Articles de revistahttp://hdl.handle.net/2117/172932019-02-17T16:42:43Z2019-02-17T16:42:43ZRobust optimization for energy-efficient virtual machine consolidation in modern datacentersNasim, RobayetZola, Enrica ValeriaKassler, Andreashttp://hdl.handle.net/2117/1194092019-01-24T11:47:15Z2018-07-17T09:46:49ZRobust optimization for energy-efficient virtual machine consolidation in modern datacenters
Nasim, Robayet; Zola, Enrica Valeria; Kassler, Andreas
Energy efficient virtual machine (VM) consolidation in modern data centers is typically optimized using methods such as Mixed Integer Programming, which typically require precise input to the model. Unfortunately, many parameters are uncertain or very difficult to predict precisely in the real world. As a consequence, a once calculated solution may be highly infeasible in practice. In this paper, we use methods from robust optimization theory in order to quantify the impact of uncertainty in modern data centers. We study the impact of different parameter uncertainties on the energy efficiency and overbooking ratios such as e.g. VM resource demands, migration related overhead or the power consumption model of the servers used. We also show that setting aside additional resource to cope with uncertainty of workload influences the overbooking ration of the servers and the energy consumption. We show that, by using our model, Cloud operators can calculate a more robust migration schedule leading to higher total energy consumption. A more risky operator may well choose a more opportunistic schedule leading to lower energy consumption but also higher risk of SLA violation.
2018-07-17T09:46:49ZNasim, RobayetZola, Enrica ValeriaKassler, AndreasEnergy efficient virtual machine (VM) consolidation in modern data centers is typically optimized using methods such as Mixed Integer Programming, which typically require precise input to the model. Unfortunately, many parameters are uncertain or very difficult to predict precisely in the real world. As a consequence, a once calculated solution may be highly infeasible in practice. In this paper, we use methods from robust optimization theory in order to quantify the impact of uncertainty in modern data centers. We study the impact of different parameter uncertainties on the energy efficiency and overbooking ratios such as e.g. VM resource demands, migration related overhead or the power consumption model of the servers used. We also show that setting aside additional resource to cope with uncertainty of workload influences the overbooking ration of the servers and the energy consumption. We show that, by using our model, Cloud operators can calculate a more robust migration schedule leading to higher total energy consumption. A more risky operator may well choose a more opportunistic schedule leading to lower energy consumption but also higher risk of SLA violation.Optimal user association, backhaul routing and switching off in 5G heterogeneous networks with mesh millimeter wave backhaul linksMesodiakaki, AgapiZola, Enrica ValeriaSantos, RicardoKassler, Andreashttp://hdl.handle.net/2117/1192052019-01-24T11:47:13Z2018-07-10T11:24:55ZOptimal user association, backhaul routing and switching off in 5G heterogeneous networks with mesh millimeter wave backhaul links
Mesodiakaki, Agapi; Zola, Enrica Valeria; Santos, Ricardo; Kassler, Andreas
Next generation, i.e., fifth generation (5G), cellular networks will provide a significant higher capacity per area to support the ever-increasing traffic demands. In order to achieve that, many small cells need to be deployed that are connected using a combination of optical fiber links and millimeter-wave (mmWave) backhaul architecture to forward heterogeneous traffic over mesh topologies. In this paper, we present a general optimization framework for the design of policies that optimally solve the problem of where to associate a user, over which links to route its traffic towards which mesh gateway, and which base stations and backhaul links to switch o¿ in order to minimize the energy cost for the network operator and still satisfy the user demands. We develop an optimal policy based on mixed integer linear programming (MILP) which considers different user distribution and traffic demands over multiple time periods. We develop also a fast iterative two-phase solution heuristic, which associates users and calculates backhaul routes to maximize energy savings. Our strategies optimize the backhaul network configuration at each timeslot based on the current demands and user locations. We discuss the application of our policies to backhaul management of mmWave cellular networks in light of current trend of network softwarization (Software-Defined Networking, SDN). Finally, we present extensive numerical simulations of our proposed policies, which show how the algorithms can efficiently trade-off energy consumption with required capacity, while satisfying flow demand requirements.
