Rajib Das

Followers

Following

Public Views

David Seamon

Kansas State University

Murat Cemal Yalcintan

Msgsu

Lisa Gualtieri

Tufts University

Enrica Papa

University of Westminster

Omid Pournik

Iran University of Medical Sciences

Paul Herring

The Open University

Massimo Regoli

"Tor Vergata" University of Rome

Izet Masic, MD, PhD, FWAAS, FEASA, FIAHSI, FEFMI, FACMI

University of Sarajevo

Ronaldo George Helal

UERJ - Universidade do Estado do Rio de Janeiro / Rio de Janeiro State University

Bálint Molnár

Eötvös Loránd University

Interests

Uploads

Papers by Rajib Das

An ACO-based multi-objective optimization for cooperating VM placement in cloud data center

The Journal of Supercomputing, 2021

High-performance computing requires numerous virtual machines (VMs) to meet the demand for servic... more High-performance computing requires numerous virtual machines (VMs) to meet the demand for services of the end-users. Such computation generates high intra-network traffic among the services running in different VMs. Thus, to achieve a high throughput of the application, VMs are to be deployed in physical machines in a way that reduces the communication cost and delays and thereby meets the high quality of services demanded by the users. On the other hand, the service providers aim to consolidate VMs in a minimum number of active physical machines to reduce their operational costs. Finding an optimal solution to minimize the cost of either the deployment or communication, by themselves are NP-Hard problems. Moreover, an attempt to optimize one disregarding the other may give a solution that is better in terms of cost but inferior in terms of throughput (or vice versa). Hence, we formulate the VM placement problem as a multi-objective optimization problem and propose an Ant Colony Optimization-based VM consolidation algorithm to find a solution in real-time. We have also presented a performance comparison of the proposed algorithm with existing mono-objective and multi-objective algorithms.

Cost Minimizing Reservation and Scheduling Algorithms for Public Clouds

IEEE Transactions on Cloud Computing, 2021

Energy and Makespan Aware Scheduling of Deadline Sensitive Tasks in the Cloud Environment

Journal of Grid Computing, 2021

Energy-Efficient Scheduling of Deadline-Sensitive and Budget-Constrained Workflows in the Cloud

Distributed Computing and Internet Technology, 2020

Eccentricity of the nodes of OTIS-cube and

In this paper we have classified the nodes of OTIS-cube based on their eccentricities. OTIS (opti... more In this paper we have classified the nodes of OTIS-cube based on their eccentricities. OTIS (optical transpose interconnection system) is a large scale optoelectronic computer architecture, proposed in , that benefit from both optical and electronic technologies. We show that radius and diameter of OTIS-Qn is n + 1 and 2n + 1 respectively. We also show that average eccentricity of OTIS-cube is (3n/2 + 1). In [10], a variant of OTIS-cube, called Enhanced OTIS-cube (E-OTIS-Qn) was proposed. E-OTIS-Qn is regular of degree n + 1 and maximally fault-tolerant. In this paper we have given a classification of the nodes of E-OTIS cube and derived expressions for the eccentricities of the nodes in each class. Based on these results we show that radius and diameter of E-OTIS-Qn is n + 1 and ⌊4n + 4/3⌋ respectively. We have also computed the average eccentricity of E-OTIS-Qn for values of n upto 20.

Download

Optimal Mapping of Applications on Data Centers in Sensor-Cloud Environment

Advances in Intelligent Systems and Computing, 2018

In sensor-cloud environment, sensing-as-a-service (Sen-aaS) is an emerging service paradigm that ... more In sensor-cloud environment, sensing-as-a-service (Sen-aaS) is an emerging service paradigm that allows on-demand provisioning of sensor resources as a service in a pay-per-use model. For each application, a disjoint set of virtual sensors (VS) are consolidated in a collaborative wireless sensor network (WSN) platform distributed across the globe. The virtual sensor network (VSN) of an application, formed using VSs, may span across multiple WSNs and the base station for each of these WSNs are placed in a host on a nearest cloud data center (DC). Here, sensor cloud plays the key role to conglomerate the data from various VSs, store them in different hosts, and transmit the same to end user application as a service (Sen-aaS). In this work, we address the problem of mapping applications on the hosts that conglomerate data from various VSs and transmit it to the end user as a constraint optimization problem. The main motivation is to minimize the maximum data migration time of all applications on sensor cloud while satisfying the host’s load-balancing constraint. We have proposed an algorithm which can solve the problem optimally under certain conditions. For the general case, if the load-balancing constraint is somewhat relaxed, the maximum delay obtained by our algorithm is optimal. When the load-balancing constraint is to be strictly satisfied, the cost of our solution is slightly more than the optimal provided by an Integer Linear Program.

