基于衰落信道的无线链路调度算法研究

发布时间：2018-08-24 08:55

【摘要】：过去几十年,无线传感器网络(Wireless Sensor Networks,WSNs)得到了飞速的发展,取得了一系列的成果,极大的改善了人们的生活方式和生活质量。作为极具应用和发展前景的一项高新技术,WSNs同样存在一系列亟待解决的问题,其中链路调度问题直接影响着无线网络的容量、时间延迟、生命周期等各个方面,成为WSNs研究中的热点问题。然而现在存在的大多数的结果都是在图干扰模型或者简单的SINR(Signal to Interference plus Noise Ratio)干扰模型下得出的,由于信号具有衰落的本质特征,因此这些结果往往与实际情况存在差距,不够精确。Rayleigh衰落模型充分考虑了传输信号的干扰累加、慢衰落等对链路调度过程的影响,更加接近真实的传输环境,因此,我们选择Rayleigh衰落模型作为研究链路调度问题的干扰模型。考虑到功率控制对于链路调度的重要意义,功率控制手段的选择就显得极为重要。本文中,我们选择Oblivious功率控制作为功率控制的手段,有效降低了传感节点的能量消耗,同时减小了节点之间的相互干扰,进而提高了同时调度的通信链路的数目,降低了通信链路的传输等待时间。链路调度成功率和系统吞吐量同样是评价一个调度算法的重要指标。如果同时调度的通信链路数量太少的话,系统的吞吐量就会过低。而如果同时调度的通信链路数量过多的话,相互之间的累加干扰就会增大,链路传输的成功概率就会降低。在本文中,我们首先通过合理设置每一个调度区域的大小来尽可能的增加同时调度的通信链路的数量,同时,保证系统中的通信链路以一个合适的概率被成功调度,然后,借助差错重传机制进一步地提高通信的成功率。事实上,尽管我们是在衰落模型下研究链路调度问题,我们依然能保证取得和SINR模型下相似或者更高的成功概率,与此同时,系统的吞吐量维持在一个较高的水平上。本文中,我们提出了以下几种有效的分布式链路调度算法:基于正三角形的调度算法、基于正方形的调度算法、基于正六边形的调度算法、基于正六边形的3-染色调度算法、基于正六边形的4-染色调度算法。并分别比较了几种调度策略以及与现有结果在链路通信质量,系统的通信容量以及发送节点的平均等待时延上的差异,从而能够针对不同的系统需求更好的选择算法。理论分析和仿真表明,我们的算法不仅能否有效的避免调度冲突,提高系统利用率,而且能够获得一个较高的系统容量和较低的时间延迟,同时与最优调度相比,仅相差一个常数近似因子。
[Abstract]:In the past few decades, (Wireless Sensor Networks,WSNs (Wireless Sensor Network) has been developed rapidly, and a series of achievements have been made, which have greatly improved people's life style and quality of life. As a new high-tech WSNs with great application and development prospects, there are also a series of problems to be solved. Among them, the link scheduling problem directly affects the capacity, time delay, life cycle of wireless networks, and so on. It has become a hot issue in WSNs research. However, most of the existing results are obtained under the graph interference model or the simple SINR (Signal to Interference plus Noise Ratio) interference model. Because the signal has the essential characteristics of fading, these results are often different from the actual situation. The influence of interference accumulation and slow fading on the link scheduling process is fully considered in the Rayleigh fading model, which is closer to the real transmission environment. We select the Rayleigh fading model as the interference model to study the link scheduling problem. Considering the importance of power control to link scheduling, the choice of power control means is very important. In this paper, we choose Oblivious power control as the power control method, which can effectively reduce the energy consumption of sensor nodes, reduce the mutual interference between nodes, and then improve the number of communication links that are scheduled at the same time. The transmission latency of communication link is reduced. Link scheduling success rate and system throughput are also important indicators to evaluate a scheduling algorithm. If too few communication links are scheduled at the same time, the system throughput will be too low. If the number of communication links scheduling at the same time is too large the cumulative interference between each other will increase and the probability of successful link transmission will be reduced. In this paper, we first increase the number of simultaneously scheduled communication links by reasonably setting the size of each scheduling area, and at the same time, ensure that the communication links in the system are successfully scheduled with a suitable probability, and then, The success rate of communication is further improved by means of error retransmission mechanism. In fact, although we study the link scheduling problem in the fading model, we can still guarantee a similar or higher probability of success to the SINR model, while the throughput of the system is maintained at a high level. In this paper, we propose the following effective distributed link scheduling algorithms: triangular scheduling algorithm, square based scheduling algorithm, hexagon based 3-coloring scheduling algorithm. 4-coloring scheduling algorithm based on hexagonal. Several scheduling strategies are compared and compared with the existing results in terms of link communication quality, system capacity and the average waiting delay of the sending node, so that the algorithm can be better selected according to different system requirements. Theoretical analysis and simulation show that our algorithm can not only effectively avoid scheduling conflicts and improve system utilization, but also can obtain a higher system capacity and lower time delay, at the same time compared with the optimal scheduling. The difference is only one constant approximation factor.
【学位授予单位】：曲阜师范大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP212.9;TN929.5

【共引文献】