الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 230 |  |  |
| | | from | 230 | | |
Abstract
The traditional power grid is a centralized power system that is managed, monitored, and controlled only by the governmental sector. The core structure of this grid didn’t rely on the Information and Communication Technology (ICT) to gather any information from the composed elements. Therefore, there was no real-time information exchange between the utility stations and customer’s meters. Multiple blackouts occurred in the absence of this real-time exchange regarding the electricity demand or the electricity generation. In addition, the customer was not able to actively participate in monitoring or controlling his power consumption. Therefore, it was very important to collect such real-time information in a new novel framework and protocols. This framework must maintain a highly reliable and secure electricity infrastructure to meet future demand growth. Therefore, in the 21st century, some countries such as China, Brazil, USA, and India were seen as pioneers of the “Smart Grid” development. The Smart Grid has a unique nature and special characteristics. Its advanced operational requirements and massive scale of heterogeneous components participating in its construction, raised many challenges in terms of network and security management. For examples, a numerous attacks of different categories such as Denial of Service (DoS) may perpetrate entire grid or any specific point of its components to disrupt grid main functions and make the service unstable or unavailable. Therefore, last few years, the Smart Gird development increased the use of digital information and controls technology to improve the availability, sustainability, and security of the electric grid. Recent research works were presented to introduce “SDN-enabled Smart Grid” by coupling the Smart Grid and Software Defined Networking (SDN) to construct any of the following forms; “SDN-enabled NANs”, “SDN-enabled SCADA”, and “SDN-enabled Microgrids”. However, given the large scale nature of Smart Grid and the special architecture of SDN network, it was expecting to meet new hard challenges especially in the network and security constraints. For example, most of the proposed researches noted the lack of an adequate or realtime “SDN-enabled SCADA” simulators to validate the presented solutions. Additionally, a few of the existing simulation tools offers limited functionalities of the spectrum of energy domain. Moreover, in SDN environments, SDN network security needs to be everywhere within the whole system architecture. Therefore, within the context of “SDN-enabled Microgrids”, securing SDN network must be delivered as a service to protect the integrity, availability, and privacy of all interconnected grid resources and information. For example, the SDN controller is typically the primary target for attackers to disrupt the SDN network management. In addition, large scale networks must maintain an up-to-date suitable number of SDN controllers. These controllers must be well protected and distributed among the network nodes to reduce the network latency and improve the network availability, reliability and Quality of Service (QoS). As a result of all these challenges, “SDN-enabled Smart Grid” is not yet adopted by many industrial entities and still a new topic being presented by the research work. Therefore, this IV thesis is presented to address the enhancement of “SDN-enabled Smart Grid” development. In addition, it is presented to contribute to its network and security management. Two main challenges are addressed in this thesis. First, the “SDN-enabled SCADA” analysis and simulation techniques. Second, securing the “SDN-enabled Microgrids” controller placements against malicious attacks. The main contributions of this thesis that address these challenges can be summarized as follows: The 1st contribution: Implement PYGRID; a complete software development and assessment framework for gridaware software defined networking. PYGRID simulator is developed to facilitate studying the effectiveness of software defined networks-based Smart Grids SCADA elements Develop a Proactive SDN Security Module (PSM) on-top-of PYGRID SDN network architecture. PSM acts as a security application to mitigate one of the most devastating attacks in Smart Grid (DoS attacks). PYGRID automatically transforms a complete PLCs circuit and connections into an emulated virtual hosts in Mininet SDN environment, and not into a virtual machines hosted on top of any commercial hypervisor platforms (such as Microsoft Hyper-V, VMWare ESXi, and Citrix). It is adapted to host a data warehouse and flow management modules to store and regulate the circuit function according to real data specifications. In addition, PYGRID is equipped to host a real top-level developed SDN application to manage and control both the operational and resilience designs. Finally, PYGRID is developed to monitor and display this network status in real-time to act as real Human Machine Interface (HMI) and simplify the network management. Therefore, it may be considered as a simple simulator to test and validate the operational function of some power circuit before live deployments. The 2nd contribution: Develop SD-CPC; SDN Controller Placement Camouflage based on stochastic game for Moving-target Defense (MTD). SD-CPC relies on the Zero-Sum game as a stochastic game to guide the MTD solution. In addition, it is adapted to define the network vulnerabilities, evaluate the risk level of the system in real-time using Bayesian Attack Graph (BAG), and dynamically change SDN controller(s) locations. Further, UK Bulk Demand Points (BDP) network capacity map is used to simulate a sample real mini-grid topology. The results of this simulation are analyzed to initially achieve the optimal and substitute SDN controller placements within the grid and then migrate the SDN controller across these nodes based on SD-CPC security analysis.