DEFINING AND MEASURING ROBUSTNESS IN WIRELESS SENSOR COMMUNICATION FOR TELEMEDICINE A Thesis Presented to The Graduate Faculty of The University of Akron In Partial Fulfillment of the Requirements for the Degree Master of Science Sudha Bhattarai August, 2008 ii DEFINING AND MEASURING ROBUSTNESS IN WIRELESS SENSOR COMMUNICATION FOR TELEMEDICINE Sudha Bhattarai Thesis Approved: Accepted: _______________________________ _______________________________ Advisor Dean of the College Dr. Kathy J. Liszka Dr. Ronald F. Levant _______________________________ _______________________________ Faculty Reader Dean of the Graduate School Dr. Timothy W. O'Neil Dr. George R. Newkome _______________________________ _______________________________ Faculty Reader Date Dr. Tim Marguish _______________________________ Department Chair Dr. Wolfgang Pelz iii ABSTRACT Wireless sensor networks are useful in a wide range of applications. A network used for a telemedicine application must be able to endure various noise factors present in the system. Dropped packets during mote communication are an important parameter that can be used to determine the robustness of the system. Robustness, in this context, is the degree to which a system is insensitive to small perturbations. We propose mote technology as a potential source for wireless communication of medical signals in body area networks. In order to determine if the proposed system is useful, a robustness metric is developed specifically for a telemedicine application. The contribution of this research is to provide a framework so that with further research, we can evaluate whether the mote hardware will deliver acceptable performance given perturbations in a wireless environment. iv ACKNOWLEDGEMENTS Thanks to Dr .Kathy Liszka for her immense support and direction in this project. Special thanks to Malinda J. Server for her help in understanding medical terms. Thanks to the committee member for their evaluation and comments on my work. I am also indebted to Mike Ritcher for making his experience with this research available to me. I would like to thank my husband Suraj Adhikari for his help in installation of motes. v TABLE OF CONTENTS Page LIST OF TABLES ………………………………………………………… …………viii LIST OF FIGURES………………………………………………………………………ix CHAPTER I. INTRODUCTION…………………………………………………………………….1 1.1. Overview…………………………………………………………………… …1 1.2. Eldercare .……………………………………………………………………… 2 1.3. The Holy Grail ………………………………………………………………… 3 1.4. Wireless Issues ………………………………………………………………….4 II. ROBUSTNESS ……………………………………………………………………….6 2.1. Performance Features ………………………………………………………… 6 2.2. Perturbation Parameters …………………………………………………………9 2.3. Impact of Perturbation Parameters on Performance Features …………………10 2.4. Analysis to Determine Robustness …………………………………………….10 2.5. Robust Research.……………………………………………………………….11 III. SYSTEM DESCRIPTION………………………………………………………… 13 vi 3.1. Hardware Description and Terminology……………………………………….13 3.1.1. Motes ……………………………………………………….……… 14 3.1.2. Mica2Dot ……………………………………………………………14 3.1.3. Mica2 ……………………………………………………………… 15 3.1.4. MIB510 …………………………………………………………… 16 3.2. Software Description and Terminology ……………………………………….17 3.2.1. TinyOS………………………………………………… …… …….18 3.2.2. nesC …………………………………………………… ……… ….18 3.2.3. Oscilloscope ……………………………………… ……… ………19 3.2.4. Blink ………………………………………………… … …………20 3.2.5. Serial Forwarder ……………………………… ……… ………… 20 3.2.6. Packet Listener ………………………………… …………… ……21 3.2.7. TOSBase ……………………………………………………….……21 3.2.8. OscilloscopeRF …………………………………… ……… …… 23 IV. PHYSICAL EXPERIMENTS …… …………………………………………….24 4.1. Objectives and limitations of the experiments ………………… ……………24 4.2. Preliminary experiments ……………………………………… …………….25 4.3. Communication issues …………………………………………… …………27 4.4. Data collection ……………………………………………………… ……….29 vii 4.5. Discussion …………………………………………………………………… 37 V. CONCLUSION AND FUTURE WORK……… ……………………………….… 38 REFERENCES ………………………………………………………………………….40 APPENDICES …………………………………………………………… ……………42 APPENDIX A. USER MANUAL ………………………………………… ………… 43 APPENDIX B. SERIALFORWARDER ……………………………………… …….51 APPENDIX C. PACKETLISTENER ……………………………….……… …………55 APPENDIX D. TOSBASE ………………………………….………………… ………57 APPENDIX E. OSCILLOSCOPERF ……………………………………… ………….62 APPENDIX F. DATA TABLES …………………………………………… …………65 viii LIST OF TABLES Table Page 4.1 Mote tests of one hour at 90 feet ……………………………… ………………36 F.1 Mote 1 readings for 1 minute ……………………………………… ………… 66 F.2 Mote 2 readings for 1 minute ……………………………………… ………….68 F.3 Mote 4 readings for 1 minute …………………………………………… …… 69 F.4 Mote 5 readings for 1 minute …………………………………………… …… 71 F.5 Mote 1 readings for 5 minutes ………………………………………………… 73 F.6 Mote 2 readings for 5 minutes …………………………………… ………… 74 F.7 Mote 4 readings for 5 minutes ………………………………… ………………75 F.8 Mote 5 readings for 5 minutes …………………………………… ………… 76 F.9 Motes reading for an hour ……………………………………… …………… 77 ix LIST OF FIGURES Figure Page 1.1. A conceptual biosensor shirt with sensors exposed on the outside for illustration …………………………………………………………….…………….4 3.1. Mica2Dot Motes …………………………………………………………….…….15 3.2. Mica2 mote ……………………………………………………………….……….15 3.3. MIB510 Hardware ………………………………………….…………………… 17 3.4. Block Diagram of the MIB510 …………………………………….…………… 17 3.5 Oscilloscope GUI to represent the graphical view of motes communication …………………………………………………………………….19 3.6 SerialForwarder reading the packet information ………….………………………20 3.7 Mote data from listen.java ………………………………………….…………… 22 4.1 One minute trials for four motes at a distance of 15 feet…………………… … 30 4.2 One minute trials for four motes at a distance of 30 feet……………… ……… 31 4.3 One minute trials for four motes at a distance of 45 feet…………….…….