Range Query Techniques Based On Key-Value Storage In DHT

The database of key-value type is non-relational database, which has a wide rangeof applications. In particular, in the context of large-scale and high-concurrentapplications, and in the process of huge unstructured content information, key-valuestorage systems play an important role. The storage model of DHT (Distributed HashTable) has its figuration long before, that may be traced back to its original source P2Psystems. A basic standpoint of this paper is to take DHT model into the database ofkey-value type. Data query technology is crucial in the key-value storage.In the paper, range queries on the storage of DHT are studied. Firstly, the existingtechnologies of resource search are analyzed and researched, including the typicaltechnologies of resource search on DHT, range queries technologies based on the treestructure and range queries on multiple properties, and the technical features of thesequeries are summarized. Then, the unique feature of range query is analyzedthoroughly and the range query strategy based on the resource detection mechanism isproposed. This method takes the fact of multiple properties into account. It makes useof the Hilbert space filling curve to map multi-dimensional attributes intoone-dimensional keywords, uses the consistent hashing to solve the distribution ofnodes and the store of resource, and takes the Bloom Filter and resource distributionneighbor table to construct the index information, which augment the amount ofinformation each node grasps so that it is helpful to save hops in the network. In thepaper, the method of range query based on the resource detection mechanism isintroduced in detail, the deployment of index information and the update of index aregiven, the algorithm of range query corresponding to the index strategy is describedand a brief analysis on the performance of the algorithm is furnished in the last. Finally,the simulation experiments are done to aim at the range query algorithm based onresource detection mechanism. The results indicate the range query algorithm based onresource detection mechanism is effective.