CMSC 818Z

Spring 2007

Information-Centric Design of Systems

 

            In this course we will discuss several technologies that have been developed within the last several years, and present these developments from the information-centric point of view. Open problems for each technology will also be presented. Students are expected to take any of the open problems discussed in class and study them further as a project/ term paper. The course will have a final exam. It will count for MS/Ph.D. course work. It will also count for MS Comp.

 

The following technologies will be discussed.

  1. Information Dynamics: This is a framework for studying the basic nature of information and its implications. This framework makes a distinction between the information and its representation, implicit and explicit information, value of information and its changes with time. In design of systems this framework addresses the issues of information needs.
  2. Time Synchronization: Time plays a critical role in most systems and accurate knowledge of time can be used in a variety of ways to improve systems. In a distributed environment developing a common knowledge of time often poses challenges. Techniques for clock synchronization which can be scaled to the Internet scale will be discussed.
  3. Location Technologies: With the location-aware computing becoming popular it is essential that the location be determined with required degree of accuracy indoors and outdoors. There are several technologies that have been developed for location determination. We will discuss two of them:
    1. RSSI Based: In this technology we derive the location by measuring the received signal strength for a wireless signal from, for example, a few 802.11b access points.
    2. PinPoint Technology: In this technology the propagation time is used as a measure for distance. Pinpoint requires stable, not synchronized, clocks and can determine the location with high degree of accuracy.
  4. Location-Aware Computing: We have been developing Rover technology as a base to build examples of location-aware applications. Several new approaches have been introduced in the design of Rover. For example, it uses an action/operation paradigm for managing the processing which is significantly more efficient than thread-based system.
  5. Cyclone Technology: Using ideas from real time computing this technology provides a time-based management of switches and assures a loss free, jitter free delivery of data packets end to end. A novel idea introduced in this technology is that of synchronization, in which each node has to have a stable clock and by not sending frames back to back the synchronization of the whole network is maintained.
  6. Performance of Networks: Our approach of studying the performance of networks is based on using fine grained measurement of end-to-end delays on a network and then explaining the behavior of the network using deterministic modeling techniques. The available bandwidth estimates are also created using this approach. We will also study the performance of wireless networks, reflecting some measurements which are being conducted at present.
  7. Ad-Hoc Networking: For applications such as sensor networks we have developed SONIC technology which permits a set of nodes to self-organize into a two level hierarchy.

Additional topics will be covered based on the availability of time.