相似度自适应估计的物联网实体高效搜索方法

针对现有相似实体搜索方法缺乏对于观测序列长度的自适应性，且搜索过程数据存储开销过大，搜索结果准确性较低的问题，该文提出相似度自适应估计的物联网实体高效搜索方法(SAEES)。首先，设计了轻量级观测序列分段表示方法，对传感器采集的实体原始观测序列进行轻量级分段压缩表示，以降低实体观测序列的存储开销。然后，提出了观测序列相似度自适应估计方法，实现对不同观测序列长度的实体相似性的准确估计。最后，设计了高效的相似实体搜索匹配方法，依据所估计的实体相似度进行实体的准确搜索匹配。仿真结果表明，所提方法可大幅提高相似实体搜索的效率。

Efficient Search Method for IoT Entities with Similarity Adaptive Estimation

The existing similar entity search method has poor adaptability to the length of the observed sequence, and the data storage overhead in the search process is too large, and the accuracy of the search result is insufficient. To this end, an efficient search method is proposed for the IoT Entity Search with Similarity Adaptive Estimation (SAEES). Firstly, in order to reduce the storage overhead of the entity observation sequence, a lightweight method of segmentation representation of the observation sequence is designed to perform a lightweight segmentation compression representation of the original observation sequence of the entity collected by the sensor. Then, in order to achieve an accurate estimation of the similarity of entities with different observation sequence lengths, an adaptive estimation method for observation sequence similarity is proposed. Finally, by exploiting the designed efficient similar entity search matching method, the exact search matching of the entity is completed according to the estimated entity similarity. The simulation results show that the proposed method can greatly improve the efficiency of similar entity search.