无线异构网的高效自适应资源分配与优化

发布时间：2018-05-03 00:41

本文选题：频谱效率 + 能效　；参考：《西南交通大学》2014年博士论文

【摘要】：未来高容量和强覆盖性的通信网络中,由宏基站、中继站和微微基站等低功率站点构成的异构网受到了广泛关注。过去的十几年,大部分的研究都是通过设计不同的网络架构来改善小区的频谱效率进而提高小区的网络吞吐率。但为了满足用户日益增长的数据需求和无处不在的接入需求,急需对网络基础设施进行快速的扩张,由此引入了巨大的能量消耗,随着引入低功率站点,网络的能耗单元也相应地增加,所以无线异构网的能效问题变得更为关键。如何在保障用户服务质量(QoS)的情况下,保证系统拥有高能效,也是现今通信技术发展的趋势。本文以下行链路的无线资源分配为基本研究对象,结合时域、频域、功率域等方面的技术,以高的频谱效率或高能效为目标,对异构网中,中继网络和宏-微微网络的路径选择、资源调度及分配、干扰协调等算法进行优化设计。首先,提出一种解码转发(DF)中继系统中联合部分频率复用和自适应调制技术(AMC)的载波分配方案。该方案采用了全双工、双天线的DF中继模式来提高时隙资源利用率,利用AMC技术对干扰容忍的特性,动态复用基站(BS)和中继站(RS)所使用的频率来提高频谱利用率,从而获得高能效的传输速率。这种联合部分频率复用方法和全复用中继模式的方案比起单天线半双工中继模式,能提升较大的系统吞吐率。其次,基于离散调制水平集合,提出了非再生中继多用户网络的中继选择和载波、比特分配的联合优化方案。该方案在用户速率限制的条件下,通过使发送功率最小来达到提升能效的目的。首先,基于初始问题的目标,提出了一个基于放大转发(AF)中继系统的低复杂度的中继选择方案。然后根据中继选择所得结果,将中继用户的两跳链路虚拟成直传链路,由此将中继小区问题转换成传统的蜂窝小区问题,进行载波、比特的分配。最后,考虑中继最大功率的限制,对超负荷的中继进行适当的调整。通过分析结果可知,本文所提次优方案,在降低了初始问题算法复杂度的同时,所得的结果接近最优方案。而且,本文所提中继选择方案较传统的中继选择方案,降低了系统的功耗。再次,基于离散调制水平集合和机会网络编码,提出了再生中继网络中一种中继选择和载波、比特分配的联合优化方案。该方案根据信道状态,机会地引入网络编码从而提高频谱利用率,在用户速率限制的条件下,通过使每帧的发送功率最小来达到提升能效的目的。该方案将中继选择和载波、比特分配进行分步求解,得到一个次优解。其中,利用本文所提中继选择方案,会使更多的用户选择中继,这为网络编码提供了更多的机会。而所提的机会网络编码分配方案根据信道状态机会地进行网络编码,所以比固定网络编码分配方案和无网络编码方案大大节约了系统的功耗。然后,基于DF中继和AF中继的正交频分多路复用(OFDM)频率选择性链路,分别提出了最优能效的功率分配方案。这两个方案在证明了最优解存在且唯一的情况下,考虑了电路所耗功率的影响。其中,DF中继链路的功率方案利用DF中继解码转发的特性,将两跳链路分离进行求解,通过比较选取最好的分配方式,该方法可求出最大化能效的最优功率解,是DF中继链路能效上界；而AF中继链路的功率分配方案包含两个步骤,先根据信道进行载波分配,并根据载波配对的结果,将AF中继链路虚拟成一跳频选链路,然后利用一跳频选链路的方法进行功率分配,该AF中继链路能效自适应优化方法可获得逼近最优解的次优解。此外,DF中继和AF中继的链路方案都考虑了速率和功率限制下的功率分配方法,并分析了其对能效的影响。接下来,在总功率和用户速率均有限制的要求下,提出了用户非均匀分布的AF中继小区下行链路中继选择和功率分配联合优化方案。利用AF中继链路的虚拟增益,将中继问题转换为单跳小区问题,并用注水算法求得次优的功率分配。该算法可较好地逼近最优解,得到总能效的下边界。此外,由于电路功耗在系统总能耗中占有较大比例,为了进一步节约能耗,从而提高能效,该方案引入了中继和基站的休眠模式,根据用户分布情况关闭负载较低的站点,从而以较小的容量损失,获得更高的能效。最后,在由宏小区和微微小区组成的同频异构网中,提出了一种时域资源分配与低功率-几乎空白子帧(ABS)功率分配的联合优化方案。该方案通过动态协调宏小区之间、微小区之间和隶属不同小区的宏微小区之间的干扰,解决了用户的小区选择问题、资源分配问题,以及低功率ABS的功率分配问题。研究结果显示,动态ABS功率与时域资源联合分配方案相比静态低功率ABS功率分配和无功率ABS方案,大大提升了整个系统的性能。
[Abstract]:In the future high capacity and strong coverage communication networks, the heterogeneous networks of low power stations, such as Acer station, relay station and micro base station, have received extensive attention. In the past decade, most of the studies have improved the frequency efficiency of the cell by designing different network architectures to improve the cell network throughput. The growing data requirements and ubiquitous access requirements of the foot users are in urgent need of rapid expansion of the network infrastructure, which leads to the introduction of huge energy consumption. With the introduction of low power stations, the energy consumption units of the network are also increased accordingly. So the energy efficiency of the wireless isomerism becomes more critical. In the case of service quality (QoS), it is also the trend of modern communication technology to ensure the system has high energy efficiency. The wireless resource allocation in this paper is the basic research object, combining the technology of time domain, frequency domain and power domain, with the aim of high frequency efficiency or high efficiency, and the relay network and macro micro network in the heterogeneous network. The path selection of the collaterals, the resource scheduling and distribution, the interference coordination and other algorithms are optimized. First, a carrier allocation scheme for the joint part frequency reuse and adaptive modulation (AMC) in the DF relay system is proposed. The scheme uses full duplex and dual antenna DF relay mode to improve the utilization of time slot resources, and uses AM The C technology improves the spectrum utilization by dynamically multiplexing the frequency of the base station (BS) and the relay station (RS) to improve the efficiency of the transmission rate. This combined partial frequency reuse method and the fully multiplexed relay mode can improve the system throughput compared with the single antenna half duplex relay mode. Secondly, Based on the discrete modulation level set, a joint optimization scheme of relay selection and carrier and bit allocation for non renewable relay multiuser networks is proposed. This scheme achieves the goal of improving energy efficiency by minimizing the transmission power under the condition of user rate constraints. First, a amplification and forwarding based on the target of initial problem is proposed. AF) the relay selection scheme of the relay system is low complexity. Then, according to the result of the relay selection, the