Signal processing for distributed nodes in smart networks

Abstract

With increasing environmental concern for energy conservation and mitigating climate change, next generation smart networks are bound to provide improved performance in terms of security, reliability, and energy efficiency. For instance, future smart networks will work in highly complex and dynamic environments and will have distributed nodes that need to interact with each other and may also interact with an energy provider in order to improve their performance. In this context, advanced signal processing tools such as game theory and distributed transmit beamforming can yield tremendous performance gains in terms of energy efficiency for demand management and signal trans-mission in smart networks. The central theme of this dissertation is the modeling of energy usage behavior of self-seeking distributed nodes in smart networks. The thesis mainly looks into two key areas of smart networks: 1) smart grid networks and 2) wireless sensor networks, and contains: an analytical framework of the economics of electric vehicle charging in smart grids in an energy constrained environment; a study of a consumer-centric energy management scheme for encouraging the consumers in a smart grid to voluntarily take part in energy management; an outage management scheme for efficiently curtailing energy from the consumers in smart grids in the event of a power outage; a comprehensive study of power control of sensors in a wireless sensor network using game theory and distributed transmit beamforming; and finally, an energy aware distributed transmit beamfoming technique for long distance signal transmission in a wireless sensor network. This thesis addresses the challenges of modeling the energy usage behavior of distributed nodes through studying the propriety of energy users in smart networks, 1) by capturing the interactions between the energy users and energy provider in smart grids using non-cooperative Stackelberg and generalized Nash games, and showing that the socially optimal energy management for users can be achieved at the solution of the games, and 2) by studying the power control of sensors in wireless sensor networks, using a non-cooperative Nash game and distributed transmit beamforming that demonstrates significant transmit energy savings for the sensors. To foster energy efficient transmission, the thesis also studies a distributed transmit beamforming technique that does not require any channel state information for long distance signal transmission in sensor networks. The contributions of this dissertation are enhanced by proposing suitable system models and appropriate signal processing techniques. These models and techniques can capture the different cost-benefit tradeoffs that exist in these networks. All the proposed schemes in this dissertation are shown to have significant performance improvement when compared with existing solutions. The work in this thesis demonstrates that modeling power usage behavior of distributed nodes in smart networks is both possible and beneficial for increasing the energy efficiency of these networks.