Through the analysis to the DDoS（distributed denial of service） attack, it will conclude that at different time segments, the arrive rate of normal SYN （Synchronization） package are similar, while the abnormal pac...Through the analysis to the DDoS（distributed denial of service） attack, it will conclude that at different time segments, the arrive rate of normal SYN （Synchronization） package are similar, while the abnormal packages are different with the normal ones. Toward this situation a DDoS defense algorithm based on multi-segment timeout technology is presented, more than one timeout segment are set to control the net flow. Experiment results show that in the case of little flow, multi-segment timeout has the ability dynamic defense, so the system performance is improved and the system has high response rate.展开更多
This paper presents the combustion characteristics in hybrid rocket motors with multisegmented grain through three-dimensional numerical simulations.Multi-segmented grain is composed of several thin grains with two or...This paper presents the combustion characteristics in hybrid rocket motors with multisegmented grain through three-dimensional numerical simulations.Multi-segmented grain is composed of several thin grains with two or more ports.The numerical model consists of Navier-Stokes equations with turbulence,solid fuel pyrolysis,chemical reactions,a fluid–solid coupling model and a regression rate model.The simulations adopt 90%Hydrogen Peroxide(HP)and PolyEthylene(PE)as the propellant combination.The effects of the rotation,port number,fuel grain segment number and mid-chamber length on the flow field and combustion performances are analyzed.The results indicate that the multi-segmented grain configuration can strengthen the flow field,and the regression rate and combustion efficiency are enhanced.Take the cases with two grain segments and three ports for example,the regression rate is increased by 32.4%-45.1%and the combustion efficiency increases by 6%-8.6%in different rotation angles.展开更多
We consider a single-server vacation queueing system that operates in the following manner. When the server returns from a vacation, it observes the following rule. If there is at least one customer in the system, the...We consider a single-server vacation queueing system that operates in the following manner. When the server returns from a vacation, it observes the following rule. If there is at least one customer in the system, the server commences service and serves exhaustively before taking another vacation. If the server finds the system empty, it waits a fixed time c. At the expiration of this time, the server commences another vacation if no customer has arrived;otherwise, it serves exhaustively before commencing another vacation. Analytical results are derived for the mean waiting time in the system. The timeout scheme is shown to be a generalized scheme of which both the single vacation and multiple vacations schemes are special cases, with c=∞and c=0, respectively. The model is extended to the N-policy vacation queueing system.展开更多
Thermal rectification is an exotic thermal transport phenomenon,an analog to electrical rectification,in which heat flux along one direction is larger than that in the other direction and is of significant interest in...Thermal rectification is an exotic thermal transport phenomenon,an analog to electrical rectification,in which heat flux along one direction is larger than that in the other direction and is of significant interest in electronic device applications.However,achieving high thermal rectification efficiency or rectification ratio is still a scientific challenge.In this work,we performed a systematic simulation of thermal rectification by considering both efforts of thermal conductivity asymmetry and geometrical asymmetry in a multi-segment thermal rectifier.It is found that the high asymmetry of thermal conductivity and the asymmetry of the geometric structure of multi-segment thermal rectifiers can significantly enhance the thermal rectification,and the combination of both thermal conductivity asymmetry and geometrical asymmetry can further improve thermal rectification efficiency.This work suggests a possible way for improving thermal rectification devices by asymmetry engineering.展开更多
Accurately estimating of Retransmission TimeOut (RTO) in Content-Centric Networking (CCN) is crucial for efficient rate control in end nodes and effective interface ranking in intermediate routers. Toward this end, th...Accurately estimating of Retransmission TimeOut (RTO) in Content-Centric Networking (CCN) is crucial for efficient rate control in end nodes and effective interface ranking in intermediate routers. Toward this end, the Jacobson algorithm, which is an Exponentially Weighted Moving Average (EWMA) on the Round Trip Time (RTT) of previous packets, is a promising scheme. Assigning the lower bound to RTO, determining how an EWMA rapidly adapts to changes, and setting the multiplier of variance RTT have the most impact on the accuracy of this estimator for which several evaluations have been performed to set them in Transmission Control Protocol/Internet Protocol (TCP/IP) networks. However, the performance of this estimator in CCN has not been explored yet, despite CCN having a significant architectural difference with TCP/IP networks. In this study, two new metrics for assessing the performance of RTO estimators in CCN are defined and the performance of the Jacobson algorithm in CCN is evaluated. This evaluation is performed by varying the minimum RTO, EWMA parameters, and multiplier of variance RTT against different content popularity distribution gains. The obtained results are used to reconsider the Jacobson algorithm for accurately estimating RTO in CCN. Comparing the performance of the reconsidered Jacobson estimator with the existing solutions shows that it can estimate RTO simply and more accurately without any additional information or computation overhead.展开更多
