Candidate Projects for CSCI 599

Delivery of Continuous Media in Mobile Ad Hoc Networks of Gaming Devices

Prof. Shahram Ghandeharizadeh

· Consider current studies that focus on environments resembling H2O clouds. Such studies can be extended in several directions. Each extension would qualify as a project:

1. Instead of system throughput, consider availability of data. At any given instance in time, what fraction of data is available to a node with a data-placement strategy such as Frequency-based, Byte-hit, Halo-clip, and Halo-block. Given a fixed probability of a link becoming unavailable in the network, what is the availability of data? And, how is it impacted by different placement strategies.
2. The placement strategies discussed in class are static. Overtime, the frequency of access to different clips might evolve. A project may design, implement, and evaluate replacement policies that modify the placement of data based on the sequence of requests issued either by a device or its neighbors. Such a policy must be able to choose victim objects when the storage capacity of a node is exhausted. Moreover, the autonomy of a device must be preserved. It should be able to make decisions independent of its neighbors. One may consider a variety of local and global optimization objectives.
3. What is the impact of multiple networking cards offering different bandwidths and radio ranges? What are the benefits of configuring a device with multiple cards? What are its limitations?