Date: Monday, March 5, 2001
Time: 1:15 P.M. (Talk) <========= NOTE!!
Place: Withers 402-A, NCSU Historical Campus (click for
courtesy parking request)
Speaker: David Thuente, Computer Science, Indiana University - Purdue University Fort Wayne
Modeling and Simulation of Data Link Layer Protocols for Radios
Abstract: Modeling and simulation are critical tools for
the design of communication systems and protocols. MIL-STD-188-220B
is an evolving family of data link layer protocols designed for narrow
band radio communication. Type 2 services are connection oriented with
decoupled acknowledgments based on HDLC protocols. Type 4 services
are connectionless with individual decoupled acknowledgments for every
packet received. The two most visible and perhaps viable MAC algorithms
from MIL-STD-188-220B are deterministic adaptable priority network access
delay (DAP-NAD) and radio embedded network access delay (RE-NAD) and its
extensions. Comparative studies for Type 2 and Type 4 on data only
networks were done using DAP-NAD and RE-NAD with the results indicating
when Type 2 is most advantageous. DAP-NAD was extended to include
voice and the simulations were run comparing it with RE-NAD optimized for
voice. A number of extensions and modifications to DAP-NAD, including
efficient integration of data and voice, are presented. The implicit token
passing scheme is modified so that the token effectively jumps over stations
that may have transmitted in the last period and allows heavily used nodes
or "higher priority nodes" to have increased transmission opportunities.
These modifications are shown to have markedly improved performance for
the standard Tactical Internet Division and Below (TIDB) network used to
compare other MAC algorithms as well as on 16 node randomly loaded networks.
The modifications to DAP-NAD presented have further enhanced its ability
to efficiently handle both data and voice which is recognized by calling
the MAC algorithm: Data and Voice Network Access Delay (DAV-NAD).
Finally, some open research problems are discussed.
Host: A.
Tharp, Computer Science, NCSU