【作者】 黄高勇；

【导师】 方旭明；

【作者基本信息】 西南交通大学 ， 通信与信息系统， 2014， 博士

【摘要】 正交频分复用(OFDM)/正交频分复用多址(OFDMA)和中继(Relay)技术相结合的多跳中继网络由于具有高频谱效率、增强覆盖、组网灵活,布网成本低等众多优点,已成为业界公认的下一代无线网络的主流架构。与传统蜂窝网络相比,基于OFDM/OFDMA技术的中继网络面对的环境更加复杂,使得无线资源管理问题变得更具挑战性：一方面,资源分配的维度由传统的时域、频域或空域向多维空间联合分配的方向发展,传统的资源分配策略无法直接应用到中继网络,且造成资源分配的算法复杂度和系统的信令开销增大；另一方面,无线数据业务的快速增长和大量低功率中继站点的引入使得系统功率消耗大大增加,基站能耗占所有设备能耗的一半以上,然而,过去大部分研究都是关注如何提高系统的吞吐率。随着近年“绿色通信”的提出,如何在保证用户服务质量(QoS)前提下,提高系统的能效,进而达到节能的目的,业已成为当前无线通信技术发展新的趋势。为此,本文以基于OFDM/OFDMA技术的无线多跳中继网络为研究对象,以系统节能为主要目标,研究路径选择、分组调度和资源分配等无线资源管理关键技术。首先,针对放大转发(AF)和解码转发(DF)的OFDM中继系统,分别提出一种能效最大化路径选择策略。针对三节点中继OFDM网络,在分析电路功率消耗、用户速率需求、带宽等因素在路径选择时对系统能效的影响基础上,通过将中继链路衰落值和直传链路衰落值转化为等效路径损耗值,分别提出一种基于等效路径损耗指数的能效最优路径选择准则。分析结果表明,能效最优传输路径取决于用户速率需求、链路信道条件以及系统带宽等因素,并满足本文所提出的能效最优路径判决准则。在此基础上,将三节点DF中继传输场景扩展到多用户多中继OFDMA蜂窝网下行传输系统,考虑电路功率消耗和固定用户速率需求的条件下,以能效最大化为目标,提出-种联合路径选择、子载波和功率分配的低复杂度次优算法。分析结果表明,与传统路径选择策略相比,所提路径选择方案能够显著提升系统能效。其次,针对三节点DF模式中继OFDM传输链路,提出一种系统总发射功率和用户最小速率需求受约束条件下子载波配对和功率分配的能效最优联合优化方案。该方案考虑了电路功率消耗对系统能效的影响,由于该问题属于非凸优化问题,直接求解复杂度高且难以实现,为此,提出一种分情况讨论求解策略,将非凸优化问题转化为凸优化问题求解,并证明了系统能效最优解的唯一性,再利用拉格朗日乘子法和KKT条件,通过注水算法获得系统能效最优解和功率分配结果。分析结果表明,与固定子载波配对能效最优功率分配策略相比,所提方案可以进一步提升链路可达速率和能效性能；与传统的速率最大化和功率最小化注水最优算法相比,所提方案可以根据系统总发射功率和用户速率需求约束条件自适应分配资源,以较低复杂度得到全局最优解,保证链路传输的可靠性。再次,考虑更为实际的源节点和中继节点发射功率各自约束条件,提出一种子载波配对和功率分配的联合优化方案。该方案针对三节点DF模式中继OFDM传输系统,同时考虑电路功率消耗和用户最小速率需求,以能效最大化为目标,建立联合子载波配对和功率分配的优化模型,采用分步求解策略,将非凸优化问题转化为凸优化问题,利用注水策略获得能效最优解。分析结果表明,所提方案能够在满足各项约束条件情况下自适应分配资源,达到系统能效最优,同时还可以保证链路传输可靠性,降低链路的中断概率和算法实现复杂度。接着,针对两跳中继OFDMA蜂窝网下行传输系统,提出一种自适应关联节能分组调度方案。该方案考虑不同类型业务QoS需求的差异性,用户分组队列状态,以及链路能效等综合因素设计调度优先级因子,在此基础上,基于透明中继传输场景,提出一种可以平衡接入域和透明域资源占用的比例的自适应分组调度策略,减少由于接入域和透明域的剩余资源比例失调而导致的中继用户分组数据无法及时调度的问题,进而提高系统时隙资源的利用率。分析结果表明,该方案能满足不同类型业务QoS需求,保证用户间的公平性；在保证系统吞吐量性能基础上,进一步降低系统能耗。最后,针对两跳中继OFDMA蜂窝网下行传输系统,提出一种半分布式自适应分组调度方案。该方案能够区分实时(RT)和非实时(NRT)业务的QoS需求。为了保证实时业务的QoS需求和非实时业务获得一定的受调度机会,该方案给出一种能区分业务类型的改进的比例公平分组调度优先级因子设计方案；为了降低中继传输系统的信令开销,设计了一种自适应时隙重分配触发机制,将BS和RS端定义的满意度因子与给定的满意度阈值进行比较,以决定是否触发时隙重分配过程。一旦触发时隙重分配,BS端调度器将根据BS端和各个RS端请求时隙数自适应调整BS和各个RS间的带宽分配比例。分析结果表明,所提半分布式分组调度策略一方面可以提高系统吞吐量,满足用户的QoS需求,保证用户间公平性,另一方面可以降低系统的信令开销,有效降低中继用户切换丢包率。更多还原

【Abstract】 The multi-hop relay network with the orthogonal frequency multiplexing (OFDM)/orthogonal frequency multiple access (OFDMA) and relay technologies has many advantages, such as high spectral efficiency, coverage enhancement, low energy consumption, flexiable and low cost networking, and so on. which has regarded as one of the mainstreams of the network architecture of the next generation mobile communications. Because of the superiority, it has always been viewed as the promising network architecture of the next generation mobile communications. However, this novel network architecture has to face the more complex radio environment and the serious challenges about the radio resource management in comparison with the traditional cellular networks, which focuses on two main aspects:i) the traditional resource allocation problem is derived from the traditional simple dimension such as time domain, frequency domain or space domain to multi-dimensional direction, causing the corresponding resource allocation strategies can not be directly applied to relay networks, which greatly increases the complexity of resource allocation and the signaling overhea; ii) with the rapid growth of wireless traffic volume and the deployment of the great many low power relay, the energy consumption has experienced the significant increase, which can take away more than half of the energy consumption for all equipments. Unlike