Abstract: Aiming at the defects of backoff window adjustment mechanisms in Binary Exponential Backoff（BEB） algorithm and fairness of each node when data transmited, this paper proposes a Periodic Two-step sampling Exponential Backoff（PTEB） algorithm. PTEB firstly introduces the concept of the period sampling, on this basis, introduces a competitive channel parameters [Qc] and network congestion parameter [Qb]. [Qb] and [Qc] are calculated in the sampling period, and defines the contention window adjustment formula based on the values of these two parameters. Adjustment of the window needs to go through two stages, each stage corresponds to different formulas. Simulation results show that PTEB algorithm can effectively reduce the collision number of nodes, increase the network throughput rate and fairness between the nodes.

Key words: Binary Exponential Backoff（BEB）, periodic, sample, contention window, throughput rate, collision, fairness

摘要： 针对二进制指数退避算法（BEB）中退避窗口调整机制的不足以及在各个节点传输数据时公平性差的问题，提出了周期性采样的两步指数退避算法（PTEB）。PTEB算法首先引入了采样周期的概念，在此基础上又引入了信道竞争能力参数[Qc]和网络拥挤参数[Qb]。在采样周期内计算[Qc]与[Qb]的值，并根据这两个参数的值定义了调整竞争窗口的公式。窗口的调整需要经过两个阶段来完成，不同的阶段分别对应不同的公式。仿真结果表明PTEB算法能够有效地减少节点的碰撞次数，增加网络的吞吐率和节点的公平性。

关键词: 二进制指数退避算法（BEB）, 周期, 采样, 竞争窗口, 吞吐率, 碰撞, 公平性