Scalable Generation of Synthetic GPS Traces with Real-Life Data Characteristics

Abstract

Abstract. Database benchmarking is most valuable if real-life data and workloads are available. However, real-life data (and workloads) are often not publicly available due to IPR constraints or privacy concerns. And even if available, they are often limited regarding scalability and variability of data characteristics. On the other hand, while easily scalable, synthetically generated data often fail to adequately reflect real-life data characteristics. While there are well established synthetic benchmarks and data generators for, e.g., business data (TPC-C, TPC-H), there is no such up-to-date data generator, let alone benchmark, for spatiotemporal and/or moving objects data. In this work, we present a data generator for spatiotemporal data. More specifically, our data generator produces synthetic GPS traces, mimicking the GPS traces that GPS navigation devices generate. To this end, our generator is fed with real-life statistical profiles derived from the user base and uses real-world road network information. Spatial scalability is achieved by choosing statistics from different regions. The data volume can be scaled by tuning the number and length of the generated trajectories. We compare the generated data to real-life data to demonstrate how well the synthetically generated data reflects real-life data characteristics.

Waste not… Efficient co-processing of relational data

“…To evaluate the approach, we used two benchmarks: the spatial range query benchmark [19] that we used to evaluate the BWD prototype [8] and a representative subset of the TPC-H benchmark. As mentioned before, we believe that with other, orthogonal optimization of individual operators, the performance can be improved further.…”

Supply and Demand Aware Synthetic Data Generation for On-demand Traffic with Real-world Characteristics

“…In these generators, the origin and destination nodes of a trip are selected based on a uniform distribution or by relying on user input weights for the di erent regions. In [5], the generator mimics real-life GPS traces by using statistical pro les derived from historical GPS data. MNTG [21] proposed an extensible web-based tra c generator that acts as a wrapper over existing generators, but allow users to easily con gure and use them for arbitrary spatial road regions.…”

“…In addition, micro-scale simulators [17] have been proposed for simulating vehicle drivers with di erent driving styles and accounting for parameters such as tra c lights for studying the evolution of tra c in a network. These simulators focus on the accuracy of the proposed models rather than scalability to large data volumes [5].…”

“…online check-ins [33]). Research on tra c generators [5,6,10,15,17,21,29,35] has addressed the problem of generating synthetic spatio-temporal datasets for various applications such as route nding, object tracking, and motion prediction. However, these generators cannot be used for studying the problem of supply-demand imbalance since they focus only on the demand side.…”