【作者】 刘强；

【导师】 孙松林；

【作者基本信息】 北京邮电大学 ， 信息与通信工程， 2022， 博士

【摘要】 随着物联网的快速发展,物联网设备的成本效率、能源效率和频谱效率将成为制约物联网设备大规模普及的关键因素。环境后向散射通信(Ambient backscatter communication,AmBC)由于其低成本低功耗的特点受到了学术界和工业界的广泛关注。在AmBC系统中,后向散射设备(Backscatter Device,BD)无需主动射频(Radio Frequency,RF)链路和电池供电便可以将信号调制到环境RF信号上,这将极大地促进低成本低功耗物联网的发展。但AmBC的无线传播环境一般是不可控性的,这将影响AmBC可达和速率的提升。智能反射面(Intelligent Reflecting Surface,IRS)可以通过软件控制的反射单元来重新配置无线传播环境而受到了广泛的关注。因此,利用IRS对AmBC系统的无线环境进行调整以增强AmBC系统的性能是一个潜在的研究方向。本文的主要目标是通过IRS对AmBC系统的无线传播环境进行重新配置以增大AmBC系统可达和速率。本文的研究主要针对以下三个场景进行展开。首先,本文对无IRS辅助的AmBC系统进行资源分配以最大化AmBC系统可达和速率,为后续IRS对AmBC系统的增强提供基础的对比。然后,本文考虑了一种特殊情况下的IRS辅助AmBC场景,在这个场景中IRS与RF源之间存在遮挡,所以IRS需要对AmBC系统的反射链路进行增强以最大化AmBC系统的可达和速率。最后,本文考虑更为一般情况下的IRS辅助AmBC场景,在这个场景中所有链路不存在遮挡,所以IRS对AmBC系统进行增强时需要同时考虑对反射链路和入射链路的增强。针对上述三种场景,本文的主要工作和贡献如下:1)本文考虑无IRS辅助的AmBC场景,提出了一种基于非正交多址的AmBC系统资源分配方案以最大化AmBC系统的可达和速率。在多BD场景下,为了最大化AmBC系统的可达和速率,本文提出了基于BD分组数和BD分配方案、BD反射系数和BD解码顺序的联合优化问题。为了解决上述优化问题,本文首先提出了静态BD分组优化方案,然后基于静态BD分组优化方案提出了低复杂度的动态BD分组优化方案。在所提出的静态BD分组优化方案中,本文给出了 BD最优解码顺序方案和BD最优反射系数的闭式解。然后,本文在静态分组优化方案的基础上提出了动态BD分组优化方案,并给出了最优BD分组数和最优BD分配方案。最后,仿真证明了所提动态BD分组优化方案优于基准方案的性能。2)本文考虑特殊情况下的IRS辅助的AmBC场景,即IRS与RF源之间存在遮挡的情况。为了最大化AmBC系统的可达和速率,本文提出了 IRS辅助反射链路增强的AmBC系统模型,形成了基于IRS相移、BD反射系数和BD解码顺序的联合优化问题。在单BD AmBC场景中,本文优先给出了令AmBC系统的反射链路增益最大的IRS相移,然后根据给定的IRS相移给出了 BD的最优反射系数。所提方案不需要求解半正定规划问题,因此具有低计算复杂度的特点。在多BD AmBC场景中,为了保障所提方案的稳定性,本文首先提出了基于半正定松弛的多BDIRS相移的初始化方案。然后,基于给定的IRS相移给出了多BD反射系数优化方案和基于SDR的多BD IRS相移优化方案。随后,通过交替执行多BD反射系数优化方案和多BDIRS相移优化方案以获得最优的AmBC系统可达和速率。最后,仿真证明了所提IRS辅助反射链路增强NOMA-AmBC方案优于无IRS辅助NOMA-AmBC 方案。3)本文考虑更为一般情况下的IRS辅助的AmBC场景,在这个场景中所有链路都不存在遮挡。IRS对AmBC系统进行增强时,需要同时考虑对反射链路和入射链路的增强。为了最大化AmBC系统的可达和速率,本文提出了 IRS辅助双射链路增强AmBC系统模型,形成了基于IRS相移、BD反射系数和BD解码顺序的联合优化问题。为了解决上述优化问题,本文首先给出了单BD AmBC场景的解决方案,然后基于单BD AmBC场景的解决方案给出了多BD AmBC场景的解决方案。首先,本文引入了 RF-BD链路增益因子对AmBC的入射链路进行松弛,主要目的是在保证入射链路有一定增益的情况下最大化反射链路增益。然后,基于不同的RF-BD链路增益因子给出了BD反射系数和IRS相移的优化解。最后,给出令AmBC系统可达和速率最大的BD反射系数和IRS相移。与单BD AmBC解决方案不同的是,多BD AmBC解决方案需要考虑基于NOMA的多用户解调的约束条件对BD反射系数进行调整。仿真证明了所提的IRS辅助双射链路增强NOMA-AmBC解决方案优于IRS辅助反射链路增强NOMA-AmBC解决方案和无IRS辅助NOMA-AmBC解决方案。更多还原