Complex product development will inevitably face the design planning of the multi-coupled activities, and overlapping these activities could potentially reduce product development time, but there is a risk of the addi...Complex product development will inevitably face the design planning of the multi-coupled activities, and overlapping these activities could potentially reduce product development time, but there is a risk of the additional cost. Although the downstream task information dependence to the upstream task is already considered in the current researches, but the design process overall iteration caused by the information interdependence between activities is hardly discussed; especially the impact on the design process＇ overall iteration from the valid information accumulation process. Secondly, most studies only focus on the single overlapping process of two activities, rarely take multi-segment and multi-ply overlapping process of multi coupled activities into account; especially the inherent link between product development time and cost which originates from the overlapping process of multi coupled activities. For the purpose of solving the above problems, as to the insufficiency of the accumulated valid information in overlapping process, the function of the valid information evolution （VIE） degree is constructed. Stochastic process theory is used to describe the design information exchange and the valid information accumulation in the overlapping segment, and then the planning models of the single overlapping segment are built. On these bases, by analyzing overlapping processes and overlapping features of multi-coupling activities, multi-segment and multi-ply overlapping planning models are built; by sorting overlapping processes and analyzing the construction of these planning models, two conclusions are obtained： （1） As to multi-segment and multi-ply overlapping of multi coupled activities, the total decrement of the task set development time is the sum of the time decrement caused by basic overlapping segments, and minus the sum of the time increment caused by multiple overlapping segments; （2） the total increment of development cost is the sum of the cost increment caused by all overlapping process. And then, based on overlapping degree analysis of these planning models, by the V1E degree function, the four lemmas theory proofs are represented, and two propositions are finally proved： （1） The multi-ply overlapping of the multi coupled activities will weaken the basic overlapping effect on the development cycle time reduction （2） Overlapping the multi coupled activities will decrease product development cycle, but increase product development cost. And there is trade-off between development time and cost. And so, two methods are given to slacken and eliminate multi-ply overlapping effects. At last, an example about a vehicle upper subsystem design illustrates the application of the proposed models; compared with a sequential execution pattern, the decreasing of development cycle （22%） and the increasing of development cost （3%） show the validity of the method in the example The proposed research not only lays a theoretical foundation for correctly planning complex product development process, but also provides specific and effective operation methods for overlapping multi coupled activities.展开更多
This paper presents an enhanced version of the standard shooting method that enables problems with two unknown parameters to be solved.A novel approach is applied to the analysis of the natural vibrations of Euler-Ber...This paper presents an enhanced version of the standard shooting method that enables problems with two unknown parameters to be solved.A novel approach is applied to the analysis of the natural vibrations of Euler-Bernoulli beams.The proposed algorithm,named as two-parameter multiple shooting method,is a new powerful numerical tool for calculating the natural frequencies and modes of multi-segment prismatic and non-prismatic beams with different boundary conditions.The impact of the axial force and additional point masses is also taken into account.Due to the fact that the method is based directly on the fourth-order ordinary differential equation,the structures do not have to be divided into many small elements to obtain an accurate enough solution,even though the geometry is very complex.To verify the proposed method,three different examples are considered,i.e.,a three-segment non-prismatic beam,a prismatic column subject to non-uniformly distributed compressive loads,and a two-segment beam with an additional point mass.Numerical analyses are carried out with the software MATHEMATICA.The results are compared with the solutions computed by the commercial finite element program SOFiSTiK.Good agreement is achieved,which confirms the correctness and high effectiveness of the formulated algorithm.展开更多