2018-07-10T11:24:55ZMesodiakaki, AgapiZola, Enrica ValeriaSantos, RicardoKassler, AndreasNext generation, i.e., fifth generation (5G), cellular networks will provide a significant higher capacity per area to support the ever-increasing traffic demands. In order to achieve that, many small cells need to be deployed that are connected using a combination of optical fiber links and millimeter-wave (mmWave) backhaul architecture to forward heterogeneous traffic over mesh topologies. In this paper, we present a general optimization framework for the design of policies that optimally solve the problem of where to associate a user, over which links to route its traffic towards which mesh gateway, and which base stations and backhaul links to switch o¿ in order to minimize the energy cost for the network operator and still satisfy the user demands. We develop an optimal policy based on mixed integer linear programming (MILP) which considers different user distribution and traffic demands over multiple time periods. We develop also a fast iterative two-phase solution heuristic, which associates users and calculates backhaul routes to maximize energy savings. Our strategies optimize the backhaul network configuration at each timeslot based on the current demands and user locations. We discuss the application of our policies to backhaul management of mmWave cellular networks in light of current trend of network softwarization (Software-Defined Networking, SDN). Finally, we present extensive numerical simulations of our proposed policies, which show how the algorithms can efficiently trade-off energy consumption with required capacity, while satisfying flow demand requirements.A fast robust optimization-based heuristic for the deployment of green virtual network functionsMarotta, AntonioZola, Enrica ValeriaD'Andreagiovanni, FabioKassler, Andreashttp://hdl.handle.net/2117/1162202019-01-24T10:52:44Z2018-04-12T21:22:32ZA fast robust optimization-based heuristic for the deployment of green virtual network functions
Marotta, Antonio; Zola, Enrica Valeria; D'Andreagiovanni, Fabio; Kassler, Andreas
Network Function Virtualization (NFV) has attracted a lot of attention in the telecommunication field because it allows to virtualize core-business network functions on top of a NFV Infrastructure. Typically, virtual network functions (VNFs) can be represented as chains of Virtual Machines (VMs) or containers that exchange network traffic which are deployed inside datacenters on commodity hardware. In order to achieve cost efficiency, network operators aim at minimizing the power consumption of their NFV infrastructure. This can be achieved by using the minimum set of physical servers and networking equipment that are able to provide the quality of service required by the virtual functions in terms of computing, memory, disk and network related parameters. However, it is very difficult to predict precisely the resource demands required by the VNFs to execute their tasks. In this work, we apply the theory of robust optimization to deal with such parameter uncertainty. We model the problem of robust VNF placement and network embedding under resource demand uncertainty and network latency constraints using robust mixed integer optimization techniques. For online optimization, we develop fast solution heuristics. By using the virtualized Evolved Packet Core as use case, we perform a comprehensive evaluation in terms of performance, solution time and complexity and show that our heuristic can calculate robust solutions for large instances under one second.
2018-04-12T21:22:32ZMarotta, AntonioZola, Enrica ValeriaD'Andreagiovanni, FabioKassler, AndreasNetwork Function Virtualization (NFV) has attracted a lot of attention in the telecommunication field because it allows to virtualize core-business network functions on top of a NFV Infrastructure. Typically, virtual network functions (VNFs) can be represented as chains of Virtual Machines (VMs) or containers that exchange network traffic which are deployed inside datacenters on commodity hardware. In order to achieve cost efficiency, network operators aim at minimizing the power consumption of their NFV infrastructure. This can be achieved by using the minimum set of physical servers and networking equipment that are able to provide the quality of service required by the virtual functions in terms of computing, memory, disk and network related parameters. However, it is very difficult to predict precisely the resource demands required by the VNFs to execute their tasks. In this work, we apply the theory of robust optimization to deal with such parameter uncertainty. We model the problem of robust VNF placement and network embedding under resource demand uncertainty and network latency constraints using robust mixed integer optimization techniques. For online optimization, we develop fast solution heuristics. By using the virtualized Evolved Packet Core as use case, we perform a comprehensive evaluation in terms of performance, solution time and complexity and show that our heuristic can calculate robust solutions for large instances under one second.A meta-analysis on classification model performance in real-world datasets: an exploratory viewGómez Guillen, DavidRojas Espinosa, Alfonsohttp://hdl.handle.net/2117/1148702019-01-24T10:52:30Z2018-03-06T19:31:03ZA meta-analysis on classification model performance in real-world datasets: an exploratory view
Gómez Guillen, David; Rojas Espinosa, Alfonso
The No Free Lunch (NFL) Theorem imposes a theoretical restriction on optimization algorithms and their equal average performance on different problems, under some particular assumptions. Nevertheless, when brought into practice, a perceived “ranking” on the performance is usually perceived by engineers developing machine learning applications. Questions that naturally arise are what kinds of biases the real world has and in which ways can we take advantage from them. Using exploratory data analysis (EDA) on classification examples, we gather insight on some traits that set apart algorithms, datasets and evaluation measures and to what extent the NFL theorem, a theoretical result, applies under typical real-world constraints.