………31 4.4 One minute trials for four motes at a distance of 60 feet………………… …… 32 4.5 One minute trials for four motes at a distance of 75 feet…………… …… … …32 4.6 One minute trials for four motes at a distance of 90 feet………………………… 33 x 4.7 Five minute trial for four motes at a distance of 15 feet………….…………… …33 4.8 Five minute trial for four motes at a distance of 30 feet……………….……… …34 4.9 Five minute trial for four motes at a distance of 45 feet……………………… …34 4.10 Five minute trial for four motes at a distance of 60 feet………….…………… …35 4.11 Five minute trial for four motes at a distance of 75 feet……………………… …35 4.12 Five minute trial for four motes at a distance of 90 feet…………….………… …36 A.1 MIB510 programming board …………………………………………………… 44 A.2 One end of the cable connected with the mib510 board ……………………… …44 A.3 Another end of the cable connected with the computer …………… ……………45 A.4 Mica2 ………………………………………………………………… …………45 A.5 Mica2 in mib510 board ……………………………………………………… ….46 A.6 Mica2dot motes ………………………………………………………… ……… 47 A.7 Mica2dot connection with the Mib510 board ……………………….…………….47 A.8 Red light on mica2dot is blinking ………………………………………….…… 48 A.9 Mib510 connected with the computer …………………………………….……….50 A.10 Results after running listen tool ……………………………………………….… 50 [...]... wired biosensors connected to a wireless sensor mote acting as a controller for processing and communication The biosensor suite could include non-contact ECG sensors with blood pressure and body temperature sensors for diagnostic support A small handheld device with a much larger memory, processing capacity and short range wireless communication capability is the focal point for the controlling mote... of a set of equations for the definition of robustness in a system of wireless sensors intended to serve a homebound patient Robustness can be defined as a factor that determines if a system of interest is stable in the presence of unpredictable changes in the input In tandem, software was developed to test a set of wireless sensors, then data was collected and analyzed to validate the proposed robust... analyzed and discussed Finally, conclusions are drawn in chapter five and a plan for future work is presented 5 CHAPTER II ROBUSTNESS In general terms, robustness is the extent that a system can continue to function in the presence of small perturbations, i.e., when faults are introduced Ali et al., [6] have developed an interesting general mathematical formulation to create a metric for robustness in a... non-invasive technique, although skin spike and ingestible sensors exist Skin contact electrodes are applied directly to the skin with a conductive adhesive gel to maintain contact for high quality traces These irritate the skin, and are not meant for long term use or extensive movement Commercially available devices for home monitoring typically use 3-lead, hard-wired, skin contact ECG sensors that can be worn... Software Description and Terminology We also developed a set of programs, written in C, for packet communication between the Mica2 and Mica2Dot motes The code is based on software supplied by the 17 open source community supporting TinyOS and the Crossbow motes It is included in Appendices B, C, D, and E 3.2.1 TinyOS The system was developed using TinyOS, a small, open-source operating system developed... source mote to a destination mote in a tiny wireless sensor network We are interested in the pattern and reasons behind the loss The need to understand the pattern of packet loss in sensor networks is so that one can ultimately design an efficient network for the home healthcare environment of interest For these experiments, we use a Crossbow MIB 510 base station, as described in Chapter Three, connected... takes sensor readings and sends a group of readings over the UART The program was modified to send information in an RF signal Our modified version of the OscilloscopeRF sends the static data with the last byte increasing every time Appendix E contains the code No live sensor data was collected Instead, we used this program to create data in flash memory and transmit This research concentrates on the communication. .. Berkeley, to support sensor networks [13] It is an event driven operating system that handles the power consumption and radio networking [14] It is built to let the user focus on writing applications to acquire and react to sensor data It is built around the “hurry up and sleep” methodology because of battery life and energy awareness The majority of time the operating system maintains the mote in low power... (2 bytes) o ADC data readings (10 readings of 2 bytes each) The listen.java program has been modified to display the mote number and last byte of the packet message The Timer class is implemented in the code for running the data collection for 1 minute, 5 minute and 60 minute test intervals The code is attached in Appendix C A screenshot of the output from a mote test is shown in Figure 3.7 3.2.7 TOSBase... What we define as a source mote is capable of collecting sensor data from its sensor board connecter interface What we define as a destination mote collects the data from the source mote (or motes) and sends the data to a base station for storage Our base station is a laptop for this research 3.1.2 Mica2Dot Mica2Dot is a third generation, Quarter-Sized (25mm), Wireless Platform for Smart Sensor, which . DEFINING AND MEASURING ROBUSTNESS IN WIRELESS SENSOR COMMUNICATION FOR TELEMEDICINE A Thesis Presented to The Graduate Faculty of The University of Akron In Partial. Requirements for the Degree Master of Science Sudha Bhattarai August, 2008 ii DEFINING AND MEASURING ROBUSTNESS IN WIRELESS SENSOR COMMUNICATION FOR TELEMEDICINE Sudha. a wireless sensor mote acting as a controller for processing and communication. The biosensor suite could include non-contact ECG sensors with blood pressure and body temperature sensors for