A vortex domain wall's(VW) magnetic racetrack memory's high performance depends on VW structural stability,high speed, low power consumption and high storage density. In this study, these critical parameters w...A vortex domain wall's(VW) magnetic racetrack memory's high performance depends on VW structural stability,high speed, low power consumption and high storage density. In this study, these critical parameters were investigated in magnetic multi-segmented nanowires using micromagnetic simulation. Thus, an offset magnetic nanowire with a junction at the center was proposed for this purpose. This junction was implemented by shifting one portion of the magnetic nanowire horizontally in the x-direction(l) and vertically(d) in the y-direction. The VW structure became stable by manipulating magnetic properties, such as magnetic saturation(M_(4)) and magnetic anisotropy energy(K_(u)). In this case, increasing the values of M_(4) ≥ 800 kA/m keeps the VW structure stable during its dynamics and pinning and depinning in offset nanowires,which contributes to maintenance of the storage memory's lifetime for a longer period. It was also found that the VW moved with a speed of 500 m/s, which is desirable for VW racetrack memory devices. Moreover, it was revealed that the VW velocity could be controlled by adjusting the offset area dimensions(l and d), which helps to drive the VW by using low current densities and reducing the thermal-magnetic spin fluctuations. Further, the depinning current density of the VW(J_(d)) over the offset area increases as d increases and l decreases. In addition, magnetic properties, such as the M_(4) and K_(u),can affect the depinning process of the VW through the offset area. For high storage density, magnetic nanowires(multisegmented) with four junctions were designed. In total, six states were found with high VW stability, which means three bits per cell. Herein, we observed that the depinning current density(J_(d)) for moving the VW from one state to another was highly influenced by the offset area geometry(l and d) and the material's magnetic properties, such as the M_(4) and K_(u).展开更多
A novel flat-flat resonator consisting of two crystals（Nd：YAG ＋ Nd：YVO4） is established for power scaling in a diode-end-pumped solid-state laser. We systematically compare laser characteristics between multi-seg...A novel flat-flat resonator consisting of two crystals（Nd：YAG ＋ Nd：YVO4） is established for power scaling in a diode-end-pumped solid-state laser. We systematically compare laser characteristics between multi-segmented（Nd：YAG ＋ Nd：YVO4） and conventional composite（Nd：YAG ＋ Nd：YAG） crystals to demonstrate the feasibility of spectral line matching for output power scale-up in end-pumped lasers. A maximum continuous-wave output power of 79.2 W is reported at 1064 nm, with Mx2= 4.82, My2= 5.48, and a pumping power of 136 W in the multi-segmented crystals（Nd：YAG ＋ Nd：YVO4）. Compared to conventional composite crystals（Nd：YAG ＋ Nd：YAG）, the optical-optical conversion efficiency of multi-segmented crystals（Nd：YAG ＋ Nd：YVO4） from 808 nm to 1064 nm is enhanced from 30% to 58.8%,while the laser output sensitivity as affected by the diode-laser temperature is reduced from 55% to 9%.展开更多
高性能水声路由协议要求协议具有高吞吐量的同时,能耗尽可能小,这是提高能源利用效率的关键.针对这一问题,提出基于能量优化和超时传输的水声传感器网络路由协议.首先,路由寻找阶段融合能量均衡策略并利用退避机制尽可能接收多的路由请...高性能水声路由协议要求协议具有高吞吐量的同时,能耗尽可能小,这是提高能源利用效率的关键.针对这一问题,提出基于能量优化和超时传输的水声传感器网络路由协议.首先,路由寻找阶段融合能量均衡策略并利用退避机制尽可能接收多的路由请求再进行回复来控制路由开销;然后,数据传输阶段引入超时传输机制,利用主备份路由协同传输并监听和判断链路有效性,融合碰撞避免机制规避碰撞;最后,通过仿真和水下实验与按需平面距离向量路由协议(Ad Hoc on-demand distance vector routing,AODV)和矢量转发路由协议(vector based forwarding,VBF)进行对比.仿真实验结果表明:所提出协议相比于AODV与VBF,吞吐量和数据包成功接收率均有提升,且具有更低的单位字节能耗;水下实验结果表明:相比于AODV,所提出协议吞吐量提升了11.84%,数据包成功接收率提升了15.62%,单位字节能耗可降低0.10 J.展开更多
Accurate early classification of elephant flows(elephants)is important for network management and resource optimization.Elephant models,mainly based on the byte count of flows,can always achieve high accuracy,but not ...Accurate early classification of elephant flows(elephants)is important for network management and resource optimization.Elephant models,mainly based on the byte count of flows,can always achieve high accuracy,but not in a time-efficient manner.The time efficiency becomes even worse when the flows to be classified are sampled by flow entry timeout over Software-Defined Networks(SDNs)to achieve a better resource efficiency.This paper addresses this situation by combining co-training and Reinforcement Learning(RL)to enable a closed-loop classification approach that divides the entire classification process into episodes,each involving two elephant models.One predicts elephants and is retrained by a selection of flows automatically labeled online by the other.RL is used to formulate a reward function that estimates the values of the possible actions based on the current states of both models and further adjusts the ratio of flows to be labeled in each phase.Extensive evaluation based on real traffic traces shows that the proposed approach can stably predict elephants using the packets received in the first 10% of their lifetime with an accuracy of over 80%,and using only about 10% more control channel bandwidth than the baseline over the evolved SDNs.展开更多
基金Supported by the Natural Science Foundation ofHebei Province (F2004000133)
文摘Through the analysis to the DDoS（distributed denial of service） attack, it will conclude that at different time segments, the arrive rate of normal SYN （Synchronization） package are similar, while the abnormal packages are different with the normal ones. Toward this situation a DDoS defense algorithm based on multi-segment timeout technology is presented, more than one timeout segment are set to control the net flow. Experiment results show that in the case of little flow, multi-segment timeout has the ability dynamic defense, so the system performance is improved and the system has high response rate.
