认知无线传感网的感知接入与资源分配研究

CORC > 沈阳自动化研究所 > 中国科学院沈阳自动化研究所 > 工业控制网络与系统研究室

题名	认知无线传感网的感知接入与资源分配研究
作者	许驰
学位类别	博士
答辩日期	2017-05-27
授予单位	中国科学院沈阳自动化研究所
授予地点	沈阳
导师	于海斌
关键词	认知无线传感网频谱感知频谱接入资源分配能量收集
其他题名	Research on Sensing-Access and Resource Allocation of Cognitive Radio Sensor Networks
学位专业	控制理论与控制工程
中文摘要	本文通过深入研究CRSN的感知接入和资源分配两大方面内容，分别实现了异构CRSN的实时频谱感知和连续频谱接入，以及层叠CRSN的持续能量供给和最优资源分配。具体来说，所进行的研究工作以及取得的创新性成果如下：（1）针对传统的同构CRSN架构下节点同时承担频谱感知和数据传输功能而无法进行连续数据传输的问题，结合CRSN的低成本、低功耗、分布式自组织、短距离多跳传输等特点，提出了全新的异构CRSN架构及其帧结构。该架构将频谱感知和数据传输功能分配给不同的异构网络节点，无需全网节点配备认知无线电收发机，在降低设备成本和网络能耗的同时，实现了实时的频谱感知和连续的数据传输，进而提高了感知可靠性和吞吐量。在此基础上，针对主用户动态变化导致的频谱接入不连续问题以及单节点感知可靠性低而影响自身和主用户通信的问题，提出了一种实时多信道协作频谱感知策略，使得CRSN可以维护一个连续可用的空闲授权信道集，并通过自适应信道切换实现对宽带授权频谱的连续可靠接入。然后，在充分保护主用户的前提下，以最大化吞吐量为目标，对网络节点数量和决策融合准则进行了联合优化，进一步提高了吞吐量和能效。（2）针对CRSN的频谱利用率低及频谱接入不连续问题，在充分考虑频谱的空域和时域特性的基础上，定义了自由机会和共享机会，建立了全新的二维空时机会模型。在此基础上，结合CRSN的低功耗、短距离多跳传输、时延受限等特点，提出了基于空域位置感知和时域频谱感知的空时联合频谱接入策略。该策略全面考虑了节点的硬件最大发射功率、节点对主用户的最大干扰功率以及主用户对节点的干扰，无需信道切换即可实现CRSN对单个授权信道的连续频谱接入，进而满足实时性应用。作为比较，考察了单纯的空域接入策略和时域接入策略。其中，在频谱共享和最大无干扰发射功率两种空域接入策略的基础上，提出了一种适合二维空时机会模型的混合空域接入策略。为了衡量网络的时延特性，分别推导了多跳CRSN在Rayleigh和Nakagami衰落信道下采用不同频谱接入策略时的端到端中断概率，并得到了其准确表达式。通过与其他策略进行比较，证明了空时联合接入策略具有更好的鲁棒性、抗干扰性和中断性能，可以有效提高频谱利用率。（3）针对CRSN的能量受限和时延受限问题，结合CRSN的低功耗、短距离多跳传输等特点，建立了基于无线充电技术的CRSN与主用户网络的层叠共存架构。通过部署专门发射能量信标的能量接入点，并根据信道状态信息和网络性能需求为网络节点进行按需无线充电，有效解决了网络的能量受限问题。这样，网络节点根据收集到的能量以及主用户的最大干扰功率约束，在授权信道与主用户进行并行传输。为了衡量网络的时延特性，推导并得到了多跳CRSN在Rayleigh衰落信道下的端到端中断概率的准确表达式和近似表达式，考察了中断饱和现象。在此基础上，以最小化端到端中断概率为目标，对无线充电功率和无线充电时间进行了联合资源分配。由于最优资源分配结果具有硬件不可实现性，故提出了一种迭代优化算法，获得了硬件可实现的近似最优资源分配结果。在不浪费网络资源的前提下，实现了网络的长期灵活部署。（4）针对CRSN的能量受限问题，结合CRSN的低功耗、短距离多跳传输等特点，构建了CRSN与主用户网络的绿色层叠共存架构。其中，CRSN以主用户发射的射频信号为能量源进行绿色能量收集，并在不影响主用户正常通信的前提下，接入授权信道与主用户进行并行传输。这样，CRSN可以同时利用主用户的能量和频谱资源，提升能效和谱效。受限于网络节点所收集到的能量及其对主用户的最大干扰功率，以最大化端到端吞吐量为目标，建立了关于多个节点的传输时间和发射功率的联合资源分配问题。考虑到问题的非凸性，经变量代换将原问题等价转换为关于传输时间和消耗能量的凸优化问题后，提出了基于迭代拉格朗日对偶分解的联合最优时间和功率分配算法，实现了最优的资源分配。
英文摘要	By deeply studying the two general aspects of sensing-access and resource allocation in CRSN, this paper realizes instantaneous spectrum sensing and continuous spectrum access for heterogeneous CRSN, and sustainable energy supply and optimal resource allocation for underlay CRSN. More specifically, the works carried out in this paper and their novelties are listed as follows: (1) For the conventional homogeneous CRSN that cannot perform continuous data transmission due to the nodes’ dual roles for both spectrum sensing and data transmission, we propose a novel heterogeneous CRSN architecture with a new frame structure by considering CRSN’s characteristics of low-cost, low-power, distributed and ad-hoc, short-range and multi-hop transmission. This architecture separates spectrum sensing and data transmission into different heterogeneous nodes, and does not need to equip every nodes with CR transmitter. In this way, both the devices cost and the energy consumption are decreased. Meanwhile, we achieve instantaneous spectrum sensing and continuous data transmission, which increase both sensing reliability and throughput. Then, to cope with the discontinuous spectrum access due to the dynamics of primary user (PU) and the impacts on CRSN and PU due to the low sensing reliability of a single node, we propose an instantaneous multi-band cooperative spectrum sensing scheme, by which the CRSN can hold an available idle licensed channel set and perform continuous and reliable spectrum access in the licensed wideband by adaptive channel switching. Then, with sufficient protection of PUs, we jointly optimize the number of nodes and the rule of decision fusion, which further increases throughput and energy efficiency. (2) For the low spectrum utilization and discontinuous spectrum access of CRSN, we define free opportunity and sharing opportunity, and formulate a novel