【Abstract】 With the rapid development of the Internet of Things(IoT),the cost efficiency,spectrum efficiency and energy efficiency of IoT devices will become the key factors restricting the large-scale popularization of IoT devices.Ambient backscatter communication(AmBC)has received extensive attention from academia and industry due to its low cost and low power consumption.In the AmBC system,the backscatter device(BD)can modulate the signal to the ambient RF signal without an active radio frequency(RF)link and battery power,which will greatly facilitate the development of low-cost and low-power IoT.However,the wireless propagation environment of AmBC is generally uncontrollable,which will affect the performance of AmBC.Intelligent reflecting surface(IRS)has received extensive attention by reconfiguring the wireless propagation environment through software-controlled reflective elements.Therefore,using the IRS to adjust the wireless environment of the AmBC system to enhance the performance of the AmBC system is a potential research direction.The main goal of this thesis is to reconfigure the wireless propagation environment of AmBC through the IRS to increase the achievable sum rate of the AmBC system.The research in this thesis mainly focuses on the following three scenarios.First,this thesis allocates resources to the AmBC system without IRS assistance to maximize the achievable sum rate of the AmBC system,and provides a basic comparison for the subsequent enhancement of the AmBC system by the IRS.Then,this thesis considers an IRS-assisted AmBC scenario in a special case,in which there is occlusion between the IRS and the ambient RF source,so the IRS needs to enhance the reflection link of the AmBC system to maximize the achievable sum rate of the AmBC system.Finally,this thesis considers the more general IRS-assisted AmBC scenario.In this scenario,all links are not blocked.Therefore,when the IRS enhances the AmBC system,it is necessary to consider both the enhancement of the reflection link and the incident link.For the above three scenarios,the main work and contributions of this thesis are as follows:1)This thesis considers the AmBC scenario without IRS assistance,and proposes a resource allocation scheme for the AmBC system based on non-orthogonal multiple access(NOMA)to maximize the achievable sum rate of the AmBC system.In the multi-BD scenario,in order to maximize the achievable sum rate of the AmBC system,this thesis proposes a joint optimization problem based on the number of BD grpups,the BD allocation scheme,the BD reflection coefficient and the BD decoding order.In order to solve the above optimization problems,this thesis first proposes a static BD grouping optimization scheme,and then proposes a lowcomplexity dynamic BD grouping optimization scheme based on the static BD grouping optimization scheme.In the proposed static BD grouping optimization scheme,we give the optimal BD decoding order scheme and the closed-form solutions for BD reflection coefficient.Then,this thesis proposes a dynamic BD grouping optimization scheme based on the static grouping optimization scheme,and gives the optimal BD grouping number and optimal BD allocation scheme.Finally,simulations demonstrate that the proposed dynamic BD grouping optimization scheme outperforms the benchmark scheme.2)This thesis considers an IRS-assisted AmBC scenario in a special case,where there is occlusion between the IRS and the RF source.The IRS needs to enhance the reflection link of the AmBC system to maximize the achievable sum rate of the AmBC system.Therefore,this thesis proposes an IRS-assisted reflection link enhanced AmBC system model,and forms a joint optimization problem based on IRS phase shift,BD reflection coefficient and BD decoding order to maximize the achievable sum rate of the AmBC system.In the single-BD AmBC scenario,this thesis first gives the IRS phase shift that maximizes the gain of the reflection link of the AmBC system,and then gives the optimal reflection coefficient of the BD according to the given IRS phase shift.The proposed scheme does not need to solve semi-definite programming(SDP)problems,so it has the characteristics of low computational complexity.In the multi-BD AmBC scenario,in order to ensure the stability of the proposed scheme,this thesis first proposes a multi-BD IRS phase shift initialization strategy based on semi-definite relaxation(SDR).Then,based on the given IRS phase shift,a multi-BD reflection coefficient optimization strategy and an SDR-based multi-BD IRS phase shift optimization strategy are given.Subsequently,the optimal achievable sum rate of the AmBC system are obtained by alternately executing the multi-BD reflection coefficient optimization strategy and the multi-BD IRS phase-shift optimization strategy.Finally,simulations show that the proposed NOMA-AmBC scheme with IRSassisted reflection link enhancement outperforms the NOMA-AmBC scheme without IRS assistance.3)This thesis considers the more general IRS-assisted AmBC scenario,in which there is no occlusion on each link.When the IRS enhances the AmBC system,it is necessary to consider the enhancement of the reflection link and the incident link at the same time.In order to maximize the achievable sum rate of the AmBC system,this thesis proposes an IRS-assisted dual-radio link enhanced AmBC system model,which forms a joint optimization problem based on IRS phase shift,BD reflection coefficient and BD decoding order.In order to solve the above optimization problem,the solution of single BD AmBC scenario is given first,and then the solution of multi-BD AmBC scenario is given based on the solution of single BD AmBC scenario.First,this thesis introduces the RF-BD link gain factor to relax the incident link of AmBC,the main purpose is to maximize the gain of the reflection link while ensuring that the incident link has a certain gain.Then,the optimal BD reflection coefficient and IRS phase shift are given based on different RF-BD link gain factors.Finally,the optimal BD reflection coefficient and IRS phase shift that maximizes the achievable sum rate of the AmBC system is given.Different from the single-BD AmBC solution,the multi-BD AmBC solution needs to adjust the BD reflection coefficient considering the constraints of NOMA-based multi-BD decoding.Simulations demonstrate that the proposed IRS-assisted dual-radio link enhanced NOMA-AmBC solution is superior to the IRS-assisted reflection link enhanced NOMAAmBC solution and the non-IRS-assisted NOMA-AmBC solution.更多还原