Efficient relocation of mobile sensors for target coverage

2017 International Conference on Advances in Computing, Communications and Informatics (ICACCI), 2017

The coverage of interest points (targets) by mobile sensors in Wireless Sensor Networks is a very... more The coverage of interest points (targets) by mobile sensors in Wireless Sensor Networks is a very important and challenging problem. As the sensors are deployed randomly, initial placements of sensors may not guarantee coverage of all targets. In this paper, we consider the problem of covering all the targets in the surveillance region by relocating a subset of the sensors with the objective of minimizing the sum of the distance moved by the sensors. We have proposed an algorithm for the problem which runs in two phases. In the first phase a subset of sensors are relocated to new positions to ensure coverage of all the targets. The sensors which cover one or more targets after the first phase are called assigned sensors. In the second phase some of the unassigned sensors are relocated to ensure connectivity i.e. a path should exist from each assigned sensor to the base station (BS). The general case of the problem, where it is possible to cover multiple targets by a single sensor, finding the optimal placement of sensors is NP complete. We have designed a heuristic for relocating the sensors which runs in polynomial time and ensures coverage and connectivity. We have conducted experiments to evaluate the performance of our algorithm. The cost of our solution is very close to the optimal for the first phase of the algorithm. Also, the total cost (phase one plus phase two) is significantly less than that obtained by other recent works on this problem.

Resource Scheduling of Workflow Tasks in Cloud Environment

2019 IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS), 2019

In a business or scientific application, most of the jobs consist of multiple computational tasks... more In a business or scientific application, most of the jobs consist of multiple computational tasks, having precedence relationships among the tasks. Such a job is generally represented by a workflow. Efficient resource allocation for the tasks of a workflow-based job is a challenging area of research, especially in a public cloud environment. At one end, we need to complete the job at a lower cost by reducing the peak demand of virtual resources for the tasks. At the other end, we need to minimize the makespan of the workflow to achieve high throughput. This research focuses on minimizing the peak demand of the virtual resources to complete the job within its minimum makespan. Determining an optimal solution to the problem is NP-Hard. Hence, we have proposed a polynomial-time heuristic, called Resource-aware Workflow Scheduling (RaWS). We have also analyzed the performance of the proposed algorithm by comparing it with the existing algorithms, namely Heterogeneous Earliest-Finish-Time (HEFT) and Dynamic Heterogeneous Earliest-Finish-Time (DHEFT). The proposed algorithm notably reduces both the peak demand of virtual resources and the makespan of the workflow.

Efficient virtual machine placement in cloud environment

2017 International Conference on Advances in Computing, Communications and Informatics (ICACCI), 2017

Placement of Virtual Machines (VMs) on physical servers is a challenging area of research and has... more Placement of Virtual Machines (VMs) on physical servers is a challenging area of research and has received considerable interest among researchers in the field of Cloud Computing. This paper focuses on the reduction of cost in placing a set of VMs considering the constraints of available cloud resources. We have proposed a polynomial time heuristic to place VMs satisfying the constraints with the objective of minimizing the number of active servers. When the number of VMs and servers are small, one restricted case of the problem can be solved optimally by forming an Integer Programming Problem (IPP). The performance of the proposed heuristics are evaluated considering different types of scenarios and it is observed that the maximum or average number of servers required by one of the proposed heuristics is nearly equal to that given by IPP based solution.

Balanced Graph Partitioning based I/O Intensive Virtual Cluster Placement in Cloud Data Center

2021 12th International Conference on Computing Communication and Networking Technologies (ICCCNT), 2021

Cost Effective Algorithms for Participant Selection Problem in Mobile Crowd Sensing Environment

2018 Fifth International Conference on Parallel, Distributed and Grid Computing (PDGC), 2018

Recent developments in the areas of smart phones and wearable devices have paved the way to many ... more Recent developments in the areas of smart phones and wearable devices have paved the way to many emerging applications in Mobile Crowd Sensing (MCS) domain. There has been a phenomenal increase in the number of people using these devices. In this work we consider an MCS application to cover a set of points of interest (targets) with the objective of selecting a minimum set of participants among the mobile devices. This problem, coined as Participant Selection Problem (PSP) to provide desired coverage can be solved optimally using an Integer Linear Programming (ILP). But for large problem size, solving the ILP becomes impractical from the point of execution time as well as the requirement of having the necessary information about all the participants in one place. In this paper, we have proposed a distributed participants selection algorithm (DPSA) to be run at each participants where, the participants after exchanging messages with its neighbors, decides to remain active or idle. We have shown that, those participants which selects themselves as active suffice to cover all the desired targets. We have run a series of experiments to measure the performance of the proposed algorithm where we measure along with number of participants selected also the total energy consumption. Simulation results reveals the close proximity of DPSA compared to the optimal one providing the same coverage.

