Reliability analysis and performance optimization of wireless networks and its applications

• Main Author: Dong, Qiumin
• Other Authors: School of Computer Engineering
• Format: Theses and Dissertations
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Wireless communication systems; DRNTU::Engineering::Electrical and electronic engineering::Electric power::Auxiliaries, applications and electric industries
• Online Access: https://hdl.handle.net/10356/60750
• Institution: Nanyang Technological University

Nowadays, as unlicensed spectrum becomes a scarce resource, the concept of cognitive radio is proposed as a feasible solution. In a cognitive radio network, the channels that belong to licensees, also named as primary users, can only be accessed by secondary users (i.e., unlicensed users) opportunistically when they are not occupied by primary users. As cognitive radio is utilized to exploit the potential capacity of spectrum, the analysis of reliability and quality of service regarding the packet transmission of secondary users arises as a critical issue. However, due to estimation error, noise or system state fluctuation, parameters such as channel state (i.e., occupied or idle), channel quality and so on may be uncertain to secondary users. As the first issue studied in our research, the quality of service measures of a cognitive radio network are evaluated with the assumption of uncertain parameters so that the measures in the worst-case and best-case can be obtained and used to analyze the reliability of the cognitive radio network. Dynamic spectrum access of secondary users is arranged according to medium access control protocols. A two-dimensional Markov chain is applied to model the packet transmission of secondary users. A robust optimization method that can solve the Markov chain with uncertainty is utilized to obtain the lower and upper bounds of performance measures. On the other hand, as information and communication technologies are implemented in smart grid and other systems. Cognitive radio networks and other wireless networks are profoundly penetrated into such systems. The impact of unreliable or imperfect wireless networks on the performance of power usage behavior in smart grid has merit of evaluation. Therefore, the second issue in our research is the deferrable load scheduling under imperfect data communication. Given that the real-time pricing is implemented to realize the demand response in smart grid, a deferrable load can be scheduled optimally so that the power consumption cost will be minimized. However, if the price information is not available occasionally due to imperfect data communication or cyber attacks, the scheduling performance will be compromised. To overcome this problem, a partially observable Markov decision process based deferrable load scheduling algorithm together with the implementation of a standby alternative channel is proposed to improve the reliability of the data communication in smart grid. The numerical results show that the scheduling performance is improved in the case when the scheduler lacks of actual price information. Moreover, the impact of imperfect data communication on plug-in hybrid electric vehicle charging scheduling performance of a charging station is quantitatively analyzed in the last part of our research. The scheduling objectives are to minimize the power consumption cost incurred by charging batteries and to allow as many as possible plug-in hybrid electric vehicles to be served. A battery replacement strategy is adopted to save drivers' time. The numerical results show that if there are backup channels that can also be used to receive price information, the degraded scheduling performance due to imperfect data communication can be improved. On the other hand, a multi-stage stochastic programming model is formulated to solve the power supply cost optimization problem for the grid operator. Renewable source generation as well as utilization of energy storage device are integrated into the supply optimization problem. In summary, this dissertation investigates the upper bounds and lower bounds of quality of service measures for the dynamic spectrum access of secondary users in a cognitive radio network. The impact of imperfect data communication channel on the performance of deferrable load scheduling and plug-in hybrid electric vehicle charging scheduling is analyzed. Some methods for mitigating such negative impacts are proposed.