The objective of this research is to develop an integrated data gathering mechanism that utilizes new physical layer techniques and a non-conventional energy model for data gathering in wireless sensor-actuator networks. The project combines protocol and algorithm design with test bed implementation and evaluation.
Intellectual Merit: This research addresses several closely coupled issues in data gathering in wireless sensor-actuator networks. First, the research considers how to plan the moving path of a mobile actuator based on the distribution of sensors and load balancing among sensors to prolong network lifetime. Second, the research considers how to utilize special properties of sensor node batteries to improve energy-efficiency. Third, the research considers how the new physical layer techniques, in particular multiple packet transmission, can be applied to improve data throughput and reduce energy consumption in sensor-actuator networks. The project seeks to develop novel cross-layer algorithms that have the potential to greatly increase data throughput and network lifetime. A wireless sensor test bed is used to further validate feasibility and performance of the algorithms that are developed. The research hopes to impact fundamental design principles and infrastructure considerations for future sensor-actuator networks.
Broader Impacts: The outcome of this research has the potential to be applicable to a wide spectrum of applications, including environmental, military, space, health care, home and other commercial areas. A project goal is to promote the participation of female engineering students. The test bed developed as part of the project is an important facility for related projects. The important findings of this project are to be disseminated to the research community via conference and journal publications, a web site, and industry-university partnerships.
In this project, we consider a wireless sensor network consisting of a large number of sensors and a limited number of mobile data collectors. In such a network, mobile collectors take over the burden of routing from sensors, roaming over the sensing area and collecting data from nearby sensors via short-range wireless communications. We have designed a series of efficient mobile data gathering schemes and techniques in such sensor networks, which aim to prolong network lifetime and shorten data gathering latency. Moving path planning with multi-hop relays. We propose a moving path planning algorithm by adopting a divide and conquer method, which recursively determines a turning point on the path. The moving path of the mobile collector is planed dynamically based on the distribution of sensors, and load balancing among sensors is performed along with the moving path planning to prolong network lifetime. Single-hop data gathering. To achieve uniform energy consumption among sensors, in this scheme, the mobile collector is scheduled to traverse the transmission range of each sensor such that data from each sensor can be collected via single-hop transmission. However, this approach typically results in significantly increased latency due to the low moving velocity of the mobile collector. Hence, we focus on minimizing the length of a data gathering tour by formulating it into an optimization problem. A heuristic algorithm is proposed to provide a practically good solution to the problem. Mobile data gathering with controlled mobility and SDMA technique. In this scheme, we apply the latest physical layer technique, Space-Division-Multiple-Access (SDMA), to sensor networks, which enables multiple sensors to upload data simultaneously to the mobile collector so that data uploading time can be greatly shortened. To better enjoy the benefit of SDMA, mobile collector may have to visit some specific locations where more sensors are compatible, which may adversely prolong the moving tour. We propose an optimal solution that minimizes data gathering latency by exploring a tradeoff between the shortest moving tour and full utilization of SDMA. We also propose a region-division and tour-planning algorithm when multiple collectors are employed. Bounded relay hop mobile data gathering scheme. In this scheme, we study the inherent tradeoff between energy saving and data gathering latency of mobile data gathering, by achieving a balance between the relay hop count of local data aggregation and the moving tour length of the mobile collector. We propose a polling-based mobile collection approach and formulate it into an optimization problem. Specifically, a subset of sensors are selected as polling points that buffer the locally aggregated data and upload the data to the mobile collector when it arrives. In the meanwhile, when sensors are affiliated with these polling points, it is guaranteed that the relaying of any packet is bounded within a given number of hops. Distributed optimization algorithms. To achieve optimal performance for anchor-based mobile data gathering, we formulate two optimization problems in a utility maximization framework, which considers fixed or variable sojourn time of the collector at each anchor point. Based on dual decomposition and subgradient methods, we design distributed algorithms to solve these problems. Extensive numerical results demonstrate that the algorithms can achieve fast convergence to reach optimality. Layered mobile data collection. To achieve scalability, long network lifetime and low data collection latency, we propose a three-layer (sensor, cluster head and mobile collector) framework called LBC-MU which employs distributed load balanced clustering and MIMO uploading techniques. Lossy wireless sensor networks. We develop distributed clustering algorithms to address lossy wireless links in sensor networks. Simulations conducted under a realistic link model demonstrate that the proposed clustering algorithm can reduce total energy consumption and prolong network lifetime significantly compared to algorithms that do not consider the lossy nature of wireless links. Data compression. We apply compression to reduce data size and sensitive data delay in data collection. Based on the observation that compression does not always reduce packet delay due to various network and hardware configurations, we design an adaptive algorithm to make on-line decisions such that compression is only performed when it improves overall performance. Experimental results show that the algorithm can adapt to network dynamics and maximize compression benefit. Wireless testbed. We develop a wireless testbed to validate data gathering mechanisms and algorithms designed for wireless sensor networks. The testbed has two types of devices: wireless node and mobile collector. Distinguished features of the wireless node include: reliable multi-rate wireless transmission for different applications; enhanced computing capability provided by an ultra-low power microcontroller and a flash-based FPGA; extended memory space for data buffering; low power consumption to ensure long-term operation; localization with the assistance of GPS modules and received signal strength indicator; in-field upgrading through the cooperation of microcontroller and FPGA. With CPU+FPGA architecture for the computing module and industry standard interfaces, the testbed can be easily utilized for other research work.
