Search

Abstract: A very important problem in many wireless ad-hoc networks, including wireless sensor networks, is positioning or the determination of geographical locations of the wireless nodes. Positioning is used both in infrastructural aspects of sensor networks, like geographic routing and topology maintenance, and in applications like wildlife tracking. Connectivity-based positioning algorithms in mobile wireless systems are studied in this work. These algorithms compute node positions based only on the connectivity, i.e. the neighborhood information of each node. Many algorithms have been proposed for positioning in stationary node systems and bounds on positional error of algorithms have been derived. The design and analysis of positioning algorithms for mobile node systems is a more challenging problem. Node mobility increases the amount of positional information available to a positioning algorithm. The work in this paper establishes a bound on the positional error for connectivity-based algorithms in mobile systems. The formulation from the analysis is used to investigate the benefit of this additional positional information on reducing positional error. There is a limit to the usefulness of positional information from previous node positions due to movement. This captures an important performance tradeoff: historical positional information can yield reduced positional error but requires more connectivity information from the network which requires greater computational resources. [Copyright &y& Elsevier]

• Severity marked failures in an operating system over time are presented as a stochastic risk process. • Risk processes are compared using orders of stochastic dominance. • The components (rate, mean severity, severity exceedance) are dimensions in a probability consequence diagram (PCD). • The PCD is presented as a Risk Cube. • The methods are illustrated with data on operational incidents in an Airport Navigational System. The failure events for a system constitute a marked point process, with severity marks defined by the consequences of failure. In this paper the process is summarized with the triplet: {failure rate, severity mean, severity exceedance}. Those component measures locate the process status in a 3 - dimensional probability consequence diagram. If the dimensions are stratified into intervals, the diagram is displayed as a risk cube. Composite risk measures, which are consistent with orders of stochastic dominance, are defined by integration of the risk process. The composite measures will determine equivalence classes of risk processes, and benchmarks for the measures mark regions of high/low risk on the cube. The risk cube methodology is applied to incident data for an airport navigation system. The location of an individual airport triplet on a risk cube, with regions determined by a family of airports, illustrates the potential for identifying unusual performance in terms of failure events. [ABSTRACT FROM AUTHOR]

The purpose of this paper is to develop safety performance measures and test the measures on data for air traffic management failure events. Failure events are classified by the severity of the consequence of occurrence, resulting in the rate of occurrence in severity categories. The safety measures are standard statistics calculated from this “distribution” for comparison of airport operations by stochastic ordering. For comparisons a benchmark is developed from the aggregation of failure data on a set of comparable airports. Airport performance is then compared with the benchmark using the defined safety measures. The benchmark comparison was implemented with failure data for major airports in Canada from 2005 to 2012. The results show a number of patterns and anomalies and some airports perform poorly in comparison to a class of similar operations. We conclude by suggesting benchmarking safety measures as a natural addition to the information system on aviation safety compiled by a national regulatory body to unravel anomalies such as implementation problems of a safety management system. [ABSTRACT FROM AUTHOR]

Wireless sensor networks (WSN) have great potential in ubiquitous computing. However, the severe resource constraints of WSN rule out the use of many existing networking protocols and require careful design of systems that prioritizes energy conservation over performance optimization. A key infrastructural problem in WSN is localization-the problem of determining the geographical locations of nodes. WSN typically have some nodes called seeds that know their locations using global positioning systems or other means. Non-seed nodes compute their locations by exchanging messages with nodes within their radio range. Several algorithms have been proposed for localization in different scenarios. Algorithms have been designed for networks in which each node has ranging capabilities, i.e., can estimate distances to its neighbours. Other algorithms have been proposed for networks in which no node has such capabilities. Some algorithms only work when nodes are static. Some other algorithms are designed specifically for networks in which all nodes are mobile. We propose a very general, fully distributed localization algorithm called range-based Monte Carlo boxed (RMCB) for WSN. RMCB allows nodes to be static or mobile and that can work with nodes that can perform ranging as well as with nodes that lack ranging capabilities. RMCB uses a small fraction of seeds. It makes use of the received signal strength measurements that are available from the sensor hardware. We use RMCB to investigate the question: 'When does range-based localization work better than range-free localization?' We demonstrate using empirical signal strength data from sensor hardware (Texas Instruments EZ430-RF2500) and simulations that RMCB outperforms a very good range-free algorithm called weighted Monte Carlo localization (WMCL) in terms of localization error in a number of scenarios and has a similar computational complexity to WMCL. We also implement WMCL and RMCB on sensor hardware and demonstrate that it outperforms WMCL. The performance of RMCB depends critically on the quality of range estimation. We describe the limitations of our range estimation approach and provide guidelines on when range-based localization is preferable. [ABSTRACT FROM AUTHOR]