Adaptive Network Control for Wireless Multimedia Communication
Timothy X Brown

Currently, wireless communication systems are focussed on a single
traffic type such as voice circuits in cellular and short messages in
pagers. The state of the so-called PCS revolution is really an
incremental improvement on existing cellular telephone service. Before
the promise of PCS-wireless transport of multimedia such as voice,
data, and video traffic anywhere, anytime-is possible, significant
technologies must be developed. This proposal will research one such
technology.

My work is focussed on building intelligent, adaptive methodologies for
controlling and managing multimedia traffic. This proposal will take my
work in a new direction to address the burgeoning wireless
environment.

I propose developing an adaptive multiplexing approach to guarantee
quality of service (QoS) to multimedia traffic in a wireless
communication network. My objective is to enable new mobile multimedia
services that can support QoS guarantees while maximizing the
utilization of the valuable wireless bandwidth through research in the
following areas:

  o Statistical classification of what combinations of traffic sources
    can and can not meet QoS guarantees despite traffic, spatial, and
    channel variability.

  o Reinforcement learning methods to learn which calls should be
    accepted to maximize network utilization.