relay user's two hop link is virtual to direct link, thus the relay cell problem is converted into a traditional cell problem, carrier, bit allocation. Finally, the limit of the maximum power of the relay is taken into consideration. According to the analysis results, the proposed suboptimal scheme can reduce the complexity of the initial problem algorithm and close the optimal scheme. Moreover, the relay selection scheme proposed in this paper reduces the power consumption of the system compared with the traditional relay selection scheme, and again, based on the discrete modulation level set and the opportunity network. A joint optimization scheme of relay selection and carrier and bit allocation in regenerative relay network is proposed. This scheme introduces network coding in accordance with channel state to improve spectrum utilization, and achieves the goal of improving energy efficiency by minimizing the transmission work rate per frame in the condition of user rate restriction. After selection and carrier, bit allocation is solved step by step, and a suboptimal solution is obtained. Among them, the proposed relay selection scheme in this paper makes more users choose relay, which provides more opportunities for network coding. The network coding scheme and the non network coding scheme greatly save the power consumption of the system. Then, based on the orthogonal frequency division multiplexing (OFDM) frequency selective link of DF relay and AF relay, the optimal energy allocation scheme is proposed respectively. These two schemes have shown that the optimal solution exists and is unique, considering the consumption of the circuit. Among them, the power scheme of the DF relay link uses the characteristics of DF relay decoding and forwarding. The two hop link separation is solved, and the best allocation method is selected by comparison. This method can find the optimal power solution to maximize the energy efficiency. It is the energy efficiency upper limit of the DF relay link; and the power allocation scheme of the AF relay link contains two steps. First, the carrier allocation is carried out on the basis of the channel, and the AF relay link is virtual as a frequency hopping link based on the result of the carrier pair, and then the power allocation is used by a frequency hopping link selection method. The AF relay link energy efficiency adaptive optimization method can obtain the suboptimal solution of the optimal solution. In addition, both the DF relay and the AF relay link schemes are both. The power allocation method under the restriction of rate and power is considered, and its effect on energy efficiency is analyzed. Next, a joint optimization scheme for the relay selection and power allocation of the AF relay cell downlink is proposed under the requirements of the total power and the user rate, and the virtual gain of the AF relay link will be used in the virtual gain of the relay link. The following problem is converted into a single hop cell problem, and the power allocation of the sub optimal is obtained by water injection algorithm. The algorithm can better approximate the optimal solution and get the lower boundary of the total energy efficiency. In addition, the power consumption of the circuit occupies a large proportion in the total energy consumption of the system. In order to further economize on energy consumption and improve energy efficiency, the scheme introduces the rest and the rest of the base station. The sleep mode closes the low load sites according to the user distribution, thus obtaining higher energy efficiency with smaller capacity loss. Finally, a joint optimization scheme for time domain resource allocation and low power almost empty white subframe (ABS) power allocation is proposed in the same frequency heterogeneous network composed of macro cells and micro cells. The dynamic coordination between the macro cell, the small area and the macro microcell belonging to the different communities can solve the user's cell selection problem, the resource allocation problem, and the power allocation problem of the low power ABS. The results show that the dynamic ABS power and the time domain resource allocation scheme are compared with the static low power ABS power allocation and the power allocation scheme. The power less ABS scheme greatly improves the performance of the whole system.

【学位授予单位】：西南交通大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN925

【参考文献】