Victoria Manfredi

Learning hierarchical models of activity

This paper investigates learning hierarchical statistical activity models in indoor environments. The abstract hidden Markov model (AHMM) is used to represent behaviors in stochastic environments. We train the model using both labeled and unlabeled data and estimate the parameters using expectation maximization (EM). Results are shown on three datasets: data collected in lab, entryway, and home environments. The results show that hierarchical models outperform flat models.

Switching kalman filters for prediction and tracking in an adaptive meteorological sensing network

We consider the problem of configuring sen- sors in an adaptive sensor network being used to monitor meteorological features. One way to decide future sensor configurations is to base them on information currently being collected. For instance, if a meteorological sensor network is being used to monitor storms in Oklahoma, then the sensors could be dynamically configured based on the predicted storm locations. While Kalman filters and their extensions are commonly used for prediction and tracking, they have been primarily applied to objects with known or fixed dynamics such as missiles or people. We explore the advantages and limitations of using Kalman filters to track objects with nonstationary dynamics (e.g., a storm can grow in size). In particular, we focus on tracking meteorological features over time with the objective of using this information to determine where radars should focus their sensing. We present results for tracking storm cells comparing least-squares regression with Kalman filter and switching Kalman filter methods. Our results show that on average the Kalman filter methods better predict the future location of a storm centroid than does a least-squares regression algorithm currently in use for meteorological storm tracking.

Separation of Sensor Control and Data in Closed-Loop Sensor Networks

Sensor networks are prone to congestion due to bursty and high-bandwidth data traffic, combined with wireless links and many-to-one data routing to a sink. Delayed and dropped packets then degrade the performance of the sensing application. In this paper, we investigate the value of separate handling of sensor control and data traffic, during times of congestion, in a closed-loop sensor network. We first show that prioritizing sensor control traffic over data traffic decreases the round-trip control-loop delay, and consequently increases the quantity and quality of the data collected by the sensor network. We then ground our analysis in a closed-loop meteorological sensor network, focusing on a storm-tracking application running over a network of X-band radars. Our application measures reflectivity (a measure of the number of scatterers in a unit volume of atmosphere known as a voxel) and tracks storms (i.e., regions of high reflectivity) using a Kalman filter. Considering data quantity, we show that prioritizing sensor control traffic increases the number of voxels, V, that can be scanned given a constant number of reflectivity samples, Nc, obtained per voxel. Here, utility increases linearly with the number of scanned voxels. Considering data quality, we show that prioritizing sensor control traffic increases the number of reflectivity samples, N, that can be obtained per voxel given a constant number of voxels, Vc, to scan. Here, since sensing accuracy improves only as a function of √N, the gain in accuracy for the reflectivity estimate per voxel as N increases is relatively small except when prioritizing sensor control increases N significantly (such as when sensor control packets suffer severe delays). Because accuracy also degrades as a function of √N, however, and because prioritizing sensor control traffic reduces the number of control packets dropped, data degradation is mitigated. Considering the performance of the tracking application, we then show that during times of severe congestion, not prioritizing sensor control can actually lead to tracking errors accumulating over time.