arXiv subjects

Andreas Terzis

Publications and source records attributed to Andreas Terzis.

25 records · Page 2Linked to original sources

arXiv2019

The Perils of Detecting Measurement Faults in Environmental Monitoring Networks

Scientists deploy environmental monitoring net-works to discover previously unobservable phenomena and quantify subtle spatial and temporal differences in the physical quantities they measure. Our experience, shared by others, has shown that measurements gathered by such networks are perturbed by sensor faults. In response, multiple fault detection techniques have been proposed in the literature. However, in this paper we argue that these techniques may misclassify events (e.g. rain events for soil moisture measurements) as faults, potentially discarding the most interesting measurements. We support this argument by applying two commonly used fault detection techniques on data collected from a soil monitoring network. Our results show that in this case, up to 45% of the event measurements are misclassified as faults. Furthermore, tuning the fault detection algorithms to avoid event misclassification, causes them to miss the majority of actual faults. In addition to exposing the tension between fault and event detection, our findings motivate the need to develop novel fault detection mechanisms which incorporate knowledge of the underlying events and are customized to the sensing modality they monitor.

eess.SP↗

arXiv2019

Phoenix: An Epidemic Approach to Time Reconstruction

Harsh deployment environments and uncertain run-time conditions create numerous challenges for postmortem time reconstruction methods. For example, motes often reboot and thus lose their clock state, considering that the majority of mote platforms lack a real-time clock. While existing time reconstruction methods for long-term data gathering networks rely on a persistent basestation for assigning global timestamps to measurements, the basestation may be unavailable due to hardware and software faults. We present Phoenix, a novel offline algorithm for reconstructing global timestamps that is robust to frequent mote reboots and does not require a persistent global time source. This independence sets Phoenix apart from the majority of time reconstruction algorithms which assume that such a source is always available. Motes in Phoenix exchange their time-related state with their neighbors, establishing a chain of transitive temporal relationships to one or more motes with references to the global time. These relationships allow Phoenix to reconstruct the measurement timeline for each mote. Results from simulations and a deployment indicate that Phoenix can achieve timing accuracy up to 6 ppm for 99% of the collected measurements. Phoenix is able to maintain this performance for periods that last for months without a persistent global time source. To achieve this level of performance for the targeted environmental monitoring application, Phoenix requires an additional space overhead of 4% and an additional duty cycle of 0.2%.

cs.NI↗

arXiv2019

Sundial: Using Sunlight to Reconstruct Global Timestamps

This paper investigates postmortem timestamp reconstruction in environmental monitoring networks. In the absence of a time-synchronization protocol, these networks use multiple pairs of (local, global) timestamps to retroactively estimate the motes' clock drift and offset and thus reconstruct the measurement time series. We present Sundial, a novel offline algorithm for reconstructing global timestamps that is robust to unreliable global clock sources. Sundial reconstructs timestamps by correlating annual solar patterns with measurements provided by the motes' inexpensive light sensors. The surprising ability to accurately estimate the length of day using light intensity measurements enables Sundial to be robust to arbitrary mote clock restarts. Experimental results, based on multiple environmental network deployments spanning a period of over 2.5 years, show that Sundial achieves accuracy as high as 10 parts per million (ppm), using solar radiation readings recorded at 20 minute intervals.

math.NA↗

arXiv2016

High Frequency Remote Monitoring of Parkinson's Disease via Smartphone: Platform Overview and Medication Response Detection

Objective: The aim of this study is to develop a smartphone-based high-frequency remote monitoring platform, assess its feasibility for remote monitoring of symptoms in Parkinson's disease, and demonstrate the value of data collected using the platform by detecting dopaminergic medication response. Methods: We have developed HopkinsPD, a novel smartphone-based monitoring platform, which measures symptoms actively (i.e. data are collected when a suite of tests is initiated by the individual at specific times during the day), and passively (i.e. data are collected continuously in the background). After data collection, we extract features to assess measures of five key behaviors related to PD symptoms -- voice, balance, gait, dexterity, and reaction time. A random forest classifier is used to discriminate measurements taken after a dose of medication (treatment) versus before the medication dose (baseline). Results: A worldwide study for remote PD monitoring was established using HopkinsPD in July, 2014. This study used entirely remote, online recruitment and installation, demonstrating highly cost-effective scalability. In six months, 226 individuals (121 PD and 105 controls) contributed over 46,000 hours of passive monitoring data and approximately 8,000 instances of structured tests of voice, balance, gait, reaction, and dexterity. To the best of our knowledge, this is the first study to have collected data at such a scale for remote PD monitoring. Moreover, we demonstrate the initial ability to discriminate treatment from baseline with 71.0(+-0.4)% accuracy, which suggests medication response can be monitored remotely via smartphone-based measures.

cs.CY↗

arXiv2014

Hadoop in Low-Power Processors

In our previous work we introduced a so-called Amdahl blade microserver that combines a low-power Atom processor, with a GPU and an SSD to provide a balanced and energy-efficient system. Our preliminary results suggested that the sequential I/O of Amdahl blades can be ten times higher than that a cluster of conventional servers with comparable power consumption. In this paper we investigate the performance and energy efficiency of Amdahl blades running Hadoop. Our results show that Amdahl blades are 7.7 times and 3.4 times as energy-efficient as the Open Cloud Consortium cluster for a data-intensive and a compute-intensive application, respectively. The Hadoop Distributed Filesystem has relatively poor performance on Amdahl blades because both disk and network I/O are CPU-heavy operations on Atom processors. We demonstrate three effective techniques to reduce CPU consumption and improve performance. However, even with these improvements, the Atom processor is still the system's bottleneck. We revisit Amdahl's law, and estimate that Amdahl blades need four Atom cores to be well balanced for Hadoop tasks.

cs.DC↗

arXiv2009

Model-Based Event Detection in Wireless Sensor Networks

In this paper we present an application of techniques from statistical signal processing to the problem of event detection in wireless sensor networks used for environmental monitoring. The proposed approach uses the well-established Principal Component Analysis (PCA) technique to build a compact model of the observed phenomena that is able to capture daily and seasonal trends in the collected measurements. We then use the divergence between actual measurements and model predictions to detect the existence of discrete events within the collected data streams. Our preliminary results show that this event detection mechanism is sensitive enough to detect the onset of rain events using the temperature modality of a wireless sensor network.

cs.NI↗

arXiv2007

Life Under Your Feet: An End-to-End Soil Ecology Sensor Network, Database, Web Server, and Analysis Service

Wireless sensor networks can revolutionize soil ecology by providing measurements at temporal and spatial granularities previously impossible. This paper presents a soil monitoring system we developed and deployed at an urban forest in Baltimore as a first step towards realizing this vision. Motes in this network measure and save soil moisture and temperature in situ every minute. Raw measurements are periodically retrieved by a sensor gateway and stored in a central database where calibrated versions are derived and stored. The measurement database is published through Web Services interfaces. In addition, analysis tools let scientists analyze current and historical data and help manage the sensor network. The article describes the system design, what we learned from the deployment, and initial results obtained from the sensors. The system measures soil factors with unprecedented temporal precision. However, the deployment required device-level programming, sensor calibration across space and time, and cross-referencing measurements with external sources. The database, web server, and data analysis design required considerable innovation and expertise. So, the ratio of computer-scientists to ecologists was 3:1. Before sensor networks can fulfill their potential as instruments that can be easily deployed by scientists, these technical problems must be addressed so that the ratio is one nerd per ten ecologists.

cs.DB↗