CS-541

Wireless Sensor Networks

Postgraduate

Wireless Sensor Networks are an emerging type of networks and a multi-disciplinary field of research and development. Their distinct characteristic from data networks is that their reasoning of existence is to monitor the physical space and provide access to remote or hostile environments, in a cost-effective manner, without the necessity of any previous infrastructure (e.g. LAN, optical fibres, etc). Representative applications fields are related, but not limited to: environmental and wild-life monitoring, industrial control, smart buildings, smart grid and water management, search and rescue applications, modern telemedicine, etc.

WSN are therefore a multidisciplinary field that combines signal processing, networking, embedded programming, and data management. CS-541 focuses on the theory, design and development aspects of such networks with the primary objective to provide a spherical introduction to this field, and additionally cover special topics from a signal processing and networking perspective. In a nutshell, CS-541 will cover the following topics: (a) network standards and connectivity / coverage aspects; (b) Distributed signal processing and machine learning; (c) Programming, simulation and data analysis.
 
Course Syllabus. Low-power Personal and Body Area Networks, IEFT RPL & uIP standard, Critical Transmission Power and Asymptotic Connectivity, Sensing Coverage in Convex / non-Convex environments, Deterministic and Probabilistic Sensor Deployment, Synchronization / FSP, Bio-inspired networking methods for dense sensor networks. Distributed algorithms for acquisition, storage and processing: Consensus and Gossip algorithms, Distributed Data Compression, Network Coding Schemes. Modelling and Learning of Spatio-temporal data: Compressed Sensing, Sparse Representations, Low Rank Matrix Completion. Localization: dead-reckoning, passive, multimodal. Programming principles with Real-time Operating Systems: tinyOS / nesC, protothreads / Contiki OS, Over-the-air-programming
 
Practical information. During the semester you will be assigned 4 homeworks that will combine the theoretical and programming aspects of the material taught. The programming environment will be on MATLAB and C / Java (Contiki / Cooja  Emulator). In addition to these homeworks, you will also be assigned on individual or team projects, which will address special topics on WSN (e.g. build a new routing algorithm, perform a set of experimental studies, design etc). Prospective students will also have to present their project to the rest of their classmates in a 15-min time slot.

Grading. Homeworks: 40% (Grading scale: 0-10), Projects: 40% (Grading scale: 0-10), Final exam: 20% (Grading scale: 0-10)

6

CS-335, CS-215

http://www.csd.uoc.gr/~hy541