文摘This paper presents the combustion characteristics in hybrid rocket motors with multisegmented grain through three-dimensional numerical simulations.Multi-segmented grain is composed of several thin grains with two or more ports.The numerical model consists of Navier-Stokes equations with turbulence,solid fuel pyrolysis,chemical reactions,a fluid–solid coupling model and a regression rate model.The simulations adopt 90%Hydrogen Peroxide(HP)and PolyEthylene(PE)as the propellant combination.The effects of the rotation,port number,fuel grain segment number and mid-chamber length on the flow field and combustion performances are analyzed.The results indicate that the multi-segmented grain configuration can strengthen the flow field,and the regression rate and combustion efficiency are enhanced.Take the cases with two grain segments and three ports for example,the regression rate is increased by 32.4%-45.1%and the combustion efficiency increases by 6%-8.6%in different rotation angles.
文摘We consider a single-server vacation queueing system that operates in the following manner. When the server returns from a vacation, it observes the following rule. If there is at least one customer in the system, the server commences service and serves exhaustively before taking another vacation. If the server finds the system empty, it waits a fixed time c. At the expiration of this time, the server commences another vacation if no customer has arrived;otherwise, it serves exhaustively before commencing another vacation. Analytical results are derived for the mean waiting time in the system. The timeout scheme is shown to be a generalized scheme of which both the single vacation and multiple vacations schemes are special cases, with c=∞and c=0, respectively. The model is extended to the N-policy vacation queueing system.
基金Project supported by the National Natural Science Foundation of China(Grant No.12274355)Xiamen University Malaysia Research Fund(Grant Nos.XMUMRF/2022C9/IORI/003 and XMUMRF/2022-C10/IORI/004)。
文摘Thermal rectification is an exotic thermal transport phenomenon,an analog to electrical rectification,in which heat flux along one direction is larger than that in the other direction and is of significant interest in electronic device applications.However,achieving high thermal rectification efficiency or rectification ratio is still a scientific challenge.In this work,we performed a systematic simulation of thermal rectification by considering both efforts of thermal conductivity asymmetry and geometrical asymmetry in a multi-segment thermal rectifier.It is found that the high asymmetry of thermal conductivity and the asymmetry of the geometric structure of multi-segment thermal rectifiers can significantly enhance the thermal rectification,and the combination of both thermal conductivity asymmetry and geometrical asymmetry can further improve thermal rectification efficiency.This work suggests a possible way for improving thermal rectification devices by asymmetry engineering.
文摘Accurately estimating of Retransmission TimeOut (RTO) in Content-Centric Networking (CCN) is crucial for efficient rate control in end nodes and effective interface ranking in intermediate routers. Toward this end, the Jacobson algorithm, which is an Exponentially Weighted Moving Average (EWMA) on the Round Trip Time (RTT) of previous packets, is a promising scheme. Assigning the lower bound to RTO, determining how an EWMA rapidly adapts to changes, and setting the multiplier of variance RTT have the most impact on the accuracy of this estimator for which several evaluations have been performed to set them in Transmission Control Protocol/Internet Protocol (TCP/IP) networks. However, the performance of this estimator in CCN has not been explored yet, despite CCN having a significant architectural difference with TCP/IP networks. In this study, two new metrics for assessing the performance of RTO estimators in CCN are defined and the performance of the Jacobson algorithm in CCN is evaluated. This evaluation is performed by varying the minimum RTO, EWMA parameters, and multiplier of variance RTT against different content popularity distribution gains. The obtained results are used to reconsider the Jacobson algorithm for accurately estimating RTO in CCN. Comparing the performance of the reconsidered Jacobson estimator with the existing solutions shows that it can estimate RTO simply and more accurately without any additional information or computation overhead.
