无线体域网低能耗功率控制与调度算法研究

发布时间：2018-12-14 11:11

【摘要】：无线体域网(Wireless Body Area Network,WBAN)能够为患者提供低成本、持续、准确的医疗监护,近年来已成为远程医疗、重症监护等领域的新趋势。而受人体安全和元器件能量限制,低能耗设计在无线体域网的设计中至关重要。节点能耗按照功能分为数据采集、处理和传输三部分,其中传输能耗占比最大。功率控制和调度算法作为降低传输能耗的两个关键手段,其研究具有理论和现实意义。目前无线体域网功率控制和调度算法存在问题主要有(1)功率控制算法都是基于预先定义好的链路质量与发射功率的函数关系,这种函数关系简化了两个变量之间的数量关系,存在较大误差。(2)表征链路质量所选择的参数为信号强度(Received Signal Strength Indicator,RSSI),其过于灵敏且易受环境因素影响产生波动。(3)现有的调度算法使用网络生存期这一个指标进行衡量,未考虑具体应用下节点的传输速率和网络公平性要求。本文在传统算法的基础上,对功率控制和调度算法进行优化。在功率控制中运用反馈调节思想,在调度算法中考虑了传感器节点的不同传输要求。具体如下:(1)提出了一种基于比例、积分、微分(Proportional、Integral、Differential,PID)算法的反馈功率控制算法。算法用链路质量(Link Quality Indicator,LQI)均值表征包接收率,将网络运行时节点LQI均值与阈值的差值,通过PID算法得到功率调节量。实验结果表明,与传统的乘增加减算法和动态姿势推测算法相比,在保证97.6%包接收率的前提下,该算法的节点平均能耗分别减少17.3mw和13.7mw,网络生存期平均提高28.7%和23.4%。(2)提出了一种加入网络公平性的集中式调度算法。Sink节点根据全网各节点链路质量在时隙开始时统一选择节点调度,该问题被建模成一个加入公平性约束的马尔科夫决策过程,采用值迭代对模型求解得到最优调度策略。实验结果表明,合理选择公平性阈值可延长网络生存期。与传统算法比较,该算法延长了网络生存期。(3)针对集中式调度在开销大、不易扩展的局限性,提出了一种加入公平参数的分布式调度算法。在最大化生存期动态策略算法(Dynamic Protocol for Lifetime Maximization,DPLM)的能效指数中加入公平性参数,各节点根据当前自身的链路质量计算能效指数,根据其与退避时间的函数关系计算退避时间。实验结果表明,与DPLM算法相比,该算法在保证网络生存期的同时合理分配了传输时隙,满足了各节点的不同数据传输要求。
[Abstract]:Wireless body area Network (Wireless Body Area Network,WBAN) can provide low cost, continuous and accurate medical care for patients. In recent years, it has become a new trend in the field of telemedicine and intensive care. Due to the limitation of human safety and component energy, low energy consumption design is very important in the design of wireless body area network. Node energy consumption is divided into three parts according to its function: data acquisition, processing and transmission. As two key means to reduce transmission energy consumption, power control and scheduling algorithms are of theoretical and practical significance. At present, the main problems of power control and scheduling algorithms in wireless bulk area networks are as follows: (1) Power control algorithms are based on the predefined function relationship between link quality and transmit power. This functional relationship simplifies the quantitative relationship between two variables, and there are large errors. (2) the parameter chosen to characterize the link quality is signal strength (Received Signal Strength Indicator,RSSI. It is too sensitive and vulnerable to fluctuations due to environmental factors. (3) the existing scheduling algorithms are measured by the network lifetime, without considering the transmission rate and network fairness requirements of the nodes under specific applications. Based on the traditional algorithms, this paper optimizes the power control and scheduling algorithms. In the power control, the different transmission requirements of sensor nodes are considered in the scheduling algorithm by using the idea of feedback regulation. The details are as follows: (1) A feedback power control algorithm based on proportional, integral and differential (Proportional,Integral,Differential,PID) algorithm is proposed. The average value of link quality (Link Quality Indicator,LQI (link quality (Link Quality Indicator,LQI) is used to characterize the packet reception rate, and the difference between the average value of LQI and the threshold of the node when the network is running, and the power regulation is obtained by using the PID algorithm. The experimental results show that compared with the traditional multiplicative addition and subtraction algorithm and the dynamic postural estimation algorithm, the average energy consumption of the nodes in this algorithm is reduced by 17.3mw and 13.7 MW, respectively, while the 97. 6% packet reception rate is guaranteed. The average lifetime of the network is increased by 28.7% and 23.4%. (2) A centralized scheduling algorithm with fairness is proposed. According to the link quality of each node in the network, the Sink node selects the node scheduling uniformly at the beginning of the time slot. The problem is modeled as a Markov decision process with fairness constraints, and the optimal scheduling strategy is obtained by solving the model by value iteration. The experimental results show that the reasonable selection of fairness threshold can prolong the network lifetime. Compared with the traditional algorithms, the proposed algorithm extends the network lifetime. (3) in view of the limitations of centralized scheduling, which is expensive and difficult to extend, a distributed scheduling algorithm with fair parameters is proposed. The fairness parameter is added to the energy efficiency index of the dynamic strategy algorithm (Dynamic Protocol for Lifetime Maximization,DPLM. Each node calculates the energy efficiency index according to the current link quality, and calculates the Backoff time according to the function relationship between the energy efficiency index and the Backoff time. The experimental results show that compared with the DPLM algorithm, the proposed algorithm not only ensures the lifetime of the network, but also reasonably allocates the transmission time slots, which satisfies the different data transmission requirements of each node.
【学位授予单位】：宁波大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92

【参考文献】