2018-03-06T19:31:03ZGómez Guillen, DavidRojas Espinosa, AlfonsoThe No Free Lunch (NFL) Theorem imposes a theoretical restriction on optimization algorithms and their equal average performance on different problems, under some particular assumptions. Nevertheless, when brought into practice, a perceived “ranking” on the performance is usually perceived by engineers developing machine learning applications. Questions that naturally arise are what kinds of biases the real world has and in which ways can we take advantage from them. Using exploratory data analysis (EDA) on classification examples, we gather insight on some traits that set apart algorithms, datasets and evaluation measures and to what extent the NFL theorem, a theoretical result, applies under typical real-world constraints.DYMO self-forwarding: a simple way for reducing the routing overhead in MANETSZola, Enrica ValeriaBarceló Arroyo, FranciscoMartín Escalona, Israelhttp://hdl.handle.net/2117/1123512019-01-24T11:45:37Z2017-12-20T22:29:18ZDYMO self-forwarding: a simple way for reducing the routing overhead in MANETS
Zola, Enrica Valeria; Barceló Arroyo, Francisco; Martín Escalona, Israel
Current routing protocols in Mobile Ad hoc Networks tend to use information on the position of the nodes in order to improve their features. In fact, without this information, protocols are hardly scalable since they tend to overflow the radio media with control packets, most of them being useless at the end. This paper presents the assessment of a modification of the DYMO protocol in order to include and use positioning information. The evaluation is carried out through simulations in realistic environments and connectivity condition. The possible error in the position is seldom considered in this kind of studies but here taken into account to catch the impact of realistic GPS devices or other sources of location techniques.
2017-12-20T22:29:18ZZola, Enrica ValeriaBarceló Arroyo, FranciscoMartín Escalona, IsraelCurrent routing protocols in Mobile Ad hoc Networks tend to use information on the position of the nodes in order to improve their features. In fact, without this information, protocols are hardly scalable since they tend to overflow the radio media with control packets, most of them being useless at the end. This paper presents the assessment of a modification of the DYMO protocol in order to include and use positioning information. The evaluation is carried out through simulations in realistic environments and connectivity condition. The possible error in the position is seldom considered in this kind of studies but here taken into account to catch the impact of realistic GPS devices or other sources of location techniques.On the energy cost of robustness for green virtual network function placement in 5G virtualized infrastructuresMarotta, AntonioD'Andreagiovanni, FabioKassler, AndreasZola, Enrica Valeriahttp://hdl.handle.net/2117/1085972019-01-24T11:45:02Z2017-10-10T13:28:20ZOn the energy cost of robustness for green virtual network function placement in 5G virtualized infrastructures
Marotta, Antonio; D'Andreagiovanni, Fabio; Kassler, Andreas; Zola, Enrica Valeria
Next generation 5G networks will rely on virtualized Data Centers (vDC) to host virtualized network functions on commodity servers. Such Network Function Virtualization (NFV) will lead to significant savings in terms of infrastructure cost and reduced management complexity. However, green strategies for networking and computing inside data centers, such as server consolidation or energy aware routing, should not negatively impact the quality and service level agreements expected from network operators. In this paper, we study how robust strategies that place virtual network func- tions (VNF) inside vDC impact the energy savings and the protection level against resource demand uncertainty. We propose novel optimization mod- els that allow the minimization of the energy of the computing and network infrastructure which is hosting a set of service chains that implement the VNFs. The model explicitly provides for robustness to unknown or impre- cisely formulated resource demand variations, powers down unused routers, switch ports and servers, and calculates the energy optimal VNF placement and network embedding also considering latency constraints on the service chains. We propose both exact and heuristic methods. Our experiments were carried out using the virtualized Evolved Packet Core (vEPC), which allows us to quantitatively assess the trade-off between energy cost, robust- ness and the protection level of the solutions against demand uncertainty. Our heuristic is able to converge to a good solution in a very short time, in comparison to the exact solver, which is not able to output better results in a longer run as demonstrated by our numerical evaluation. We also study the degree of robustness of a solution for a given protection level and the cost of additional energy needed because of the usage of more computing and network elements.