基金sponsored by Jiangsu Provincial Colleges and Universities Natural Science Foundation of China (Grant No.08KJD410001)Humanities and Social Sciences Planning Fund of Ministry of Education of China (Grant No. 12YJAZH151)Humanities and Social Sciences Youth Fund of Ministry of Education of China (Grant No. 12YJCZH209)
文摘Complex product development will inevitably face the design planning of the multi-coupled activities, and overlapping these activities could potentially reduce product development time, but there is a risk of the additional cost. Although the downstream task information dependence to the upstream task is already considered in the current researches, but the design process overall iteration caused by the information interdependence between activities is hardly discussed; especially the impact on the design process＇ overall iteration from the valid information accumulation process. Secondly, most studies only focus on the single overlapping process of two activities, rarely take multi-segment and multi-ply overlapping process of multi coupled activities into account; especially the inherent link between product development time and cost which originates from the overlapping process of multi coupled activities. For the purpose of solving the above problems, as to the insufficiency of the accumulated valid information in overlapping process, the function of the valid information evolution （VIE） degree is constructed. Stochastic process theory is used to describe the design information exchange and the valid information accumulation in the overlapping segment, and then the planning models of the single overlapping segment are built. On these bases, by analyzing overlapping processes and overlapping features of multi-coupling activities, multi-segment and multi-ply overlapping planning models are built; by sorting overlapping processes and analyzing the construction of these planning models, two conclusions are obtained： （1） As to multi-segment and multi-ply overlapping of multi coupled activities, the total decrement of the task set development time is the sum of the time decrement caused by basic overlapping segments, and minus the sum of the time increment caused by multiple overlapping segments; （2） the total increment of development cost is the sum of the cost increment caused by all overlapping process. And then, based on overlapping degree analysis of these planning models, by the V1E degree function, the four lemmas theory proofs are represented, and two propositions are finally proved： （1） The multi-ply overlapping of the multi coupled activities will weaken the basic overlapping effect on the development cycle time reduction （2） Overlapping the multi coupled activities will decrease product development cycle, but increase product development cost. And there is trade-off between development time and cost. And so, two methods are given to slacken and eliminate multi-ply overlapping effects. At last, an example about a vehicle upper subsystem design illustrates the application of the proposed models; compared with a sequential execution pattern, the decreasing of development cycle （22%） and the increasing of development cost （3%） show the validity of the method in the example The proposed research not only lays a theoretical foundation for correctly planning complex product development process, but also provides specific and effective operation methods for overlapping multi coupled activities.
文摘This paper presents an enhanced version of the standard shooting method that enables problems with two unknown parameters to be solved.A novel approach is applied to the analysis of the natural vibrations of Euler-Bernoulli beams.The proposed algorithm,named as two-parameter multiple shooting method,is a new powerful numerical tool for calculating the natural frequencies and modes of multi-segment prismatic and non-prismatic beams with different boundary conditions.The impact of the axial force and additional point masses is also taken into account.Due to the fact that the method is based directly on the fourth-order ordinary differential equation,the structures do not have to be divided into many small elements to obtain an accurate enough solution,even though the geometry is very complex.To verify the proposed method,three different examples are considered,i.e.,a three-segment non-prismatic beam,a prismatic column subject to non-uniformly distributed compressive loads,and a two-segment beam with an additional point mass.Numerical analyses are carried out with the software MATHEMATICA.The results are compared with the solutions computed by the commercial finite element program SOFiSTiK.Good agreement is achieved,which confirms the correctness and high effectiveness of the formulated algorithm.
