Dynamic Spectrum Access Networks: From Dream to Reality
Event details
| Date | 26.06.2009 |
| Hour | 15:15 |
| Speaker | Prof. Allen B. MacKenzie, Virginia Tech |
| Location | |
| Category | Conferences - Seminars |
For the last several years, the idea of using dynamic spectrum access
(DSA) strategies to overcome the artificial scarcity of radio
frequency spectrum has been popular among both regulators and
researchers in wireless communications and networking. While many of
the problems associated with creating radios capable of DSA have been
solved, fundamental challenges remain in creating and managing
networks of such radio nodes. This talk will describe some of the work
that we have done to address those challenges and will present ideas
for continued progress through the pursuit of improved experimental
capabilities for wireless networking researchers.
First, we will briefly describe the motivation for DSA in general and
some of the specific applications currently envisioned for DSA
networks. Then, we will discuss the problems of topology control and
medium access control in DSA mobile ad hoc networks. Specifically, we
advocate a broader view of topology control than has been common in
the past, by including the choice of link parameters such as channel,
rate, and power in the pursuit of improving end-to-end network
performance. We further advocate the need for dynamic topology
control approaches that optimize the network topology dynamically to
accommodate current traffic. We discuss some of our initial work in
this area, which includes approaches based on game theoretic models,
genetic algorithms, and heuristics. We also describe the shortcomings
of work in this area --- most notably the failure to consider
operational issues associated with real networks. This leads to our
discussion of medium access control, in which we describe our work on
a multi-channel medium access control protocol for DSA. Again, we
point out a flaw that our work shares with much other work in this
area --- the assumption of a common control channel --- and describe
the nascent development of techniques to overcome this flaw.
With these problems as background, we discuss the need for frameworks
and platforms that make experimental work with DSA networks more
accessible to networking researchers. We describe a simple proof-of-
concept DSA network that we constructed and discuss the challenges
associated with currently available platforms for such experiments.
Finally, we discuss the requirements and potential architecture of an
accessible experimental platform for DSA networking research.
Practical information
- General public
- Free