  |
|
Contents: I am currently working in the Collaborative Embedded Processing, Actuation, and Networked Sensing (CEPANS) lab. The focus of my work is wirless sensor networking. Specifically I am looking into many-to-many routing of streaming media in large scale, low power, low bandwidth, heterogenous networks. Hardware Platform: Currently in CEPANS we are using Freescale Semiconductor's MC13192-EVB and SARD platforms. The core of the platforms is the MC9S08GT60 8-bit microcontroller and the MC13192 802.15.4 Zigbee compatible transceiver. The microcontroller can operate at 1.8V-3V and from 64kHz-40 MHz. It has 60KB of flash memory and 4KB of ram. The microcontroller is capable of 4 different low power modes with the lowest consuming only about 500 nA. It has dual 2-channel 16-bit interrupt timers and an 8-channel 10-bit analog to digital converter. The transciever has a range of about 600m with line of sight and has a rated throughput of 250Kbps. Both the SARD and EVB platforms are equipped with an RS-232 interface capable of up to 38400 baud. The EVB additionally has a USB serial interface. The SARD platform is equipped with a MMA1260D z-axis accelerometer and a MMA6260Q x and y-axis accelerometer.   More Information: I recently gave a seminar summarizing my research for the University of Wisconsin Madison Computer Engineering Department entitled: Experimental Evaluation and Demonstration of Low-Power Zigbee Wireless Sensor Networking Platforms (ppt) Abstract: Recently wireless sensor networks have become a hot area of research and development due to their large number of applications, as well as tremendous industrial, economic, and scientific potentials. To keep up with this trend, it is important to have experimental platforms for development, analysis, and verification, that are cost effective, flexible, extensible, easy to use, and more importantly, representative of actual future deployments. In this talk, I will first present a brief overview of the current state of the art in terms of wireless sensor networking platforms, the problems, and the challenges ahead. The overview will be followed by the motivation and a detailed investigations of the design, use, and evaluation of our Zigbee-based platform at UW Madison. I will present details of our ongoing efforts in utilizing and analyzing the performance of the hardware provided by Freescale Corp. and other manufacturers, our system software, available development tools and technologies, and live demonstrations of several applications to exercise the various components. The applications presented include accelerometer-based location tracking of a mobile node, distributed video surveillance and tracking, and multimedia appliance control. The following is a short paper I wrote as part of my Ph.D. qualifying exam: Heterogeneity in Large Scale Low-Power Ad-Hoc Wireless Sensor Networks (pdf) Abstract: Wireless sensor networks have recently become a growing area of research and development due to the tremendous number of applications that can greatly benefit from such systems. In the past the majority of research efforts have implicitly assumed homogenous networks. While this is true of certain applications, in this paper I will motivate the use of heterogeneous wireless sensor networks, provide some example applications from my current research where homogeneity is clearly the incorrect model, illustrate some issues I have encountered with heterogeneity in wireless sensor networks, review previous work related to these issues, and finally present a detailed description and approach to the issue of routing in high-throughput applications operating on low-bandwidth wireless platforms. This was my final masters paper summarizing the work I had done while working on my masters degree: Experimental Evaluation of Low-Power Zigbee Wireless Sensor Networks (pdf) Abstract: Recently wireless sensor networks have become a growing area of research and development due to the tremendous number of applications that can greatly benefit from such systems. To keep up with this trend, it is important to have experimental platforms for analysis and verification, that are cost effective, flexible, and easy to use. In this project, we investigate the experimental use and evaluation of a Zigbee-based wireless sensor network platform. We begin by presenting the details of the hardware and system software used (e.g. MAC and routing) and a number of applications to exercise the various components. The applications include accelerometer-based location tracking of a mobile node, video surveilance, and multimedia appliance control. We also present several experimentally observed power consumption data, where applicable. Route Mapping Tool: In order to gather useful information about the network topology and to help debug our implementation of the Bellman-Ford routing algorithm, I developed a tool to map in real time the current routes in the network. The routes are mapped from a root node that the pc plugs into via RS-232 to any other nodes in the network. The tool then displays these routes as a graphical tree. This is accomplished as follows, upon startup the tool requests the routing table of the node it is connected to. From this table it is able to make a tree of all 1-hop and 2-hop routes and also knows what nodes are in the network. After recieving the table the tool then begins requesting routes for nodes more than 2-hops from the root node. This message is recursively passed down the route until it is 2-hops away from the end node, with each node filling itself into the original request as one of the hops. When the message is 2-hops away from the end node, the node in possession of the message fills in itself, the next-hop and the end node into the message, and then sends the message back to the root node. The root node passes the message back to the mapping tool and the mapping tool adds the route to the tree. When the tool has a map of all the nodes in the network, it starts over requesting routes that are greater than 2-hops in case the network topology changes. If the master node's routing table changes in any way it resends it to the tool and the tool restarts route mapping. The last important thing to note is there is actually an intermediate tool that takes UDP packets from the pc-side tool and converts them to serial packets for the node-side application and vice-versa. Sources and binaries (note these are not final releases): Route Mapping Source for Boards (zip)(Codewarrior IDE 3.0 Project) Route Mapping Binary for Boards (zip)(S19 File) Route Mapping Source for PC (zip)(Visual C++ 6.0 Project) Route Mapping Binary for PC (zip)(EXE file) Screenshots:  
|
