| 摘要: |
| RFID 数据采集过程中漏读现象频频发生,降低了RFID(radio frequency identification)应用中查询结果的准确性.目前解决漏读问题的算法主要是以RFID 原始读数为粒度,并基于标签自身历史读数进行窗口平滑,这种作法会填补许多与查询无关的冗余数据,并且在多逻辑区域参与的复杂应用中,填补准确率较差.为解决上述问题,首次将RFID数据从数据层抽象到逻辑区域层作为处理的粒度,提出3 种基于动态概率路径事件模型的数据填补算法,通过挖掘已知的区域事件的顺序相关性来对后续发生的事件进行判断 |
| 关键词: RFID 技术 数据填补 概率路径事件模型 区域事件 漏读数据 冗余数据 |
| DOI： |
| 分类号: |
| 基金项目:Supported by the National Natural Science Foundation of China under Grant Nos.60773220, 60873009 (国家自然科学基金); the National Basic Research Program of China under Grant No.2006CB303000 (国家重点基础研究发展计划(973)) |
|
| RFID Data Interpolation Algorithm Based on Dynamic Probabilistic Path-Event Model |
|
GU Yu,YU Ge,LI Xiao-Jing,WANG Yi
|
| Abstract: |
| Missing reads occur frequently during RFID (radio frequency identification) data collection, which will reduce the accuracy of query results in RFID applications. To solve this problem, the existing algorithms mainly take primitive RFID readings as granularity and adopt window smooth strategy based on tag historical readings, which may interpolate data that the query doesn’t care about and incur inaccuracy when multiple logic areas are involved. In this paper, data are transformed from data level to logic area level as the interpolation granularity. Then three data interpolating algorithms based on the probabilistic path-event model are proposed, where the incoming events are judged and interpolated by mining the sequence correlation of known area events. Furthermore, the factor of time is considered,and thus probabilistic path-event model is developed. Abundant experiments prove the proposed algorithms have different performance advantages in different conditions and are predominant over the existing strategy in redundancy and accuracy. |
| Key words: RFID technology data interpolation probabilistic path-event model area event missing reads redundant data |
