The Theory Of Rate Schedules In Wireless Energy Harvesting Communication Systems

The energy shortage problem of wireless devices is becoming more and more prominent.Energy harvesting from environments has been explored and implemented as an alternative to supplement or even replace batteries in modern wireless communication systems.One of the most important research concerns is to develop energy-efficient data transmission scheduling algorithms.Despite the abundant research on energy-efficient rate scheduling polices in energy har-vesting communication systems,there are still many shortcomings.First,most of the literature consider isolated point-to-point data transmission,and rarely discuss the exploitation of multi-task data sharing to further improve the efficiency of energy utilization.Second,most of the research literature on the energy harvesting communication systems assume that the battery ca-pacity of the wireless devices is infinite and the devices keep operating all the time during the transmission,while in real environment,wireless devices cannot store energy in an unrestricted manner,and they can actually enter the sleep state in order to save energy.To this end,this thesis studies optimization problems in the energy harvesting wireless communication systems.First,we consider rate schedules with data sharing for energy harvest-ing devices.We formulate the problem as a transmission completion time minimization problem under constraints of dynamical data requests and energy arrivals.We discover the relationship between two problems:the completion time minimization problem and the energy consump-tion minimization problem with a given completion time.We first derive the optimal algorithm for the min-energy problem,a decomposition method is proposed to solve this problem,and then adopt it as a building block to compute the optimal solution for the min-completion-time problem.Then,we study the rate schedules for energy harvesting communication systems with sleep states,where the sleep state and battery capacity of the wireless device is considered.This thesis considers a more realistic model with sleep-state where a transmitter can switch between operating state and sleeping state,and investigates the optimal rate scheduling policies for max-imizing the data throughput under dynamical energy arrivals.The energy needed to open the transmitter and the power needed to keep the transmitter operating are both taken into consid-eration in our model.We discover that the structure of the optimal solution in sleep-state model differs significantly from that of none-sleep-state model,e.g.,the existence of energy overflow and a critical rate.We develop an optimal algorithm that maximizes the data throughput based on figuring out the structure of the optimal solution,and the optimal rate scheduling strategy is constructed by the dynamic programming technique.Finally,for the online settings of the above two problems,we have developed efficient online heuristic real-time rate scheduling algorithms to tackle online energy arrivals,and simulation results further validate the effectiveness of the proposed algorithms.In this thesis,we study the rate scheduling problems in energy harvesting communication systems in terms of two aspects:data sharing and sleep states.Our work extends the rate sched-ules of energy harvesting wireless devices,and provides a further reference for the theoretical research and practical application in this field.