文摘A vortex domain wall's(VW) magnetic racetrack memory's high performance depends on VW structural stability,high speed, low power consumption and high storage density. In this study, these critical parameters were investigated in magnetic multi-segmented nanowires using micromagnetic simulation. Thus, an offset magnetic nanowire with a junction at the center was proposed for this purpose. This junction was implemented by shifting one portion of the magnetic nanowire horizontally in the x-direction(l) and vertically(d) in the y-direction. The VW structure became stable by manipulating magnetic properties, such as magnetic saturation(M_(4)) and magnetic anisotropy energy(K_(u)). In this case, increasing the values of M_(4) ≥ 800 kA/m keeps the VW structure stable during its dynamics and pinning and depinning in offset nanowires,which contributes to maintenance of the storage memory's lifetime for a longer period. It was also found that the VW moved with a speed of 500 m/s, which is desirable for VW racetrack memory devices. Moreover, it was revealed that the VW velocity could be controlled by adjusting the offset area dimensions(l and d), which helps to drive the VW by using low current densities and reducing the thermal-magnetic spin fluctuations. Further, the depinning current density of the VW(J_(d)) over the offset area increases as d increases and l decreases. In addition, magnetic properties, such as the M_(4) and K_(u),can affect the depinning process of the VW through the offset area. For high storage density, magnetic nanowires(multisegmented) with four junctions were designed. In total, six states were found with high VW stability, which means three bits per cell. Herein, we observed that the depinning current density(J_(d)) for moving the VW from one state to another was highly influenced by the offset area geometry(l and d) and the material's magnetic properties, such as the M_(4) and K_(u).
基金Project supported by the National Defense Pre-Research Foundation of China(Grant No.9140A020105)
文摘A novel flat-flat resonator consisting of two crystals（Nd：YAG ＋ Nd：YVO4） is established for power scaling in a diode-end-pumped solid-state laser. We systematically compare laser characteristics between multi-segmented（Nd：YAG ＋ Nd：YVO4） and conventional composite（Nd：YAG ＋ Nd：YAG） crystals to demonstrate the feasibility of spectral line matching for output power scale-up in end-pumped lasers. A maximum continuous-wave output power of 79.2 W is reported at 1064 nm, with Mx2= 4.82, My2= 5.48, and a pumping power of 136 W in the multi-segmented crystals（Nd：YAG ＋ Nd：YVO4）. Compared to conventional composite crystals（Nd：YAG ＋ Nd：YAG）, the optical-optical conversion efficiency of multi-segmented crystals（Nd：YAG ＋ Nd：YVO4） from 808 nm to 1064 nm is enhanced from 30% to 58.8%,while the laser output sensitivity as affected by the diode-laser temperature is reduced from 55% to 9%.
文摘高性能水声路由协议要求协议具有高吞吐量的同时,能耗尽可能小,这是提高能源利用效率的关键.针对这一问题,提出基于能量优化和超时传输的水声传感器网络路由协议.首先,路由寻找阶段融合能量均衡策略并利用退避机制尽可能接收多的路由请求再进行回复来控制路由开销;然后,数据传输阶段引入超时传输机制,利用主备份路由协同传输并监听和判断链路有效性,融合碰撞避免机制规避碰撞;最后,通过仿真和水下实验与按需平面距离向量路由协议(Ad Hoc on-demand distance vector routing,AODV)和矢量转发路由协议(vector based forwarding,VBF)进行对比.仿真实验结果表明:所提出协议相比于AODV与VBF,吞吐量和数据包成功接收率均有提升,且具有更低的单位字节能耗;水下实验结果表明:相比于AODV,所提出协议吞吐量提升了11.84%,数据包成功接收率提升了15.62%,单位字节能耗可降低0.10 J.
基金supported by the National Natural Science Foundation of China(61962016)the Ministry of Science and Technology of China(G2022033002L)+1 种基金National Natural Science Foundation of Guangxi(2022JJA170057)Guangxi Education Department’s Project on Improving the Basic Research Ability of Young and Middleaged Teachers in Universities(2023ky0812,Research on Statistical Network Delay Predictions in Large-scale SDNs).
文摘Accurate early classification of elephant flows(elephants)is important for network management and resource optimization.Elephant models,mainly based on the byte count of flows,can always achieve high accuracy,but not in a time-efficient manner.The time efficiency becomes even worse when the flows to be classified are sampled by flow entry timeout over Software-Defined Networks(SDNs)to achieve a better resource efficiency.This paper addresses this situation by combining co-training and Reinforcement Learning(RL)to enable a closed-loop classification approach that divides the entire classification process into episodes,each involving two elephant models.One predicts elephants and is retrained by a selection of flows automatically labeled online by the other.RL is used to formulate a reward function that estimates the values of the possible actions based on the current states of both models and further adjusts the ratio of flows to be labeled in each phase.Extensive evaluation based on real traffic traces shows that the proposed approach can stably predict elephants using the packets received in the first 10% of their lifetime with an accuracy of over 80%,and using only about 10% more control channel bandwidth than the baseline over the evolved SDNs.
