study investigated the spatiotemporal characteristics of soil erosion while also identifying the controlling factors that affect crop and livestock farming.The aim was to reveal how these factors interacted to influence agriculture in two case study areas between 2000 and 2017.The results showed that soil erosion substantially declined as a consequence of soil conservation,while the output from crop farming increased,indicating that there was a synergy between soil conservation and dryland farming in both the districts,over time;and factors facilitating crop farming,such as labor force,agricultural machinery,and irrigation,appeared to exert an adverse effect on livestock farming in both areas,implying a tradeoff between livestock and crop farming in the context of programs such as the Grain for Green Program(GGP)in the Loess Hills.Such a tradeoff is a result of the competition for cropland and labor that exists due to grazing prohibition and rural-urban migration.Therefore,policy-related resolutions are required to address the land use conflicts and migration-related rural labor losing due to implementation of the GGP in the Loess Hills.
文摘This paper presents the design and implementation of a message-based distributed operating system kernel NDOS. The main purpose of the kernel is to support a distributed data processing system and a distributed DBMS. It uses the abstraction of communication between processes as basic mechanism. In NDOS, services and facilities such as message passing and process synchronization, which are related to IPC and may cause the change of the state of a process, are integrated into a single concept, an event. The initial version of NDOS ker- nel has been implemented on a full heterogeneous environment of different machines, LANs, and OSs with the original high-layered systems and applications are still provided.
基金supported by the National Natural Science Foundation of China(Grant No.11871475)the Hunan Provincial Innovation Foundation for Postgraduate(Grant No.CX20200096)the Fundamental Research Funds for the Central Universities of Central South University(Grant No.2020zzts024).
文摘In this paper,a reaction-diffusion SEI epidemic model with nonlinear incidence rate is proposed.The well-posedness of solutions is studied,including the existence of positive and unique classical solution and the existence and the ultimate boundedness of global solutions.The basic reproduction numbers are given in both heterogeneous and homogeneous environments.For spatially heterogeneous environment,by the comparison principle of the diffusion system,the infection-free steady state is proved to be globally asymptotically stable if R_(0)<1,if R_(0)>1,the system will be persistent and admit at least one positive steady state.For spatially homogenous environment,by constructing a Lyapunov function,the infect ion-free steady state is proved to be globally asymptotically stable if,R_(0)<1,and then the unique positive steady state is achieved and is proved to be globally asymptotically stable if R_(0)>1.Finally,two examples are given via numerical simulations,and then some control strategies are also presented by the sensitive analysis.