传感器网络微分环节中改进的最佳二阶近似模流量控制

摘要

Because self-organized wireless sensor network is randomly setup, data transmission among the nodes is subject to traffic jam. To address this problem, this paper proposes a method of optimum second order approximation modulus with an improved differentiation element to control data transmission. The controller applies a function of the optimum second order approximation modulus as its transfer function. It negatively feedbacks the data buffer length in a node to its input point, and controls the speed of data transmission of the up-stream node. A differential element is added to the transfer function to reduce over-modulation and speed the peak emergence. Thus no traffic overflow will occur both in transition and in permanent periods.%自组织无线传感器网络的结构组合具有随机性,当节点传输数据时,数据缓冲区很容易发生传输拥塞.为解决这个问题,该文提出对微分环节进行改进的最佳二阶近似模流量控制方法.该方法把传感器网络设计抽象成一个控制器,此控制器具有最佳二阶近似模传递函数.控制器把节点数据缓冲区的实际数据高度负反馈到输入端,根据缓冲区实际高度和理想高度的差值大小,由控制器控制上游节点发送数据的速率.为了减小系统超调量和使峰值提前,在传递函数中再加入一个微分环节,以确保网络在暂态和稳态阶段缓冲区数据都不会发生溢出.