Vehicle Assisted Energy-Efficient Data Dissemination Framework
One or more files will be made publicly available from 2023-05-24.
MetadataShow full metadata
One of the main issues of maintaining a smart city (SC) is to handle a huge volume of data generated from various data sources and to transport them to the central control units for further processing and analysis. These data sources incur not only high running cost of core networks but also results in overburdening of such networks. The high energy utilization and carbon discharges from these core networks posing serious environmental issues. Therefore, in this thesis, to overcome the issues of high energy consumption (EC) and carbon emission (CE) of congested core networks, an energy-efficient vehicle assisted data dissemination (VADD) framework to tackle big data traffic is being proposed. In the investigation, different delay tolerant data forwarding case scenarios are designed and analysed by analytical modelling and data-driven approaches. To illustrate the efficiency of our proposed framework, we present mathematical models and data forwarding schemes for Auckland City (New Zealand) and Beijing City (China) case scenarios to transfer delay tolerant data from various data sources to a control centre. One of the motivations for designing a sustainable data dissemination framework is to efficiently utilize vehicular mobility to offload data from traditional networks. However, the main contributions of this thesis are highlighted below. First, we explore the potential of the VADD framework for data dissemination, by considering the annual average daily traffic of vehicles in Auckland city. Second, we analyse the network coverage and capacity of the proposed framework by considering the displacement of vehicles in Beijing City. Third, we propose a mathematical model for data transmission, numerically analyse the data forwarding strategies for the Auckland case by considering the vehicle mobility with Poisson distribution and with different parameters (e.g effective distance, deliver probability, and vehicle density). Fourth, we use the data-driven approach to transmit an updated software code to a massive number of smart sensors in an SC by our proposed framework. For this purpose an optimal code forwarding scheme (OCFS) is presented, to select the optimal vehicle and optimal roadside unit (RSU) for code transmission. Finally, we develop an EC model for our proposed framework and use different optimization techniques to prove that our proposed framework is energy efficient, as compared to traditional networks in terms of EC and CE.