Research On Power Allocation Problem In Cognitive Relay Networks With Energy Harvesting

With the development of society and the advancement of technology,wireless mobile communication technology has evolved from the original public mobile telephone system to the fifth generation(5G)mobile communication that supports massive device connections.The biggest impetus for the development of wireless communication technology is people's constant pursuit of higher data transmission rate.In the future,aiming at higher data transmission rate,wireless communication technology will be integrated with internet and IP technology to provide users with personalized communication services with better quality.The development of wireless mobile communication technology brings not only higher data transmission rate,but also wider frequency spectrum requirement,which brings severe challenges to the already scarce radio spectrum resource.At present,most of the radio spectrum resources have been allocated to some authorized services,and the remaining small part can not be effectively used because of the characteristics of large span and fragmented distribution.Cognitive radio(CR)technology alleviates the contradiction between the scarce spectrum resource and the increasing data traffic through re-developing and utilizing the licensed spectrum resource,which effectively improves the utilization of spectrum.CR networks(CRNs)intelligently observe and learn the surrounding radio environment,and dynamically adjust the transmission parameters(such as modulation parameters,transmit power)of secondary user(SU)based on the perceived spectrum occupancy information and other radio environment characteristics,so that SU can share the authorized spectrum resource on the premise of ensuring the communication quality requirements of primary user(PU).In order to meet the quality of service(QoS)requirements of PU in CRNs and achieve the purpose of spectrum sharing,it is necessary to reasonably control the transmit power of SU to avoid harmful interference to PU.Cooperative relay communication technology constructs independent signal paths,which effectively expands the network coverage while increasing the spatial diversity gain.The application of cooperative relay technology in CRNs can solve the problem of quality degradation or even interruption of communications caused by channel fading,which improves the communication performance of the secondary networks.However,the additional power consumptions generated by the increased relay nodes also increase the energy burden of the cognitive relay network.Starting from the concept of green communications,energy harvesting(EH)technology collects energy from the surrounding environment(e.g.,solar,hydro,wind,and tidal energy)through specific devices,which alleviates the contradiction between energy consumption and energy supply.Based on the above research background,this thesis proposes several power allocation algorithms in EH-based cognitive relay networks with the consideration of spectrum sensing information uncertainties,mobile relay and channel state information(CSI)uncertainties.The major contributions of this thesis are as follows:(1)In practical communication scenarios of CRNs,due to the influence of feedback delay,channel fading and accuracy of the spectrum sensing algorithms,it is difficult for SU to obtain accurate spectrum sensing information.In view of the imperfect spectrum sensing,this thesis adopts orthogonal frequency division multiplexing(OFDM)and natural EH technologies to investigate the power allocation problem in cognitive relay networks with spectrum sensing uncertainties.First,based on overlay spectrum sharing scheme,this thesis mathematically describes the uncertain spectrum sensing information as the mis-detection probability and false alarm probability,and re-establish the interference model between PU and SU by considering the possible co-channel interference and side-channel interference.Then,the independent power control is performed on the SU transmitter and the relay node that applies the natural EH scheme.In this thesis,the power allocation algorithms with the perfect spectrum sensing and the imperfect spectrum sensing are proposed respectively to maximize the throughput of the secondary system under the constraints of the maximum transmit power of SU,the energy causality at relay,the energy storage of battery while ensuring that the interference power to PU does not exceed a certain threshold.Simulation results verify the feasibility and effectiveness of the proposed algorithm,and demonstrate that the power allocation algorithm with spectrum sensing uncertainties protects the normal communication of PU at the expense of throughput loss of the secondary system.(2)Aiming at the mobility limitation of fixed-position relay nodes(i.e.,static relay nodes)in traditional cognitive relay networks,this thesis studies the power allocation problem of cognitive mobile relay networks with EH.As for the aforementioned system model of cognitive relay networks,we make modifications from the following aspects: 1)The underlay spectrum sharing model and the natural EH process at the SU transmitter are taken into consideration.2)Replace the traditional fixed-position relay with a miniaturized unmanned aerial vehicle(UAV)with the function of relaying.3)With the consideration of the perfect CSI and accurate spectrum sensing information,a power allocation algorithm for UAV-assisted cognitive mobile relay networks is proposed.Based on the modified system model of cognitive mobile relay networks,this thesis describes four communication scenarios(i.e.,four trajectories of UAV): straight trajectory towards SU receiver,straight trajectory towards SU transmitter,hovering over the midpoint of the communication link of SU and circular trajectory.The simulation experiments and theoretical analysis of the proposed algorithm are carried out with the above four communication scenarios.Simulation results show that the proposed power allocation algorithm in EH-based cognitive mobile relay network can effectively improve the throughput performance of SU while guaranteeing the communication quality requirements of PU.(3)Considering the unavoidable CSI uncertainties in actual wireless communications,this thesis investigates a robust power allocation algorithm in EH-based cognitive relay networks.A new EH scheme that combines natural energy and radio frequency(RF)energy is designed in underlay OFDM-based cognitive relay networks.This thesis formulates a robust power allocation problem with the imperfect CSI where the ellipsoid uncertainty set and the interval uncertainty set are applied to describe the uncertain parameters.According to the principle of throughput maximization of decode and forward protocol,the original multi-variable optimization problem is transformed into a single variable optimization problem.Considering the worst case of parameter estimation,the original semi-infinite programming(SIP)robust power allocation problem is transformed into a deterministic problem with limited constraints.The optimal power allocations of the secondary system are obtained by using the Lagrange dual decomposition method and the robust optimization method.Simulation results analyze the influence of uncertain channel parameters on the communication performance of the secondary system.Comparing with the non-robust power allocation algorithm with the perfect CSI,the robust power allocation algorithm under the imperfect CSI proposed in this section strictly satisfies the interference temperature constraint of PU at the cost of less throughput loss of the secondary system,and avoids harmful interference to PU.