Gray codes for partial match and range queries

It is suggested that Gray codes be used to improve the performance of methods for partial match and range queries. Specifically, the author illustrates the improved clustering of similar records that Gray codes can achieve with multiattribute hashing. Gray codes are used instead of binary codes to map record signatures to buckets. In Gray codes, successive codewords differ in the value of exactly one bit position; thus, successive buckets hold records with similar record signatures. The proposed method achieves better clustering of similar records, thus reducing the I/O time. A mathematical model is developed to derive formulas giving the average performance of both methods, and it is shown that the proposed method achieves 0-50% relative savings over the binary codes. The author also discusses how Gray codes could be applied to some retrieval methods designed for range queries, such as the grid file and the approach based on the so-called z-ordering. Gray codes are also used to design good distance-preserving functions, which map a k-dimensional (k-D) space into a one-dimensional one, in such a way that points are close in the k-D space are likely to be close in the 1-D space.<>