Firewall: Scaling up and down in cloud.
Firewalls have been widely deployed for securing networks and are the first line for defending against malicious traffic. A firewall checks each incoming or outgoing packet to decide whether to accept or reject the packet based on its policy. Setting up a centralized firewall for a whole cloud data center is infeasible. In this project, we aim at addressing most of the actual challenges in the centralized firewall by outsourcing some parts of the main firewall as several small firewalls.
Incremental NVF Optimal Placement.
Network Function Virtualization (NFV) is a paradigm to facilitate dynamic provisioning of network services through virtualization technologies. The optimal placement of virtual network functions is known to be an NP-hard problem. Our placement objectives consist of three goals: 1) minimizing end-to-end delay for each flow, 2) maximizing the network resource utilization, 3) minimizing network function installation costs.
Reducing Energy and Carbon Cost of Multiple Cloud Providers Using Cloud Federation and Solar Energy Prediction
Nowadays, the highest amount of money paid by cloud providers is about energy and carbon cost. The computation of energy and carbon cost plays a crucial role in data centers' operational costs and also could help to reduce their carbon footprint rate. In this research, generally, we aim to reduce the overall energy and carbon cost of multiple cloud providers with the use of cloud federation, renewable energy and other effective issues such as idle host awareness, solar energy prediction and dynamic PUE.
Intelligent countermeasure selection for Information security risk assessment
These days, the risk assessment of an organization's assets is a challenging task. If organizations did not pay significant attention to the security of their data and an attack happens against them, they will face too many problems like reputation loss or legal problems. One of the best ways to improve the process of risk assessment is by using our knowledge about data extracted from the business processes. With this kind of framework, we can make an accurate risk assessment model. What is left to do is using the logs of this system to selecting the mitigations based on the previous system's state in an intelligent way, which is the subject of this research.
Improving cvss vulnerability scoring.
CVSS is one of the most popular methods for calculating vulnerability severity. One of the disadvantages of the CVSS method is the failure to use the vulnerability history, which is addressed in this research. In the history of vulnerability, a variety of metrics can be considered, for example, when the vulnerability is fixed, the number of attacks between detection and fixing. The purpose of this research is to focus on these metrics.
A security test oracle based on security features extracted from source-code.
The importance of software application's security has increased rapidly over the recent years. Software security testing helps us to check the security behavior of a program. To determine whether the program behaves securely, we need an oracle who detects the correct program's behavior from security concept. There are various approaches to construct security test oracles. In this research, we will present a security test oracle based on security features which are extracted from source code of program to identify vulnerable codes and their types.
Energy efficient workflow scheduling in cloud computing environment
The growth of energy consumption has been explosive in current data centers. This explosion has led to greater advocacy of green computing, and many efforts and works focus on the task scheduling in order to reduce energy dissipation. In order to obtain more energy reduction as well as maintain the quality of service, this research proposes a DVFS-enabled Energy-efficient Workflow Scheduling algorithm.
Incremental VM placement
Over the last few years, virtual machine placement has become a chief operation in cloud computing. Researchers confront a lot of challenges according to different aspects such as computational complexity, end-to-end delay and etc.. in our placement, we want to model the optimal solution by integer linear programming and represent a greedy algorithm to make an incremental placement independent of the cost function.
Performance Analysis in Microservice
Architecture using OpenTracing
A microservice based cloud program consists of several services, each developed separately. OpenTracing is a new standard for controlling and viewing in-app events at different levels in the cloud. In this research, we intend to use OpenTracing by calculating multi metrics such as throughput, response time, request ratio of success and error to analyze the performance of software implemented on a cloud scale and will use the results obtained to identify system bottlenecks and monitor the system trends closely.
Virtual machine placement in the cloud
environment according to maintaining the quality of service
Virtual machine placement in the cloud environment can be done with different approaches. In this research, the maintenance of service quality for a distributed application in the cloud environment is considered.
An intelligent VM placement method for minimizing energy cost and carbon emission in distributed cloud data centers
By considering energy and carbon efficient virtual machine (VM) placement approaches, not only would the total power usage and carbon emission be reduced, but the excessive amounts of data centers' expenses including energy cost and carbon tax would also be significantly declined. In this research, for overcoming this compelling issue, we introduce a novel, intelligent VM placement approach for reducing energy consumption, carbon emission, and their related costs.
Anomaly Detection on OpenStack Log Analysis
OpenStack is a free open-source cloud computing platform that includes common services such as computing, image, network, and storage services. OpenStack Log Analysis is a significant part of detecting anomaly behavior. In this research, we analyze OpenStack log and divide it into two sections, log parsing and log mining. Furthermore, different datamining algorithms are used to find any unusual patterns that cause faults.
Service Function Chain Scheduling Problem in Edge Cputing
Service function chain scheduling, which is based on virtualization of network functions, has a significant impact on system performance and service delivery to users in the cloud. In this research, we decided to address the problem of service function chain scheduling in edge networks by considering two improvements: 1) network reconfiguration and 2) optimization of network resources.
Amir Mohammad Karamzadeh
Cold Start Latency Mitigation in Serverless Computing
One of the services in cloud computing is the computational model of serverless computing. In serverless computing, user codes are hosted without knowing the details of the infrastructure. One of the challenges with this model is the cold start problem. In this project, we intend to reduce the cold start delay.
Automating the translation of high-level security requirements for cloud users into an optimal placement model
Optimal placement of security functions in the cloud by cloud providers is one of the major challenges. The reason is to translate the user needs to the placement model. In this project, we intend to perform this challenge automatically.
Dynamic selection of cloud data center optimization objective
The purpose of data center optimization can be different (e.g., energy consumption, resource efficiency, energy cost, SLA). In most data centers, these goals are fixed. In this research, we intend to have these goals dynamically and based on a set of KPI in a data center.
A fluctuations-in-resource-demand aware virtual machine placement algorithm for cloud data centers
Reducing energy consumption in data centers is one of the major challenges. In most of the past work, fluctuations in user requests in resources in the location of virtual machines are not considered. In this research, we intend to optimize the location of the virtual machine, taking into account changes in resource demand, in order to reduce migration and optimize the energy consumption of the centers.