the researches only focus on the throughput enhancement previously, more and more researchers begin to study how to improve system efficiency to save energy under the conditions of satisfying the basic quality of service (QoS) requirements, with the emergement and development of "green communication" notions. Therefore, by taking the system energy-saiving as the main target, this dissertation investgates the key technologies of the wireless resource management for the OFDM/OFDMA-based wireless multihop relay networks, including path selection, packet scheduling and resource allocation.Firstly, two optimal energy-efficient path selection schemes, which are used for the OFDM system with amplify and forward (AF), as well as decode and forward (DF), are proposed respectively. According to the analysis of the impact of circuit power consumption, user rate requirements, bandwidth and other factors which have the effect on system energy efficiency for the path selection, two optimal energy-efficient path selection criteria based on equivalent path loss exponent are proposed respectively for both AF-mode and DF-mode relay OFDM systems by converting the fading values of relay and direct links into the equivalent path loss values.. The analysis results show that the optimal energy-efficient transmission path is determined by the factors of data rate, channel conditions and bandwidth, and satisfies the decision criteria of the optimal energy-efficient path proposed in this dissertation, based on which the three node DF relay transmission scenario is extended to relay OFDMA cellular downlink transmission system with multile users and multiple relays. Considering the circuit power consumption and fixed user rate demand conditions, a low complexity suboptimal algorithm for joint path selection, subcarrier and power allocation is proposed to maximize energy efficiency of the system. The analysis results indicate that the proposed path selection scheme can improve system energy efficiency significantly compared to the traditional schemes.Secondly, under the constraints of the minimum rate and the total transmission power, a joint optimal energy-efficient resource allocation scheme with the subcarrier pairing and power allocation is proposed for the case with a three-node OFDM-based DF-mode relay link,. Moreover, the proposed scheme considers the effect of the circuit power consumption on the system energy efficiency. As this problem is a non-convex optimization problem, of which the complexity is too high to be soloved, a case discussion method is introduced. Thus the non-convex optimization problem can be converted to a convex optimization, where the uniqueness of the optimal solution for the system energy efficiency is proved. Then, by using Lagrange multiplier method, KKT condition, and water-filling algorithm, the optimal solution of the system energy efficiency and power allocation results can be obtained. The analysis results show that, compared with the optimal energy-efficient power allocation strategy with the fixed subcarrier pairing, the proposed scheme can further improve the achievable data rate of the relay link and the energy efficiency. In addition, by comparing with the traditional optimal water-filling algorithm of the rate maximization or power minimization, the proposed scheme can adaptively allocate resources under the constraints of the total transmission power regulated by the system and the user data rate requirements to obtain the global optimal