WSNs中追求低能耗的休眠调度算法的研究

发布时间：2018-06-17 23:53

本文选题：无线传感器网络 + 传感器调度　；参考：《吉林大学》2017年硕士论文

【摘要】：近年来,无线传感器因其具有体积小和价格低廉等优点被广泛地应用于国防军事、医疗健康和环境监测等领域。但无线传感器由电池供电,较小的体积限制其只能携带有限的电量,并且电池更换难度大,因此无线传感器网络的能耗问题一直被人们所关注,越来越多的学者投入到提高网络能效的研究中。通过对有向传感器网络特点分析,提出迭代优化休眠调度算法来解决有向传感器网络中的目标覆盖问题。使用二维数组表示候选解,并随机初始化指定数量的候选解添加到候选解集合中。重组函数根据当前候选解集合产生新的候选解,有利于丰富候选解集合的多样性。算法先后调用调整函数和休眠函数作用于候选解集合,调整函数尝试调整候选解中传感器节点的感知方向来减少传感器使用数量,从而节省能量;休眠函数是在调整函数结束后尝试去除候选解中过多的冗余覆盖,减少能量浪费。算法对初始候选解集合施加多轮计算、重组、调整和休眠等操作后获取较优问题解。通过实验验证,与领域内另外两种算法对比,提出的算法能够获取更长的网络生存周期。针对全向传感器网络的目标覆盖问题提出多阶段贪心覆盖算法。算法主要分为三个阶段,一是通过适应度函数选择合理的传感器节点构建覆盖集合,适应度函数综合考虑传感器节点的覆盖贡献率、对其它目标点的覆盖影响度、能量标准差等因素;二是针对第一步所构建的覆盖集合进行无用覆盖优化,在保证完成覆盖任务的前提下,进行半径优化调整来减少网络中的无用覆盖;三是冗余覆盖优化阶段,在无用覆盖优化基础上通过调整节点的感知半径大小或者调整节点的工作状态来进一步节省能量。实验结果表明,与解决相同问题的算法相比,提出的多阶段贪心覆盖算法能获取更长的网络生存周期。另外,基于迭代优化休眠调度算法和多阶段贪心覆盖算法进行了覆盖优化系统的设计与实现,对系统进行了需求分析及功能设计,系统大体上分为交互模块、数据处理模块和系统控制模块。数据处理模块负责对输入的数据进行校验和转换等操作,便于其他模块使用,系统控制模块作为系统的核心,根据输入的网络信息进行复杂的计算。系统可以对不同场景进行快速高效的覆盖方案计算。
[Abstract]:In recent years, wireless sensors have been widely used in military defense, medical health and environmental monitoring due to their small size and low cost. However, the wireless sensor is powered by battery, the small volume limits it can only carry a limited amount of electricity, and the battery replacement is difficult, so the problem of energy consumption in wireless sensor network has been concerned by people all the time. More and more scholars put into the research of improving network energy efficiency. By analyzing the characteristics of directed sensor networks, an iterative optimal sleep scheduling algorithm is proposed to solve the target coverage problem in directed sensor networks. A two-dimensional array is used to represent the candidate solution and a random number of candidate solutions are initialized to the candidate solution set. The recombination function generates new candidate solutions according to the current candidate solution sets, which is helpful to enrich the diversity of candidate solution sets. The algorithm successively calls the adjustment function and the sleep function to act on the candidate solution set. The adjustment function tries to adjust the sensing direction of the sensor node in the candidate solution to reduce the number of sensors used, thus saving energy. The dormant function tries to eliminate the redundant cover in the candidate solution after the end of the adjustment function and reduces the energy waste. The algorithm applies multiple rounds of computation, recombination, adjustment and hibernation to the initial set of candidate solutions to obtain the optimal solution. Compared with the other two algorithms in the domain, the proposed algorithm can obtain longer network lifetime. A multistage greedy coverage algorithm is proposed for target coverage in omnidirectional sensor networks. The algorithm is mainly divided into three stages. Firstly, the suitable sensor nodes are selected by fitness function to construct coverage set. The fitness function considers the coverage contribution rate of sensor nodes and the coverage impact on other target points. Secondly, the useless coverage is optimized according to the cover set constructed in the first step, and the radius is adjusted to reduce the useless coverage in the network on the premise of completing the coverage task. The third is redundancy coverage optimization stage, which can save energy further by adjusting the perceived radius of nodes or adjusting the working state of nodes on the basis of useless coverage optimization. Experimental results show that the proposed multi-stage greedy cover algorithm can obtain longer lifetime than the algorithm to solve the same problem. In addition, based on the iterative optimal sleep scheduling algorithm and the multi-stage greedy coverage algorithm, the design and implementation of the coverage optimization system are carried out, and the requirements and functions of the system are analyzed. The system is divided into interactive modules. Data processing module and system control module. The data processing module is responsible for the operation of checking and converting the input data, which is convenient for other modules to use. The system control module is the core of the system, and the complex calculation is carried out according to the input network information. The system can calculate the coverage scheme of different scenarios quickly and efficiently.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【相似文献】