Electrical and Computer Engineering ETDs

Publication Date

1-31-2013

Abstract

Over the past few years, considerable progress has been made in the area of networked robotic systems and mobile sensor networks. The vision of a mobile sensor network cooperatively learning and adapting in harsh unknown environments to achieve a common goal is closer than ever. In addition to sensing, communication plays a key role in the overall performance of a mobile network, as nodes need to cooperate to achieve their tasks and thus have to communicate vital information in environments that are typically challenging for communication. Therefore, in order to realize the full potentials of such networks, an integrative approach to sensing (information gathering), communication (information exchange), and motion planning is needed, such that each mobile sensor considers the impact of its motion decisions on both sensing and communication, and optimizes its trajectory accordingly. This is the main motivation for this dissertation. This dissertation focuses on communication-aware motion planning of mobile networks in the presence of realistic communication channels that experience path loss, shadowing and multipath fading. This is a challenging multi-disciplinary task. It requires an assessment of wireless link qualities at places that are not yet visited by the mobile sensors as well as a proper co-optimization of sensing, communication and navigation objectives, such that each mobile sensor chooses a trajectory that provides the best balance between its sensing and communication, while satisfying the constraints on its connectivity, motion and energy consumption. While some trajectories allow the mobile sensors to sense efficiently, they may not result in a good communication. On the other hand, trajectories that optimize communication may result in poor sensing. The main contribution of this dissertation is then to address these challenges by proposing a new paradigm for communication-aware motion planning in mobile networks. We consider three examples from networked robotics and mobile sensor network literature: target tracking, surveillance and dynamic coverage. For these examples, we show how probabilistic assessment of the channel can be used to integrate sensing, communication and navigation objectives when planning the motion in order to guarantee satisfactory performance of the network in realistic communication settings. Specifically, we characterize the performance of the proposed framework mathematically and unveil new and considerably more efficient system behaviors. Finally, since multipath fading cannot be assessed, proper strategies are needed to increase the robustness of the network to multipath fading and other modeling/channel assessment errors. We further devise such robustness strategies in the context of our communication-aware surveillance scenario. Overall, our results show the superior performance of the proposed motion planning approaches in realistic fading environments and provide an in-depth understanding of the underlying design trade-off space.

Keywords

Wireless sensor networks., Mobile communication systems., Robotics.

Sponsors

NSF, ARO Collaborative Technology Alliance

Document Type

Dissertation

Language

English

Degree Name

Electrical Engineering

Level of Degree

Doctoral

Department Name

Electrical and Computer Engineering

First Advisor

Mostofi, Yasamin

First Committee Member (Chair)

Mostofi, Yasamin

Second Committee Member

Abdallah, Chaouki

Third Committee Member

Fierro, Rafael

Fourth Committee Member

Tapia, Lydia

Share

COinS