EE392S Sensor Networks A Series of Ten Lectures to be given Winter quarter of the 03 04 academic year John S Wenstrand 1 Abstract Silicon sensor technology has been under active development for more than thirty years with well established technologies for sensing force temperature strain pressure flow presence of chemical and biological agents voltage current charge photons images relative motion acceleration distance and position to name a few At the same time digital silicon technology has advanced to the point where significant computational power can be integrated with analog structures permitting local interpretation of data in the very low cost low power integrated systems sometimes referred to as smart sensors A third technological vector that of low power ad hoc radio networks is currently a topic of active research This emerging technology offers the promise of low cost low power wireless interconnection of smart sensors This seminar series will explore the interplay of these technologies as enablers of sensing systems formed from networks of cooperating smart sensors A survey of sensing technology ad hoc radio networks and integration issues will suggest the range of capabilities and design trade offs for smart sensing network nodes Information theoretical and networking considerations will provide a sense of the system trade offs of power volume of data data rate and sensor count Human factors and practical considerations for widespread consumer adoption of smart sensing networks will be presented Finally future applications and barriers to adoption of sensing systems in consumer industrial medical military and medical fields will be presented Professors Dasher Dutton and El Gamal and others have offered many helpful ideas on structure content and organization of this series 1 2 Outline 2 Speakers from industry government Stanford and other academic institutions will provide diverse views of the state of the art and build bridges to the Stanford community in this emerging field It is intended that a participant in this series will be left with a broad understanding the confluence of technologies which are enabling the creation of networks of smart sensors the range of capabilities and design trade offs which govern such systems and the requirements for making these systems truly pervasive A seminar EE402S offered Spring quarter by Professor Dasher will focus on advanced sensing applications presenting a global view of emerging applications of such systems Close coordination with Professor Dasher already in progress will result in this seminar and 402S being quite complementary It is expected that the proposed seminar will initiate an enduring multidisciplinary conversation on the topic of sensor networks 2 Outline A conceptual outline for the series follows This initial outline is expected to change significantly over the next quarter with input from prospective speakers so as to better achieve the intent described above 1 Overview of Sensing Systems January 6 2004 Summarize the rich history of silicon sensors Analyze trends in silicon manufacturing technology processor technology and mixed signal design enabling low power inexpensive smart sensors Highlight the emergence of ad hoc radio network technology Discuss the size and growth of the silicon sensors in the global electronics market Present the purpose of series the conceptual framework for the course and requirements for credit 2 Silicon Sensor Technology January 13 2004 Survey existing and emerging silicon sensor technologies Highlight the large number of mature technologies Discuss the cost silicon area special fabrication and environmental requirements relevant to widespread deployment Exclude discussion of image sensors 3 Sensing and Computation Image Sensors January 20 2004 The historical vision for measurement computing and communications M C 2 and pervasive computing still rings true today Motivate the integration of smart sensors Analyze trade offs in generation of silicon technology power silicon area and yield Present examples 2 Outline 3 of smart sensors with significant integrated computation using image sensors as an example 4 Low Power Ad Hoc Radio Networks January 27 2004 Present current research and emerging standards for low power radio networks considering energy vs hop length energy per hop energy for forwarding bandwidth limitations and robustness What can be learned of topology from the network Estimate the time line for commercial availability of different types of networks 5 Information theoretical issues for collaborative networks of sensors February 3 2004 This lecture will examine constraints on system architecture for high bandwidth sensing applications such as imaging Link and network bandwidth limitations dictate the degree of local data reduction required and may bound non local collaboration 6 Energy Scavenging February 10 2004 Enumerate energy sources in the environment Quantify usable energy to be extracted from various environments Compare long lived battery operated systems to energy scavenging systems Discuss opportunities for creating energyrich environments Present examples of existing systems Summarize implications for design of future energy scavenging systems 7 Human Factors and Product Design February 17 2004 Wireless low power low cost sensing systems are requirements for rapid ultrahigh volume deployment of sensing systems Widespread acceptance of such systems requires ease of use and accommodation of human preferences This lecture will look at principles of design for adoption of new technologies in general and for sensing systems in particular Examples of good and bad design will be presented 8 Integration bounds and examples February 24 2004 Design tradeoffs cost constraints and market acceptance for a smart sensing system will be presented Alternatively a more specialized domain may be presented such as Toward an Integrated System for Interpretation of Neurological Activity Military Applications of Distributed Sensing Systems What can be done with today s sensing system technology if cost constraints are removed 9 Research Directions for Sensor Networks March 2 2004 A government sponsor of sensor network research will survey current topics of 3 Logistics 4 research in sensor networks and interesting problems in need of attention What are the key technical problems in need of solution in support of military homeland security industrial and commercial deployment of