Marionette: using RPC for interactive development and debugging of wireless embedded networks

Abstract

Quantum optical coherence tomography (QOCT) makes use of an entangled twin-photon light source to carry out axial optical sectioning. We have probed the longitudinal structure of a sample comprising multiple surfaces in a dispersion-cancelled manner while simultaneously measuring the group-velocity dispersion of the interstitial media between the reflecting surfaces. The results of the QOCT experiments are compared with those obtained from conventional optical coherence tomography (OCT). 4145-4154 (1995). 13. L. Mandel and E. Wolf, Optical Coherence and Quantum Optics (Cambridge, New York, 1995), ch. 22. 14. C. K. Hong, Z. Y. Ou, and L. Mandel, "Measurement of subpicosecond time intervals between two photons by interference," Phys. Rev. Lett. 59, 2044Lett. 59, -2046Lett. 59, (1987

Visibility Levels: Managing the Tradeoff between Visibility and Resource Consumption

“…To improve repeatability of distributed event-driven applications, EnviroLog [8] provides an event recording and replay service that helps users to debug problems that are difficult to reproduce. Tools such as Marionette [17] support logging and retrieval of runtime state for fault diagnosis. Clairvoyant [18] is an interactive source-level debugger for sensor nodes.…”

Towards a Versatile Problem Diagnosis Infrastructure for LargeWireless Sensor Networks

“…Existing problem diagnosis infrastructures for WSNs [2,8,13] proved invaluable for deploying and maintaining relatively small, real-world networks running simple, well-studied applications. However, they were not designed for rapidly emerging systems, characterized by the increasing complexity and scale, and often operating in unfamiliar settings.…”

“…However, they were not designed for rapidly emerging systems, characterized by the increasing complexity and scale, and often operating in unfamiliar settings. Although all solutions enable remote monitoring of application metrics, they require the programmer to foresee the important ones prior to the deployment [2,8], or put constraints on how the metric values are obtained [2,13]. Thus, many unanticipated problems cannot be diagnosed.…”

“…Thus, many unanticipated problems cannot be diagnosed. Moreover, metric values are collected per node, and then processed at a base station [2,8,13]. This lack of support for in-network processing and dynamic collaboration of nodes limits scalability and restricts applicability to certain problems.…”

“…None of the aforementioned diagnostic infrastructures for WSNs [2,8,11,13] addresses all these issues. The issues are, however, crucial for large-scale systems.…”