Location Based Services

Location Based Services project targets to construct a data analysis platform for location information. This platform is a kind of multi-dimensional database. It is designed to handle large amounts of data and high data insertion rates, while utilizing several characteristic properties of location information to achieve high performance and usability.

Challenges:

  1. Scalability
    First of all, these platforms should be scalable because the volume of location data grows larger and larger with time. Each location data item itself is very small in size, but total number of data points is huge.
  2. Performance
    1. Sustain high insertion rates
      Collecting location data from massive mobile phones/sensor nodes is one of big challenge. Thus, the system should be capable of handling upload requests while achieving high insertion throughput.
    2. Query processing
      Each location data point has very little information. When we utilize location data, we extract statistical properties or summaries from the dataset. Range queries and time series queries are basic and/or important queries for location based data analysis. Our first goal is to optimize these queries by introducing scalable index structures.
    3. Real-time query support
      Most applications are interested in the latest data. So being able to analyze data at it is being uploaded into the system is critical for improving the utility of the applications.
  3. Load Balancing
    Location data have a heavy spatial and/or temporal skew. For example, in large cities like New York, Los Angeles, huge amounts of location data are generated; while in rural areas, a little amount of data is generated. Overcoming skew in data distribution is also a big challenge for effective load balancing.
Project Status: 
Active