中文摘要：

针对后向散射通信广泛采用的EPC Gen2协议在传输大量数据时效率较低,且不能很好适应能量捕获条件和信道质量变化的问题,首先对EPC Gen2协议原有流程进行了优化,采用突发传输减少空闲等待时隙,引入纠删码减少确认帧等待时延;接着,推导了特定能量捕获条件和信道条件下的有效吞吐量表达式,归纳出有效吞吐量最大化问题,通过穷举法得到帧长、编码冗余度和充电时间理论最优值;然后,提出一种动态帧和充电时间在线调整策略,在运行时根据阅读器端的吞吐量测量反馈调整帧长及编码冗余度,控制充电时间,并设计了相应的数据帧和确认帧格式;最后,为验证本文方法的有效性,设置了不同能量捕获条件和信噪比条件,通过仿真得到本文所提方法采用的帧长、编码冗余度、达到的有效吞吐量和消耗的能量,与EPC Gen2协议固定帧长方案和通过数值分析得到的理论最优值进行了比较。仿真结果表明,本文所提方法相比EPC Gen2协议固定帧长方案能显著提高有效吞吐量,且吞吐量性能接近理论最优值。可见,本文对协议流程的优化及动态调整帧长及编码冗余度的策略在改善系统吞吐量性能方面是有效的。

英文摘要：

The widely-used EPC Gen2 protocol for backscatter communication is inefficient in dealing with transferring a large amount of buffered data to RFID reader. Furthermore,it is not adaptable to dynamic energy harvesting and channel condition. In order to solve these problems,firstly,the original logic of EPC Gen2 protocol was optimized by adopting burst transmission to reduce the idle slots and introducing erasure code to avoid waiting for the ACK. Secondly,the expression of the throughput under specific energy harvesting and channel conditions was derived. Then the throughput maximization problem was formulated,which was solved by the method of exhaustion. The optimal combination of the frame length,the number of redundant frames and the charging time were obtained. Thirdly,a dynamic frame and charging time adaptation（ DFCTA） scheme was proposed,which increased or decreased the frame length and the number of redundant frames at runtime based on the throughput measurement at the reader. Finally,simulations were conducted under different energy harvesting and channel conditions to evaluate the frame length,the number of redundant frames,the achieved throughput and the energy cost of the proposed DFCTA scheme. Simulation results showed that by DFCTA scheme,the throughput performance was remarkably improved compared with current fixed-frame-length scheme in EPC Gen2 protocol and it was quite close to the theoretical optimum. Consequently,the optimization of the protocol logic and the proposed DFCTA scheme were proved to be effective in improving the throughput performance.