Modeling and Engineering Algorithms for Mobile Data

Abstract

Summary. In this paper, we present an object-oriented approach to modeling mobile data and algorithms operating on such data. Our model is general enough to capture any kind of continuous motion while at the same time allowing for encompassing algorithms optimized for specific types of motion. Such motion may be available in a specific form, e.g., described by polynomials or splines, or implicitly restricted using bounds for speed or acceleration given by the application context.

A critical evaluation of location based services and their potential

“…Accordingly, a number of studies have examined how spatio-temporal databases can handle streams of location data from mobile devices. Blunck et al (2006) abstracted the problem by suggesting that the stream of positional data should be converted to a kinetic data structure for efficiency, and so a class of algorithms can be developed to handle it, e.g. to provide constraints or services such as collision detection.…”

Energy-efficient trajectory tracking for mobile devices

“…When the next heading is measured, the orthogonal distance is recalculated using the previous GPS speed reading and the difference between the two most recent heading measurements (2). When the new and previous heading measurements match, the orthogonal distance remains constant (3). Whenever the heading changes cause the trajectory error threshold to be violated, a new GPS position is requested (4).…”

“…This value was suitable for tracking with most transportation modes in our experiments. Additional knowledge about the motion abilities and patterns of the tracked target and the expected irregularity of its movement can be exploited to more specifically adapt the update time Δt based on the current estimated orthogonal distance D orth (t k ) and measured heading θ k , see also [3]. Note also, that summing in Equation 1 over signed instead of absolute heading changes, would allow converse heading changes to cancel each other out, thereby leading to later update times Δt and thus higher energy savings -at the expense, though, of a higher probability to overlook the violation of the error threshold as a result of heading measurement errors.…”

“…The power models have been derived with the help of the Nokia Energy Profiler. 3 The process for generating the power models is described in [13] and the parameter values of the power models for the Nokia N97 are described in [11].…”

“…For the representation of trajectories usually piecewise-linear curves in space and time are assumed. Alternative representations, though, such as by by probability functions and by non-linear functions have been discussed to reflect the uncertainty inherent in monitoring motion as well as the non-linearity of motion [21,23,2,3]. Furthermore, various models and algorithms for the simplification of trajectories and their impacts on the accuracy of query results and data derived from trajectories, have been proposed [4,17,2].…”