Applying the Effective Distance to Location-Dependent Data

Abstract

In wireless mobile environments, we increasingly use data that depends on the location of mobile clients. However, requested geographical objects (GOs) do not exist in all areas with uniform distribution. More urbanized areas have greater population and greater GO density. Thus the results of queries may vary based on the perception of distance. We use urbanization as a criterion to analyze the density of GOs. We propose the Effective Distance (ED) measurement, which is not a physical distance but the perceived distance varying based on the extent of urbanization. We present the efficiency of supporting location-dependent data on GOs with proposed ED. We investigate several membership functions to establish this proposed ED based on the degree of urbanization. In our evaluation, we show that the z-shaped membership function can flexibly adjust the ED. Thus, we obtain improved performance to provide the location-dependent data because we can differentiate the ED for very densely clustered GOs in urbanized areas.