Show simple item record

dc.contributor.authorShen, Hongen_US
dc.description.abstractThe range-join of two sets R and S is the set that contains all tuples (r, s) satisfying e1⩽|r-s|⩽e2, r?R and s?S. For computing the range-join of R and S in a hypercube of p processors, this paper presents an improved selection-based parallel algorithm which reduces the local memory from O(n) repaired in the previous algorithm to O(m+n/p), where |R|=m, |S|=n and p⩽max{m,n}. The new algorithm also reduces the best-case time complexity from O(m/p log2 p+n/p log m) of the previous result to O(m+n/p log2p) when m⩾plog, while maintaining the cost optimality in the worst case. Unlike the previous algorithm, our algorithm works by selecting the median of RUS to evenly partition the whole data set for divide-and-conquer join in the next phase. We present an upper bound of time complexity of the algorithm in the general case and show that the best-case time complexity of the algorithm is better than permutation-based range-join when n⩾plogp+1en_US
dc.relation.ispartofconferencenameNot Knownen_US
dc.subject.fieldofresearchENVIRONMENTAL SCIENCESen_US
dc.titleAn improved selection-based parallel range-join algorithm in hypercubesen_US
dc.typeConference outputen_US
dc.type.descriptionE1 - Conference Publications (HERDC)en_US
dc.type.codeE - Conference Publicationsen_US
gro.hasfulltextNo Full Text
gro.griffith.authorShen, Hong

Files in this item


There are no files associated with this item.

This item appears in the following Collection(s)

  • Conference outputs
    Contains papers delivered by Griffith authors at national and international conferences.

Show simple item record