Indexing Moving Objects Using Short-Lived Throwaway Indexes

Abstract

Abstract. With the exponential growth of moving objects data to the Gigabyte range, it has become critical to develop effective techniques for indexing, updating, and querying these massive data sets. To meet the high update rate as well as low query response time requirements of moving object applications, this paper takes a novel approach in moving object indexing. In our approach we do not require a sophisticated index structure that needs to be adjusted for each incoming update. Rather we construct conceptually simple short-lived throwaway indexes which we only keep for a very short period of time (sub-seconds) in main memory. As a consequence, the resulting technique MOVIES supports at the same time high query rates and high update rates and trades this for query result staleness. Moreover, MOVIES is the first main memory method supporting time-parameterized predictive queries. To support this feature we present two algorithms: non-predictive MOVIES and predictive MOVIES. We obtain the surprising result that a predictive indexing approach -considered state-of-the-art in an external-memory scenario -does not scale well in a main memory environment. In fact our results show that MOVIES outperforms state-of-the-art moving object indexes like a main-memory adapted B x -tree by orders of magnitude w.r.t. update rates and query rates. Finally, our experimental evaluation uses a workload unmatched by any previous work. We index the complete road network of Germany consisting of 40,000,000 road segments and 38,000,000 nodes. We scale our workload up to 100,000,000 moving objects, 58,000,000 updates per second and 10,000 queries per second which is unmatched by any previous work.

Spatial joins in main memory

“…It falls behind even a baseline approach, termed Binary Search, where the data points are sorted by one coordinate, upon which a nested loop with binary search (on the sorted coordinate) is used to compute the join. There is no clear winner: the best-performing approaches, namely R-Tree [4,6], CR-Tree [5], and Linearized KD-Trie [3], exhibit very similar performance.…”

“…For example, two dimensional coordinates are often encoded as two 4-byte single-precision or integer values [2,3,7,8].…”

“…Here, a static index is built, and index nested loops are used to compute join results. The following four static indexes belong to this category: R-Tree [4,6], CR-Tree [5], Linearized KD-Trie [3], and Static Grid [8]. This category reports the best performance results on average [7].…”

Thermal-Join

“…Updating the index, however, is very costly and it is typically cheaper to rebuild the index from scratch at every iteration (and to throw away the index after leading to short lived throw-away indexes [8]). Space-oriented partitioning indexes are particularly efficient in building an index as they use spatial grids or hierarchical space decomposition to index the objects.…”

Behavioral simulations in MapReduce

“…We will use this observation next to parallelize behavioral simulations efficiently in the MapReduce framework. As we will see in Section 4, the observation also enables us to apply efficient database indexing techniques, e.g., [18,37].…”