two-dimensional spatial-temporal opportunity model by fully considering the spatial and temporal properties of spectrum. Then, taking into account CRSN’s characteristics of low-power, short-range and multi-hop transmission and delay constrained, we propose a joint spatial-temporal access scheme (JSTAS) based on spatial position sensing and temporal spectrum sensing. With full consideration of the hardware maximum transmit power of node, the peak interference power at multiple PUs and the interference of PU to node, JSTAS can achieve continuous spectrum access in a single licensed channel without channel switching, which can satisfy real-time applications. As comparisons, we evaluate pure spatial access scheme and temporal access scheme. Based on spectrum sharing and maximum interference-free transmit power schemes, we propose a hybrid spatial access scheme which suits the proposed opportunity model. To evaluate the delay characteristic, we derive the end-to-end outage probability for the multi-hop CRSN with different spectrum access schemes over Rayleigh and Nakagami fading channels, and obtain their exact expressions. By comparing these schemes, we collaborate that JSTAS is more robust, anti-interference and has better outage performance than other schemes, which sufficiently enhances spectrum efficiency. (3) For the energy constraint and the delay constraint of CRSN, we formulate an underlay coexistence paradigm for PUs and CRSN with wireless charging by considering the characteristics of low-power, short-range and multi-hop transmission. By deploying an energy access point that broadcasts power beacon and performs on-demand wireless charging according to channel states and network’s requestments, we can solve the energy-contrained problem effectively. According to the harvested energies of nodes and the peak interference power at PU, the nodes in CRSN perform concurrent transmission with PU in the licensed channel. To evaluate the delay characteristic, we derive and obtain the exact and asymptotic expressions for the end-to-end outage probability of multi-hop CRSN over Rayleigh fading. Furthermore, we investigate outage saturation phenomenon. Then, aiming at minimizing end-to-end outage probability, we jointly optimize the charging power and the charging time. As the optimal resource allocation cannot be implemented by hardwares, we propose an iterative algorithm which is hardware achievable and can realize near-optimal resource allocation. In this way, without resource waste, we realize the long-term and flexible deployments of CRSN. (4) For the energy constraint of CRSN, we formulate a green underlay coexistence paradigm for PUs and CRSN by considering the characteristics of low-power, short-range and multi-hop transmission. The nodes in CRSN harvest green energy from the RF signals of PUs and perform concurrent transmission with PU in the licensed channel on the promise that PUs’ normal communication is not influenced. In this way, CRSN utilizes both the energy and the spectrum of PU, which improves energy and spectrum effiencies. Then, subject to the harvested energies of nodes and the peak interference powers at PU, we formulate the end-to-end throughput maximization problem with respect to the transmission time and the transmit powers of multiple nodes. As the problem is nonconvex, we first convert it into an equivalent yet convex problem with respect to the transmission time and the consumed energies, and then propose a joint optimal time and power allocation algorithm based on the iterative Lagrange dual decomposition. In this way, we realize optimal resource allocation.
语种	中文
产权排序	1
内容类型	学位论文
源URL	[http://ir.sia.cn/handle/173321/20567]
专题	沈阳自动化研究所_工业控制网络与系统研究室
推荐引用方式 GB/T 7714	许驰. 认知无线传感网的感知接入与资源分配研究[D]. 沈阳. 中国科学院沈阳自动化研究所. 2017.