QoS Aware Energy Efficient VM Consolidation Techniques for a Virtualized Data Center

2018 IEEE/ACM 11th International Conference on Utility and Cloud Computing (UCC), 2018

The large-scale virtualized data centers in the Cloud environment consume huge amount of energy l... more The large-scale virtualized data centers in the Cloud environment consume huge amount of energy leading to high operational costs and emission of greenhouse gases. Energy consumption of a data center can be reduced by dynamically consolidating the virtual machines (VMs) to a minimum number of physical machines, using live migration. However, the dynamic workload of virtual machines makes the VM consolidation problem more challenging. In this paper, we have proposed a prediction based migration technique for the VMs, where we perform VM migrations based on the predicted CPU utilization. Extensive simulations show that the proposed technique substantially reduces energy consumption, number of VM migrations and Service Level Agreement (SLA) violations within a data center. The performance overheads associated with excessive migration of VMs increase the time needed by the VMs to complete their jobs. So in this paper, we have also proposed a deadline aware VM migration technique, which reduces the time taken by the VMs to execute their jobs significantly, thereby improving the Quality of Service (QoS). Such improvement in QoS is achieved at the cost of slight increase in the energy consumption within the data center. However, simulation results show that appropriate setting of deadlines for the VMs, helps in achieving a trade-off between energy consumption and the QoS.

Energy and quality of service-aware virtual machine consolidation in a cloud data center

The Journal of Supercomputing, 2020

The large-scale virtualized Cloud data centers consume huge amount of electrical energy leading t... more The large-scale virtualized Cloud data centers consume huge amount of electrical energy leading to high operational costs and emission of greenhouse gases. Virtual machine (VM) consolidation has been found to be a promising approach to improve resource utilization and reduce energy consumption of the data center. However, aggressive consolidation of VMs tends to increase the number of VM migrations and leads to over-utilization of hosts. This in turn affects the quality of service (QoS) of the applications running in the VMs. Thus, reduction in energy consumption and at the same time ensuring proper QoS to the Cloud users are one of the major challenges among the researchers. In this paper, we have proposed an energy efficient and QoS-aware VM consolidation technique in order to address this problem. We have used Markov chain-based prediction approach to identify the over-utilized and under-utilized hosts in the data center. We have also proposed an efficient VM selection and placement policy based on linear weighted sum approach to migrate the VMs from over-utilized and under-utilized hosts considering both energy and QoS. Extensive simulations using real-world traces and comparison with state-of-art strategies show that our VM consolidation approach substantially reduces energy consumption within a data center while delivering suitable QoS.

Bandwidth allocation for communicating virtual machines in cloud data centers

The Journal of Supercomputing, 2020

High-performance computing in a cloud environment may require massive data transfer among some of... more High-performance computing in a cloud environment may require massive data transfer among some of the virtual machines (VMs). These VMs are deployed in physical machines (hosts) of a data center. The data transfer among the communicating VMs may use the same shared communication links of the data center. Hence, it is important to have efficient bandwidth allocation policies for different data transfer requests (DTRs) which result in better utilization of bandwidth and fair allocation among the DTRs. In this paper, a few bandwidth allocation policies are proposed and their performances are analyzed. While designing these policies, the objective is the maximization of throughput and bandwidth utilization while minimizing the service time and turnaround time. Some of the policies are based on integer linear programming (ILP) which runs in exponential time while others are based on polynomial-time heuristics. Experimental results show that the performances of heuristic-based policies are comparable to those given by ILP-based exponential time policies.

Cost-effective routing as a service in sensor-cloud

International Journal of Sensor Networks, 2020

Sensor-cloud is a collaborative platform which allows multiple wireless sensor networks (WSN) to ... more Sensor-cloud is a collaborative platform which allows multiple wireless sensor networks (WSN) to pull their resources together for better utilisation and efficiency. In this work, we consider a set of applications, each characterised by the set of targets it intends to cover and our aim is to provide routing as a service (RaaS) to all these applications in the sensor-cloud. We should be able to run data gathering for each application in parallel with the minimum use of sensor-cloud resources. This is achieved by creating a collection of data gathering trees. We have proposed an algorithm which selects a minimal set of sensors to cover the required targets and obtains a spanning forest connecting them so that, the total energy cost of all the trees in the forest is minimised. Experimental results demonstrate the advantages of serving the applications by a sensor-cloud over standalone WSNs in terms of the usage of resources.