2017-10-10T13:28:20ZMarotta, AntonioD'Andreagiovanni, FabioKassler, AndreasZola, Enrica ValeriaNext generation 5G networks will rely on virtualized Data Centers (vDC) to host virtualized network functions on commodity servers. Such Network Function Virtualization (NFV) will lead to significant savings in terms of infrastructure cost and reduced management complexity. However, green strategies for networking and computing inside data centers, such as server consolidation or energy aware routing, should not negatively impact the quality and service level agreements expected from network operators. In this paper, we study how robust strategies that place virtual network func- tions (VNF) inside vDC impact the energy savings and the protection level against resource demand uncertainty. We propose novel optimization mod- els that allow the minimization of the energy of the computing and network infrastructure which is hosting a set of service chains that implement the VNFs. The model explicitly provides for robustness to unknown or impre- cisely formulated resource demand variations, powers down unused routers, switch ports and servers, and calculates the energy optimal VNF placement and network embedding also considering latency constraints on the service chains. We propose both exact and heuristic methods. Our experiments were carried out using the virtualized Evolved Packet Core (vEPC), which allows us to quantitatively assess the trade-off between energy cost, robust- ness and the protection level of the solutions against demand uncertainty. Our heuristic is able to converge to a good solution in a very short time, in comparison to the exact solver, which is not able to output better results in a longer run as demonstrated by our numerical evaluation. We also study the degree of robustness of a solution for a given protection level and the cost of additional energy needed because of the usage of more computing and network elements.DYMO self forwarding: a simple way for reducing the routing overhead in MANETsZola, Enrica ValeriaBarceló Arroyo, FranciscoMartín Escalona, Israelhttp://hdl.handle.net/2117/1083762019-01-24T10:50:57Z2017-10-05T12:32:24ZDYMO self forwarding: a simple way for reducing the routing overhead in MANETs
Zola, Enrica Valeria; Barceló Arroyo, Francisco; Martín Escalona, Israel
Current routing protocols in Mobile Ad hoc Networks tend to use information on the position of the nodes in order to improve their features. In fact, without this information, protocols are hardly scalable since they tend to overflow the radio media with control packets, most of them being useless at the end. This paper presents the assessment of a modification of the DYMO protocol in order to include and use positioning information. The evaluation is carried out through simulations in realistic environments and connectivity condition. The possible error in the position is seldom considered in this kind of studies but here taken into account to catch the impact of realistic GPS devices or other sources of location techniques.
2017-10-05T12:32:24ZZola, Enrica ValeriaBarceló Arroyo, FranciscoMartín Escalona, IsraelCurrent routing protocols in Mobile Ad hoc Networks tend to use information on the position of the nodes in order to improve their features. In fact, without this information, protocols are hardly scalable since they tend to overflow the radio media with control packets, most of them being useless at the end. This paper presents the assessment of a modification of the DYMO protocol in order to include and use positioning information. The evaluation is carried out through simulations in realistic environments and connectivity condition. The possible error in the position is seldom considered in this kind of studies but here taken into account to catch the impact of realistic GPS devices or other sources of location techniques.Predicting expected TCP throughput using genetic algorithmHernandez Benet, CristianKassler, AndreasZola, Enrica Valeriahttp://hdl.handle.net/2117/1032192019-01-24T11:43:34Z2017-04-03T12:14:58ZPredicting expected TCP throughput using genetic algorithm
Hernandez Benet, Cristian; Kassler, Andreas; Zola, Enrica Valeria
Predicting the expected throughput of TCP is important for several aspects such as e.g. determining handover criteria for future multihomed mobile nodes or determining the expected throughput of a given MPTCP subflow for load-balancing reasons. However, this is challenging due to time varying behavior of the underlying network characteristics. In this paper, we present a genetic-algorithm-based prediction model for estimating TCP throughput values. Our approach tries to find the best matching combination of mathematical functions that approximate a given time series that accounts for the TCP throughput samples using genetic algorithm. Based on collected historical datapoints about measured TCP throughput samples, our algorithm estimates expected throughput over time. We evaluate the quality of the prediction using different selection and diversity strategies for creating new chromosomes. Also, we explore the use of different fitness functions in order to evaluate the goodness of a chromosome. The goal is to show how different tuning on the genetic algorithm may have an impact on the prediction. Using extensive simulations over several TCP throughput traces, we find that the genetic algorithm successfully finds reasonable matching mathematical functions that allow to describe the TCP sampled throughput values with good fidelity. We also explore the effectiveness of predicting time series throughput samples for a given prediction horizon and estimate the prediction error and confidence.
