分级自适应伪贝叶斯时隙ALOHA控制算法  

Ranked adaptive pseudo bayesian control algorithm for slotted ALOHA

作　　者：方飞[1,2] 余文春[1] 

机构地区：[1]内江师范学院工程技术学院,四川内江641110 [2]电子科技大学通信学院,成都611731

出　　处：《重庆邮电大学学报（自然科学版）》2016年第3期303-311,共9页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家自然科学基金(61471102);四川省教育厅重点项目资助(13ZA0005)~~

摘　　要：时隙ALOHA需要使用控制算法以保证系统获得稳定吞吐量,而控制算法的关键是精确估计系统中的节点数。针对系统中的节点数在某时刻发生剧变时,传统的伪贝叶斯控制算法(pseudo Bayesian control algorithms,PBCA)存在调节时间较长的问题,提出分阶快速自适应伪贝叶斯控制算法(ranked fast adaptive PBCA,RFA-PBCA),首先设置一个门限值nth把信道的竞争情况分成高强度和低强度2种类型,再利用游程技术将信道分成空闲(冲突)和非空闲(非冲突)2种状态。当信道检测到c(c=6)个空闲时隙时,若估计节点数n大于nth,则节点将其传输数据概率增大为原先的2倍;同理,当信道检测c(c=6)个冲突时隙后,将其传输数据概率设为原先的1/2,其他情况下则采用PBCA进行调整。仿真结果表明,RFA-PBCA能够很好适应系统节点急剧变化的应用场景,其性能明显优于传统的PBCA。The control algorithm is necessary for slotted ALOHA in order to achieve the stabsentiality of control algorithm is to derive the acute number of system nodes. The classic control algorithm suffers perform-ance loss when the system nodes sharply changes. In this paper, a ranked fast adaptive pseudo Bayesian control algorithm（RFA-PBCA） was proposed to promote the performance of ALOHA. Firstly, a threshold value is set up to divide thechannel into two types of high intensity and low intensity. Then, the channels are grouped into two states, idle （ or con-flict） and not idle （or not conflict） by the technology of run. When 6 idle channels are detemated node ra is greater than nth , all nodes update their probability of transmitting data as 2 times. Similarly, if 6 collision slots are detected, the number of estimate nodes will be doubled and the probability of transmitting data is updated as 1/2 times. In other cases, the pseudo Bayesian control algorithms （PBCA） are adopted. The results of simulation show that can PFA-PBCA well adapt to the scenarios where node of system rapidly changes, its performance outperforms than pseudo-Bayesian control algorithms.

关 键 词：时隙ALOHA 游程 分阶快速自适应 伪贝斯控制算法 

分 类 号：TN929.5[电子电信—通信与信息系统]