Download

Coverage of Targets in Mobile Sensor Networks With Restricted Mobility

IEEE Access, 2018

In this paper, we consider the problem of covering all regions of interests (targets) by relocati... more In this paper, we consider the problem of covering all regions of interests (targets) by relocating a set of mobile sensors such that total movement made by them is minimized. This problem itself is a challenging one and addressed recently by some researchers under free mobility model. We consider a more restricted version of the problem where sensors can move only in two mutually perpendicular directions. We first show that the optimal point to which a sensor must move to cover a specific target is different under this model from the one where sensors can move freely, and characterize such a point. On the basis of this observation, we have developed heuristics to solve the problem. The heuristics run in two phases; the first phase ensures coverage and the second phase, connectivity. In both the phases, the sensors can move only with restricted mobility. We have run a set of experiments to evaluate the performance of the proposed algorithm and found that the total movement made in the first phase is comparable to the solution given by an IPP (Integer Programming Problem). For the second phase, we have presented two heuristics MinCon and MinCon_m. The algorithm MinCon works by finding connected components of the graph consisting of sensor nodes. It then identifies destination locations where some sensors must be placed so that all necessary components become connected. Once the destinations are known, the problem is solved by mapping it to an LSAP (Linear Sum Assignment Problem). The other heuristic MinCon_m improves over MinCon by moving only a subset of sensors to their destinations using the solution of LSAP. It then finds the movement of the remaining sensors applying a technique used in the first phase. INDEX TERMS Connectivity, mobile sensors, rectilinear mobility, target coverage.

Download

Efficient Area Coverage in Wireless Sensor Networks Using Optimal Scheduling

Wireless Personal Communications, 2019

Wireless sensor networks generally have unique lifetime necessities. In any case, the density of ... more Wireless sensor networks generally have unique lifetime necessities. In any case, the density of the sensors may not be sufficiently substantial to fulfil the coverage requirement while meeting the lifetime constraint in the mean time. Once in a while coverage has to be traded for network lifetime. The proposed efficient pipeline based spatial temporal optimization scheduling for coverage optimization satisfies the coverage problem while meeting the lifetime constraint at the same time. In the proposed optimal scheduling, initially number of nodes in the network is clustered by using energy based one hop clustering algorithm. After the formation of clusters pipeline based spatial temporal optimization algorithm is used for the optimal scheduling. Here the optimization is improved by using trust of each sensor nodes and the area of clusters. Finally, data is aggregated through the optimally scheduled cluster nodes. The experimental results show that our proposed optimization scheduling substantially outperforms other schemes in terms of network lifetime, coverage redundancy and convergence time.

Bid Selection for Deadline Constrained Jobs over Spot VMs in Computational Cloud

Lecture Notes in Computer Science, 2016

Spot instance is an attractive pricing option to the cloud service users. But to use the spot ins... more Spot instance is an attractive pricing option to the cloud service users. But to use the spot instances, the cloud service users (CSU) have tos select an appropriate bid. The total cost of using spot instances and time to complete the user’s job are both determined by the bid chosen. Thus, selection of appropriate bid for spot instances is a critical problem for the CSUs because of the unpredictable nature of spot prices. The problem is more challenging when there is a deadline constraint. The Dynamic Bidding Algorithm (DBA) was considered to be an effective approach in this field. In this paper, we show that by combining a greedy approach for bidding with effective use of checkpointing, the total cost of using spot instances can be reduced further without violating the deadline constraint. To measure the cost of our solutions and those given by DBA with the optimum we have formulated an IPP for the given problem. In formulating the IPP it is assumed that spot prices at all future times are known. Experimental results based on some actual spot and on-demand price data show that the cost of our solution is comparable to those given by IPP. It is also observed that the cost decreases as the deadline to complete the job is extended.

Maximum lifetime scheduling for area coverage in wireless sensor networks

2015 IEEE International Conference on Advanced Networks and Telecommuncations Systems (ANTS), 2015

One important and challenging problem for sensor network is how to ensure coverage of the target ... more One important and challenging problem for sensor network is how to ensure coverage of the target area and maximize the lifetime of the network at the same time. This problem is known to be a NP-complete and there are many attempts to find an approximate solution. In this paper, we reduce the problem of area coverage to target coverage and it is possible to compute an upper bound on the lifetime achievable for target coverage. We have given an Integer Programming Problem (IPP) formulation of the problem using this upper bound. We do some approximation on the problem before applying IPP, so that solution could be obtained in reasonable time. In the experiments with number of sensors ranging from 50 to 150 we have been able to achieve the upper bound in all the cases.

GSE: a generalized full-access multistage interconnection network with minimum cost

Proceedings of 3rd International Conference on High Performance Computing (HiPC)

The conventional full-access unique-path MIN's are designed for N = 2" only. Thus, when... more