Wireless network architecture for future smart grid machine to machine communications

Nafi, N 2017, Wireless network architecture for future smart grid machine to machine communications, Doctor of Philosophy (PhD), Engineering, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Nafi.pdf Nafi.pdf Click to show the corresponding preview/stream application/pdf; Bytes
Title Wireless network architecture for future smart grid machine to machine communications
Author(s) Nafi, N
Year 2017
Abstract Transformation of the conventional power grid into an efficient power delivery network is an important advance that will benefit consumers, business and the environment by providing improved integration of renewable energy, including solar and wind. A reliable, low latency communication system is a fundamental requirement for smart power grids. To achieve bidirectional energy distribution capability and to support diverse Smart Grid (SG) applications, the modern SG requires the capacity to handle the traffic generated by machine to machine (M2M) communication infrastructure. Successful integration of numerous SG applications, renewable energy sources and Electric Vehicles (EVs) into a conventional power grid would not be possible without a communication network that has been designed to support the needs of the new and innovative renewable power generation, distribution and storage technologies. While the legacy communication infrastructure, utilized to support the existing power network, fails to support all of the SG functionalities, Software Defined Networking (SDN), based on wireless communication systems, has the potential to provide an effective solution. SDN offers a range of features that fulfill the unique requirements of the SG applications. Being a new networking paradigm, SDN remains to be implemented for SG M2M communication scenarios and there remain a number of challenges that need to be overcome.

M2M communication protocols and standards provide a starting point for the broader development of SG communication networks that can be enhanced by abstracting high-level network functionalities. The aim of this research was to carry out an in-depth study on the future SG communication networks and to propose solutions to identified limitations of existing communication networks. Keeping this intention in mind, the study first focuses on the SG application modeling techniques based on the traffic requirements and power supply load profiles. To address the dynamicity of the traffic model and demand load curve, a series of analytical models and smart algorithms were developed. SG application models were developed and evaluated using a range of scenarios reflecting typical usage. Heterogenous network architectures and efficient traffic models were developed to identify an appropriate wireless communication technology and to maximize the network performance for major SG applications. However, a careful observation of the communication networks ability to manage and control the diverse M2M communications reveals that the inadequate dynamic communication network configuration capability would be a problem for future SG applications.

M2M communication protocols and standards provide a starting point for the broader development of SG communication networks that can be enhanced by abstracting high-level network functionalities. To realize the full potential of the SGs and deployment scenarios it is essential to analyze the major applications and key requirements to develop those applications. Also, it might be necessary to select an appropriate communication technology for each of the power system domains. The study first focuses on the SG application modeling techniques based on the traffic requirement and load supply profiles of the power system. To address dynamicity of the traffic model and demand load curve, a series of analytical models and smart algorithms were developed. The developed SG application models were further evaluated using simulation scenarios and a test bed model. The challenge of selecting an appropriate wireless communication technology and maximizing network performance for major SG applications was handled by developing multiple heterogenous network architectures and efficient traffic models. A comprehensive literature review of the state of the art of SG applications and standards was carried out to develop robust network models utilizing diverse communication technologies. The literature survey immensely helped to develop two novel SG application models, Zigbee based Pilot protection scheme for a smart distribution grid and Vehicle to Grid (V2G) smart load management scheme. Application modelling included detail traffic modelling, developing smart algorithms, analytical models, user load profile analysis, simulation models and test bed setups. Furthermore, a novel WiMax Ranging scheme is presented to improve the random-access mechanism for various periodic M2M applications supported by extensive simulation based performance analysis.

Future SGs will be overwhelmed by an excessive number of sensor devices that collect various data related to the power system. In a SG Neighborhood Area Network (NAN), wireless sensor networks (WSNs) will play a key role in the development of major SG applications. The application centric WSNs require complex configurations such as well-defined access techniques, transmission and security protocols. Challenges also include development of appropriate routing protocols to tackle resource limitations and delay caused by decentralized WSNs and ad hoc based packet forwarding techniques. A careful observation of manageability and controllability of the diverse M2M network reveals that the inadequate dynamic network configuration capability of the existing SG communication network would be a key bottleneck for future SG. Thus, a novel WSN based communication framework is presented exploiting the emerging SDN networking paradigm. SDN would be beneficial for SGs in many ways. By decoupling the control plane and data forwarding plane, SDN facilitates real-time control and integration of network services and applications that can reach down into the network through the controller hierarchy. A higher degree of control over the overall SG communication network would be achievable via the dynamic programmability provided by SDN. The SDN based WSN network must be robust enough to support the adaptive energy dispatching capacity of the modern power system. The proposed communication framework incorporates novel communication features to separate the control plane and data forwarding plane within the SG communication network. This includes detailed modeling of the control and data plane communication parameters to support both delay sensitive and delay tolerant SG applications. The unique SDN features offers a platform to accommodate maximum number of SG applications with highest controllability and manageability. The performance of the SDN based future SG network is evaluated using a simulation scenario that considers realistic user load profiles, wireless standards, the SG premises geographical area and the state of the art of the SG standards.

Although the control plane enables a global view of the data plane and provides a centralized platform to control and deploy new services, physically a single controller in the controller would not be practical for SG networks. The challenges arise in terms of scalability, security and reliability, particularly in a SG environment. To increase the efficiency of the proposed SDN based WSNs for the SG NAN, the study proposed distributed controllers with a comprehensive analytical model that optimizes the number of distributed controllers to enhance performance of the proposed communication framework in the NAN domain. The proposed framework along with the analytical model derive several solutions, such as the minimum number of controllers to support the switches and M2M devices, accommodate SG applications and a differentiated flow processing technique to support all traffic types within the network. Lastly, the study focuses on developing SDN-based application specific traffic models for the smart distribution grid.

The thesis focuses on three major issues while developing a future SG communication system. Firstly, its identifies major applications and their traffic requirements at different domains of the SG. Appropriate traffic models were developed by designing robust wireless communication network models. Also, application centric smart optimization techniques are adopted to achieve maximum performance and presented with simulation results, statistical analysis and a test bed result analysis. Secondly, to facilitate the centralized controllability and programmability for supporting diverse SG applications within the SG, a novel WSNs communication framework is presented exploiting the next generation SDN paradigm. Both delay sensitive and delay tolerant SG applications were considered based on the traffic requirement to develop the SDN based WSN communication framework in the SG NAN. Smart algorithms were developed at the SDN based WSN application layer to accommodate a large number of SG applications. The framework feasibility is demonstrated by the simulations carried out to verify the model and provide a statistical analysis. Thirdly, the thesis focuses on developing a novel analytical model that can be used to determine the optimal number of distributed controllers and switches in a SG NAN domain. The proposed application centric traffic modelling techniques, SDN based wireless communication framework and analytical models in this thesis can be adapted for research into other communication networks, particularly those that are begin developed for the Internet of Things and other forms of M2M communications. Also, due to the technology agonistic characteristics of the analytical and traffic models, they can be used in the development of various wireless networks, particularly those that focus on wireless sensor networks, more generally than the broader Internet of Things.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Engineering
Subjects Wireless Communications
Power and Energy Systems Engineering (excl. Renewable Power)
Keyword(s) Smart Grid Communication
Neighborhood area network
Electric vehicles
Software defined networking
Wireless Sensor Network
Smart load management
Versions
Version Filter Type
Access Statistics: 62 Abstract Views, 24 File Downloads  -  Detailed Statistics
Created: Wed, 08 Nov 2017, 13:16:16 EST by Denise Paciocco
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us