2017-04-03T12:14:58ZHernandez Benet, CristianKassler, AndreasZola, Enrica ValeriaPredicting the expected throughput of TCP is important for several aspects such as e.g. determining handover criteria for future multihomed mobile nodes or determining the expected throughput of a given MPTCP subflow for load-balancing reasons. However, this is challenging due to time varying behavior of the underlying network characteristics. In this paper, we present a genetic-algorithm-based prediction model for estimating TCP throughput values. Our approach tries to find the best matching combination of mathematical functions that approximate a given time series that accounts for the TCP throughput samples using genetic algorithm. Based on collected historical datapoints about measured TCP throughput samples, our algorithm estimates expected throughput over time. We evaluate the quality of the prediction using different selection and diversity strategies for creating new chromosomes. Also, we explore the use of different fitness functions in order to evaluate the goodness of a chromosome. The goal is to show how different tuning on the genetic algorithm may have an impact on the prediction. Using extensive simulations over several TCP throughput traces, we find that the genetic algorithm successfully finds reasonable matching mathematical functions that allow to describe the TCP sampled throughput values with good fidelity. We also explore the effectiveness of predicting time series throughput samples for a given prediction horizon and estimate the prediction error and confidence.Optimising for energy or robustness? Trade-offs for VM consolidation in virtualized datacenters under uncertaintyZola, Enrica ValeriaKassler, Andreashttp://hdl.handle.net/2117/1031982019-01-24T11:43:11Z2017-04-03T10:16:02ZOptimising for energy or robustness? Trade-offs for VM consolidation in virtualized datacenters under uncertainty
Zola, Enrica Valeria; Kassler, Andreas
Reducing the energy consumption of virtualized datacenters and the Cloud is very important in order to lower CO2 footprint and operational cost of a Cloud operator. However, there is a trade-off between energy consumption and perceived application performance. In order to save energy, Cloud operators want to consolidate as many Virtual Machines (VM) on the fewest possible physical servers, possibly involving overbooking of resources. However, that may involve SLA violations when many VMs run on peak load. Such consolidation is typically done using VM migration techniques, which stress the network. As a consequence, it is important to find the right balance between the energy consumption and the number of migrations to perform. Unfortunately, the resources that a VM requires are not precisely known in advance, which makes it very difficult to optimise the VM migration schedule. In this paper, we therefore propose a novel approach based on the theory of robust optimisation. We model the VM consolidation problem as a robust Mixed Integer Linear Program and allow to specify bounds for e.g. resource requirements of the VMs. We show that, by using our model, Cloud operators can effectively trade-off uncertainty of resource requirements with total energy consumption. Also, our model allows us to quantify the price of the robustness in terms of energy saving against resource requirement violations.
The final publication is available at Springer via http://dx.doi.org/10.1007/s11590-016-1065-x
2017-04-03T10:16:02ZZola, Enrica ValeriaKassler, AndreasReducing the energy consumption of virtualized datacenters and the Cloud is very important in order to lower CO2 footprint and operational cost of a Cloud operator. However, there is a trade-off between energy consumption and perceived application performance. In order to save energy, Cloud operators want to consolidate as many Virtual Machines (VM) on the fewest possible physical servers, possibly involving overbooking of resources. However, that may involve SLA violations when many VMs run on peak load. Such consolidation is typically done using VM migration techniques, which stress the network. As a consequence, it is important to find the right balance between the energy consumption and the number of migrations to perform. Unfortunately, the resources that a VM requires are not precisely known in advance, which makes it very difficult to optimise the VM migration schedule. In this paper, we therefore propose a novel approach based on the theory of robust optimisation. We model the VM consolidation problem as a robust Mixed Integer Linear Program and allow to specify bounds for e.g. resource requirements of the VMs. We show that, by using our model, Cloud operators can effectively trade-off uncertainty of resource requirements with total energy consumption. Also, our model allows us to quantify the price of the robustness in terms of energy saving against resource requirement violations.La espectroscopía raman aplicada a la identificación de materiales pictóricosRuiz Moreno, SergioYúfera Gomez, José ManuelSoneira Ferrando, M. JoséBreitman Mansilla, Mónica CeliaMorillo Bosch, M. PazGràcia Rivas, Ignaciohttp://hdl.handle.net/2117/978762019-01-24T10:48:15Z2016-12-07T14:46:16ZLa espectroscopía raman aplicada a la identificación de materiales pictóricos
Ruiz Moreno, Sergio; Yúfera Gomez, José Manuel; Soneira Ferrando, M. José; Breitman Mansilla, Mónica Celia; Morillo Bosch, M. Paz; Gràcia Rivas, Ignacio
2016-12-07T14:46:16ZRuiz Moreno, SergioYúfera Gomez, José ManuelSoneira Ferrando, M. JoséBreitman Mansilla, Mónica CeliaMorillo Bosch, M. PazGràcia Rivas, Ignacio