Improved Data Locality Using Morton-order Curve on the Example of LU Decomposition

Perdacher, Martin; Plant, Claudia; Böhm, Christian

doi:10.1109/bigdata50022.2020.9378385

Cited by 2 publications

(1 citation statement)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From Lemma 7 we know that a single POI results in an in-path oracle of size O( 12 n). For each entry, we store a block-pair (A, B) corresponding to the quadtree blocks as a four-dimensional Morton code [72,[74][75][76]. In this scheme, a pair of Morton codes of the same level can be bit-interleaved to form a 64-bit number with the level of the block stored in the most-significant bit position of the code.…”

Section: Storing the In-path Oraclementioning

confidence: 99%

In-Path Oracles for Road Networks

Ghosh

Sankaranarayanan

Khatter

et al. 2023

IJGI

View full text Add to dashboard Cite

Many spatial applications benefit from the fast answering to a seemingly simple spatial query: “Is a point of interest (POI) ‘in-path’ to the shortest path between a source and a destination?” In this context, an in-path POI is one that is either on the shortest path or can be reached within a bounded yet small detour from the shortest path. The fast answering of the in-path queries is contingent on being able to determine without having to actually compute the shortest paths during runtime. Thus, this requires a precomputation solution. The key contribution of the paper is the development of an in-path oracle that is based on precomputation of which pairs of sources and destinations are in-path with respect to the given POI. For a given road network with n nodes and m POIs, an O(m×n)-sized oracle is envisioned based on the reduction of the well-separated pairs (WSP) decomposition of the road network. Furthermore, an oracle can be indexed in a database using a B-tree that can answer queries at very high throughput. Experimental results on the real road network POI dataset illustrate the superiority of this technique compared to a baseline algorithm. The proposed approach can answer ≈ 1.5 million in-path queries per second compared to a few hundred per second using a suitable baseline approach.

show abstract

Section: Storing the In-path Oraclementioning

confidence: 99%