摘要： 为解决RFID系统中的标签碰撞问题，在分析经典二进制搜索算法的基础上，提出一种基于二进制搜索的改进算法以实现标签的快速识别。算法采用栈和队列技术、后退策略、碰撞位跟踪及并行处理思想，在碰撞处理中，读写器只需发送一个三维向量命令，标签以部分碰撞位参数为应答，减少传输数据量和识别次数；利用队列存储已识别的ID，将读写操作与碰撞处理并行，省去单独读写时间；采用模块并行处理，使读写器具有双核处理能力。测试结果表明，该算法在减少传输数据量和识别次数上优于经典二进制算法。

关键词: 射频识别, 二进制搜索算法, 后退策略, 碰撞位跟踪, 并行处理

Abstract: To solve the problem of tags collision in RFID system, a new anti-collision algorithm based on classical binary-search algorithms is proposed to improve identification efficiency. In order to reduce query frequency and data interchange in the collision processing, the new algorithm adopts the techniques of stack and queue storage, back strategy, collision-bit tracking and parallel processing, etc. The algorithm can make reader only send a three-dimensional-vector order, and tags respond to the reader with some of collision-bit parameters. The algorithm utilizes the queue to store identified ID, and makes both read-write operations and collision processes work simultaneously, thus read-write time is left out. The reader has dual-core processor by using parallel processing modules. Performance result shows that this algorithm is significantly superior to the classical binary algorithms in reducing data interchange and query frequency.

Key words: RFID, binary-search algorithm, back strategy, collision-bit tracking, parallel processing

中图分类号: 

• TP391