This paper studies indexing multidimensional data based on the Borel Cayley graph. In order to provide the scalable query method, a special distributed hash table based on Borel Cayley graph is applied to key-value-based data-storing system. By taking advantage of the two different representations for the same one Borel Cayley graph, we present an efficient way for distributed multidimensional queries. Compared with other ring-like indexing method, experimental results demonstrate that Borel Cayley-based index method reduces the routing table length and improves the index maintenance efficiency. 1 INTRODUCTION The latest research report from International Data Corporation (IDC) shows that the global revenue of big data and business analytics (BDA) was $189.1 billion in 2019, and it will grow more than $80 billion in 2022. There are hundreds of different big data applications in many application fields, such as purchase transactions, social networks, and healthcare system. Facing such distributed, unstructured big data, it is a challenging task to manage such amounts of heterogeneous data efficiently. GFS 1,2 proposed by Google company, Open source implementations HBASE 3 and HDFS of Hadoop, Dynamo 4 of Amazon, Cassandra 5 of Facebook are the representative massive data management systems. These systems store data in key-value models as elementary cloud infrastructure. Without proper indexing method, key-value-based storing system needs to scan the entire data set for every query request by using parallel processing methods, for example, MapReduce. By using key-value pairs in key, key-value storing system usually is a one-dimensional index. Proceeding one-dimensional range query can be efficiently achieved by almost every distributed file system. However, when query requirement is multidimensional, it is difficult for the key-value data-storing method to support complex queries, such as DBMS-like query and range query. To address the problems mentioned above, we present a method for indexing information on the key-value data-storing system. Based on the architectures, the current proposed indexing strategies can be classified into master-slave architecture and peer-to-peer based overlay network architecture. In master-slave architecture, every node in the cloud storing system maintains both the global index and local index. EMINC 6 and MD-HBase 7 are two typical master-slave architectures for key-value storing. However, each node will get a
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.