中文摘要：

针对无线射频识别（RFID）单元的分布情况可能会导致出现多种类型的冲突以及现存算法的系统成本较高等问题，提出多个数据通道完成阅读器和标签之间的通信。在接收问询流程中，阅读器首先等待一个伪随机时间，然后检测两个序列，如果这两个条件都不满足，阅读器将会等待伪随机时间后再次尝试。在接收过程中，如果存在任何可利用的数据通道，阅读器就会选取不同的数据通道。该设置可以确保避免发生隐藏和暴露终端问题。与邻域友好型防冲突算法（NFRA）、多通道MAC协议（MCMAC）算法和冲突避免（Dica）的分布式算法相比，实验结果显示提出的算法降低了标签识别过程中花费的时间，而且集中控制过程额外的空间成本更低，因此在最小化网络开销的情况下增加了成功询问的比例。

英文摘要：

Concerning that the placement of the elements of radio frequency identification （RFID） may result in several types of collisions, and the system cost of existing algorithms is high, this paper proposed a way using several data channels to com- plete the communication between the reader and the tag. In an inquiry process, a reader wait for a random time firstly, and then detected the two sequences. If these two conditions were not fit, the reader would wait for pseudo random time and try again. In the process of acceptance, if there were any available data channels, the reader would select a different data channel. This setting ensured that the hidden and exposed terminal problems were avoided. Compared with neighborhood friendly reader anti-collision （NFRA） algorithm, multi-channel MAC protocol （MCMAC） algorithm and collision avoidance distributed （Dica） algorithm, experimental results show that the proposed algorithm reduces more time needed for tags＇ identification, and the extra space cost in centralized control process is less, thus increasing the rate of successful interrogations while minimizing the network overhead.