solution with lower complexity and ensure the reliability of link transmission.Thirdly, considering a more realistic scenario with the maximum power constraints of source node and relay node, a joint optimization scheme of subcarrier pairing and power allocation is proposed. Besides, in view of the constraints of circuit power consumption and the minimum user rate requirement, a joint optimization model of the subcarrier pairing and power allocation is formulated to obtain the optimal energy efficiency of the system. By converting the non-convex optimization problem to the the convex optimization problem, a case discussion method is proposed to solve this optimization problem and obtain the optimal solution using water-filling algorithm. The analysis results show that, the proposed scheme can adaptively allocate the resources under various constraints and achieve optimal energy efficiency of the system, which can also guarantee the transmission reliability and reduce the link outage probability as well as the algorithm complexity.Fourthly, for two-hop relay cellular OFDMA downlink transmission systems, an adaptive energy-saving-related packet scheduling scheme is proposed. In order to take the comprehensive factors, such as the differences between the QoS requirements for different types of services, status of packet queues, and the energy efficiency of links, into account, a proposed scheme design a priority factor of packet scheduling. On the basis of this, an adaptive packet scheduling strategy is put forward to balance the proportion of the occupied slots between the access zone and transparent zone, and reduce the probability that the packets in the packet queues of relay users cannot be timely scheduled which results from the imbalance of the proportion of leftover slots between the access zone and transparent zone, and further improve the utilization of system slot resource. The analysis results show that, the proposed scheme can satisfy the QoS requirements of the different types of services and ensure the fairness among user. Furtherly, it can reduce the energy consumption of the system under guaranteeing the system throughput performance.Lastly, with respect to two-hop relay OFDMA cellular downlink transmission systems, a semi-distributed adaptive packet scheduling scheme is proposed, by which the QoS requirements of real-time (RT) traffic and non-real-time (NRT) traffic can be distinguished.. To guarantee the QoS requirements of RT traffic and obtain certain scheduling opportunities for NRT traffic, a modified proportional fair scheduling priority factor is designed, which can distinguish the traffic types. In order to reduce the signaling overhead of relay transmission systems, an adaptive slot reallocation triggering mechanism is proposed, in which the slot reallocation process is determined whether it could be triggered by comparing the defined satisfaction factors in BS and RS with the given priority factor in advance.. Once the time slot reallocation is triggered, the scheduler in BS will timely adjust the proportion of bandwidth allocating to BS and each RS according to the requested slot number of BS and each RS. The analysis results show that, the proposed semi-distributed packet scheduling strategy is capable of improving the system throughput, meeting the QoS requirements of users, ensuring the fairness among users, modetating the signaling overhead of the system, and effectively reducing the handover packet loss rate of the relay users.更多还原