【作者】 梁爽；

【导师】 房至一；

【作者基本信息】 吉林大学 ， 计算机系统结构， 2022， 博士

【摘要】 随着无线通信系统的迅速发展,各类用户对系统的性能要求不断增加。在无线通信系统中采用天线阵列可以有效提升系统容量和信号增益,同时也可以显著提高通信能效。辐射方向图优化通常是天线阵列优化中的主要问题,然而,该问题是一个复杂的非线性优化问题,采用传统方法对其求解开销较大。进一步,将天线阵列引入到无线传感器网络(WSNs)中并采用协作波束成形可以大幅提高单个节点的通信距离和通信能效,然而,由于WSNs中节点随机分布导致的位置误差将产生较大的方向图旁瓣电平,进而影响系统通信性能。WSNs通信有效性面临的另一个重要问题是其节点能量受限,采用基于无人机(UAV)的无线充电技术对节点进行能量补充可以有效提升传统无线传感器网络的生命周期,然而,无人机本身也将面临能量受限问题,因此如何在基于UAV的无线可充电传感网络(WRSNs)中对充电无人机(CUAVs)进行调度至关重要。本文针对无线通信系统中天线阵列方向图优化、基于虚拟天线阵列的WSNs协作波束成形优化和WRSNs中CUAV调度优化三个方面的问题开展工作,并分别设计了适用于求解上述问题的仿生智能计算方法,主要学术贡献和创新如下:1.基于不同几何模型的天线阵列方向图优化方法(1)提出一种基于改进迁移和自适应变异策略的生物地理优化算法(BBOIMAM),用以解决直线形天线阵列(LAA)、圆形天线阵列(CAA)和随机形天线阵列(RAA)的方向图单目标优化问题。BBOIMAM算法通过引入广义正弦迁移模型、精英学习迁移策略和基于弹簧振动的自适应变异策略改善了标准生物地理优化算法(BBO)的搜索性能,标准测试函数集CEC2017和CEC2020实验验证了提出算法的有效性;其次,基于BBOIMAM设计不同几何模型天线阵列中阵元激励电流优化方法,用以抑制不同天线阵列产生波形图的最大旁瓣电平;最后,仿真实验结果证明了该算法的有效性。(2)构建以优化天线阵列旁瓣电平和零陷为目标的多目标方向图优化问题MBPOP,并提出一个改进的多目标差分进化算法(IMOEAD),用以解决构建的优化问题。该算法通过引入正态分布交叉(NDX)、Lévy飞行和最优解选择机制用以改善算法的多样性和全局搜索能力,并通过标准多目标测试函数集ZDT测试验证了提出算法的有效性。最后,给出基于IMOEAD算法的LAA旁瓣电平抑制和零陷控制的多目标优化方法,及其有效性验证。2.WSNs中基于虚拟天线阵列的协作波束成形性能优化方法(1)针对传感器节点静止的WSNs,提出一种基于虚拟圆环天线阵列(CCAA)的联合旁瓣抑制方法JSSA。该方法首先给出基于虚拟CCAA的能量最优组阵节点选择方法;然后,基于改进的鸡群优化(VPCSO)算法,给出以圆环天线阵列为引导阵列的阵元激励电流优化方法,用以抑制虚拟天线阵列协作波束成形的最大旁瓣电平。其中VPCSO算法将位置学习机制和变异机制引入到传统的鸡群优化(CSO)算法中,前者用以提高算法的局部搜索能力,后者用以提高算法的全局搜索能力;最后,给出基于VPCSO算法的JSSA的性能、能耗分析,以及电磁仿真分析。(2)针对节点可移动的无线传感器网络(MWSNs),首先构建以同时减少MWSNs中分布式协作波束成形(DCB)节点的最大旁瓣电平(SLL)、激励电流值和移动能耗为目标的DCB多目标联合优化问题;其次,提出一种分布式并行布谷鸟搜索算法(DPCSA),该算法引入了集群分布、并行计算和全局协作更新机制,并通过标准测试函数集CEC2017验证了所提算法的有效性;最后,给出基于DPCSA的DCB联合优化问题优化方法,并验证了其有效性和稳定性。3.WRSNs中充电UAV充电性能优化方法提出WRSNs充电效率优化问题以延长WSNs的生命周期。首先,构建CUAVs部署优化问题CUAVDOP,该问题以联合增加CUAVs充电范围内的传感器节点数量,最大化传感器节点和CUAVs之间的充电效率,并最小化CUAVs的总飞行能耗为目标;其次,提出一种改进的萤火虫算法(IFA),该算法引入基于对立的学习模型、吸引模型和自适应步长算子,使其更适合于求解提出的优化问题。进一步,给出基于IFA的联合优化问题求解方法;最后,给出使用CUAVs进行无线可充电传感器网络充电的最大充电距离和应用场景分析。更多还原

