Pervasive computing software adapts its behavior according to the changing contexts. Nevertheless, contexts are often noisy. Context inconsistency resolution provides a cleaner pervasive computing environment to context-aware applications. A faulty context-aware application may, however, mistakenly mix up inconsistent contexts and resolved ones, causing incorrect results. This paper studies how such faulty context-aware applications may be affected by these services. We model how programs should handle contexts that are continually checked and resolved by context inconsistency resolution, develop novel sets of data flow equations to analyze the potential impacts, and thus formulate a new family of test adequacy criteria for testing these applications. Experimentation shows that our approach is promising.

Two complications frequently arise in real-world applications, motion and the contamination of data by outliers. We consider a fundamental clustering problem, the k-center problem, within the context of these two issues. We are given a finite point set S of size n and an integer k. In the standard k-center problem, the objective is to compute a set of k center points to minimize the maximum distance from any point of S to its closest center, or equivalently, the smallest radius such that S can be covered by k disks of this radius. In the discrete k-center problem the disk centers are drawn from the points of S, and in the absolute k-center problem the disk centers are unrestricted. We generalize this problem in two ways. First, we assume that points are in continuous motion, and the objective is to maintain a solution over time. Second, we assume that some given robustness parameter 0 < t ≤ 1 is given, and the objective is to compute the smallest radius such that there exist k disks of this radius that cover at least ⌈tn ⌉ points of S. We present a kinetic data structure (in the KDS framework) that maintains a (3 + ε)-approximation for the robust discrete k-center problem and a (4 + ε)-approximation for the robust absolute k-center problem, both under the assumption that k is a constant. We also improve on a previous 8-approximation for the non-robust discrete kinetic k-center problem, for arbitrary k, and show that our data structure achieves a (4 + ε)-approximation. All these results hold in any metric space of constant doubling dimension, which includes Euclidean space of constant dimension.

We report on work in progress upon the verification of contextaware applications written in C-based languages. We recognize that context-aware programs are generally either middleware-based and multithreaded, or driven by asynchronous events, and focus on identifying the program points in which the contextual updates impact the application behaviour. Inheriting from related work on the validation of context-aware applications, we implement a technique for detecting context-aware program points over SatAbs, a tool for generic verification of specifications of ANSI C/C++ programs. We then briefly review future work with regard to the automatic verification of these programs

Abstract—Pervasive computing systems often use middleware as a means to communicate with the changing environment. However, the interactions with the context-aware middleware as well as the interactions among applications sharing the same middleware may introduce faults that are difficult to reveal by existing testing techniques. Our previous work proposed the notion of context diversity as a metric to measure the degree of changes in test inputs for pervasive software. In this paper, we present a case study on how much context diversity for test cases relates to fault-based mutants in pervasive software. Our empirical results show that conventional mutation operators can generate sufficient candidate mutants to support test effectiveness evaluation of pervasive software, and test cases with higher context diversity values tend to have higher mean mutation scores. On the other hand, for test cases sharing the same context diversity, their mutation scores can vary significantly in terms of standard derivations. Keywords—context diversity; mutation analysis; pervasive computing I.

Abstract not found

In this thesis, the theoretical analysis of real-world motivated problems regarding objects in motion is considered. Specifically, four major results are presented addressing the issues of robustness, data collection and compression, realistic theoretical analyses of this compression, and data retrieval. Robust statistics is the study of statistical estimators that are robust to data outliers. The combination of robust statistics and data structures for moving objects has not previously been studied. In studying this intersection, we consider a problem in the context of an established kinetic data structures framework (called KDS) that relies on advance point motion information and calculates properties continuously. Using the KDS model, we present an approximation algorithm for the kinetic robust k-center problem, a clustering problem that requires k clusters but allows some outlying points to remain unclustered. For many practical problems that inspired the exploration into robustness, the KDS model is inapplicable due to the point motion restrictions and the advanceflight plans required. We present a new framework for kinetic data that allows calculations on moving points via sensor-recorded observations. This new framework

• The average price of the top 50 paid