%0 Journal Article
%T 基于主动链路质量估计的采集树协议
Collection Tree Protocol Based on Active Link Quality Estimation
%A 洪芳晗
%A 施伟斌
%A 谢梦
%A 杨宸
%J Software Engineering and Applications
%P 321-331
%@ 2325-2278
%D 2025
%I Hans Publishing
%R 10.12677/sea.2025.142029
%X 采集树协议CTP依赖数据包和信标的传输估计链路质量，在此基础上选择多跳传输路径，在长时间未进行传输的情况下，无法更新链路质量估计，当采用休眠机制时，难以利用信标序号估计链路质量。本文提出一种基于主动链路质量估计的采集树协议ALE-CTP，对CTP的链路估计方法做出改进，通过在本地采集干扰和噪声强度并计算输入特征，各节点能够利用基于线性回归的模型LEAPS主动估计链路质量。LEAPS采用近似的信干噪比样本的1~3阶原点矩作为输入特征，由于链路质量估计不依赖于传输过程，即使长时间没有数据或信标传输，也能对实时的链路质量进行准确估计。在TinyOS中实现了ALE-CTP，并在真实环境中进行了测试，实验结果显示，与采用4Bit的CTP相比，在有较强干扰的条件下，ALE-CTP具有更高的切换成功率和较短的响应时间，交付率提高了23%，时延降低13%。
The Collection Tree Protocol (CTP) estimates link quality based on the transmission of data packets and beacons, and selects multi-hop transmission paths accordingly. In the absence of transmission for a long time, the link quality estimation cannot be updated. When a sleep mechanism is employed, it is difficult to estimate link quality using beacon sequence numbers. In this paper we propose a collection tree protocol ALE-CTP based on active link quality estimation, which improves the link estimation method of CTP. By collecting interference plus noise intensity and calculating input features locally, each node can actively estimate link quality using the offline trained model LEAPS. LEAPS is based on linear regression and uses the first to third order origin moments of the approximate signal to interference plus noise ratio (SINR) samples as input features. It makes link quality estimation independent of transmission process, enabling accurate real-time estimation of link quality even in the absence of data or beacon transmissions over extended periods. ALE-CTP was implemented in TinyOS and tested in a real environment. The experimental results showed that compared with CTP using 4Bit, ALE-CTP had a higher switching success rate and shorter response time under strong interference conditions. The delivery rate was increased by 23% and the latency was reduced by 13%.
%K 采集树协议，
%K 链路质量估计，
%K 线性回归，
%K 近似的信干噪比
Collection Tree Protocol
%K Link Quality Estimation
%K Linear Regression
%K Approximate SINR
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=112350