【Abstract】 With the wireless communication system developed rapidly,the performance requirements of all kinds of users are increasing.The application of array antenna in wireless communication system can not only improve the system capacity and gain effectively,but also improve the communication energy efficiency significantly.Among array antenna optimization problems,the radiation pattern optimization is usually considered as the main problem.However,due to this problem is a complex nonlinear optimization problem,a large overhead will be incurred during the process to solve that by traditional methods.The introduction of array antenna and collaborative beamforming in wireless sensor networks(WSNs)can greatly improve communication distance and energy efficiency of a single node.However,the location errors caused by the random distribution of nodes in WSNs,will generate the high sidelobe level(SLL)of the beam pattern,and then directly affects the communication performance of the system.In addition,energy limitation of sensor nodes is also a significant problem of WSNs.Wireless charging technology based on unmanned aerial vehicle(UAV)can supplement the energy of nodes and effectively improve the life cycle of traditional WSNs.However,UAV itself will also face the problem about limited energy,thus,how to schedule the charging UAV(CUAV)is the key issue in the rechargeable wireless sensor networks(WRSNs)based on UAV.Therefore,this paper focuses on three problems: array antenna pattern optimization in wireless communication system,collaborative beamforming(CB)optimization in WSNs based on virtual array antenna and CUAV scheduling optimization in WRSNs.Moreover,the bio-inspired computing methods suitable for solving the above problems are designed respectively.The main contributions and innovations are as follows:1.Array antenna pattern optimization method based on different geometric models(1)A biogeography-based optimization based on improved migration and adaptive mutation strategy(BBOIMAM)is proposed to solve the single objective optimization problems on pattern of linear antenna array(LAA),circular antenna array(CAA)and random antenna array(RAA).BBOIMAM algorithm improves the local and global search ability of standard biogeography-based optimization algorithm(BBO)by introducing generalized sinusoidal migration model,elite learning migration strategy and adaptive mutation strategy based on spring vibration.The effectiveness of the proposed algorithm is verified by the experiments on CEC2017 and CEC2020 standard test function sets.Secondly,based on BBOIMAM,the optimization methods of array element excitation current in antenna array of different geometric models are designed to suppress the maximum sidelobe level(SLL)of different antenna arrays beam pattern.The experimental results show the effectiveness of the algorithm.(2)A multi-objective pattern optimization problem MBPOP aiming at optimizing the SLL and NULL of antenna array is constructed,and an improved multi-objective evolutionary algorithm based on decomposition(IMOEAD)is proposed to solve the proposed optimization problem.The diversity and global search ability of the standard algorithm are improved by introducing normal distribution crossover(NDX),Lévy flight and optimal solution selection mechanism.The effectiveness of the proposed algorithm is verified by the standard multi-objective test function set ZDT.Finally,the multi-objective optimization method on SLL suppression and NULL control of LAA based on IMOEAD algorithm is given,and its effectiveness is also verified.2.Performance optimization method of CB in WSNs based on virtual antenna array(1)For WSNs with static sensor nodes,a joint SLL suppression method JSSA based on virtual concentric circles antenna array(CCAA)is proposed.First,JSSA gives the calculation method for the location of the energy optimal array nodes.Second,the method shows that how to use the CCAA as the guide array to select the nodes of the VNAA.Third,an improved chicken swarm optimization(VPCSO)algorithm is proposed to optimize the excitation current of the selected array nodes.VPCSO algorithm introduces location learning mechanism and mutation mechanism into the traditional chicken swarm optimization(CSO)algorithm.The former is used to improve the local search ability,and the latter is used to improve the global search ability.Finally,the performance,energy consumption analysis and electromagnetic simulation analysis of JSSA based on VPCSO algorithm are given.(2)For WSNs with mobile nodes,a distributed collaborative beamforming(DCB)multi-objective joint optimization problem is constructed to optimizes the maximum SLL,excitation current and mobile energy consumption of DCB nodes,simultaneously.Secondly,a distributed parallel cuckoo search algorithm(DPCSA)is also proposed.The algorithm introduces cluster distribution,parallel computing and global cooperative update mechanism,and then the effectiveness of the proposed algorithm is verified by the CEC2017 standard test function set.Finally,the optimization method of DCB joint optimization problem based on DPCSA,and its effectiveness and stability verification are given.3.Charging performance optimization method of CUAV in WRSNAiming at the charging efficiency optimization in WRSN,the deployment optimization problem of CUAVS(CUAVDOP)is constructed.The goal of CUAVDOP is to jointly increase the number of sensor nodes within the charging range of CUAVS,maximize the charging efficiency between sensor nodes and CUAVS,and minimize the total motion energy consumption of CUAVS.Secondly,a new improved firefly algorithm(IFA)is proposed.The algorithm introduces opposition-based learning model,attraction model and adaptive step operator to make it more suitable for solving the proposed optimization problem.Thirdly,the optimization method of CUAVDOP based on IFA is given.Finally,the analysis about maximum charging distance and application scenario of charging with CUAVS in WRSN are given.更多还原