James Smith
Ph.D. FinalJun 27, 2008, 10:30AM
ECE conference room
Robust Resource Allocation in Heterogeneous Parallel and Distributed Computing Systems
Abstract: In a heterogeneous distributed computing environment, it is often advantageous
to allocate system resources in a manner that optimizes a given system
performance measure. However, this optimization is often dependent on system
parameters whose values are subject to uncertainty. Thus, an important research
problem arises when system resources must be allocated given uncertainty in
system parameters. Robustness can be defined as the degree to which a system
can function correctly in the presence of parameter values different from
those assumed. In this research, we define mathematical models of robustness in
both static and dynamic stochastic environments. In addition, we model
dynamic environments where estimates of system parameter values are provided as
point estimates where these estimates are known to deviate substantially from
their actual values.
The main contributions of this research are (1) mathematical models of
robustness suitable for dynamic environments based on single estimates of
system parameters (2) a mathematical model of robustness applicable to
environments where the uncertainty in system parameters can be modeled
stochastically, (3) a demonstration of the use of this metric to design
resource allocation heuristics in a static environment, (4) a mathematical
model of robustness in a stochastic dynamic environment, (5) we demonstrate the
utility of this dynamic robustness metric through the design of resource
allocation heuristics, (6) the derivation of a robustness metric for
evaluating resource allocation decisions in an overlay network along with
a near optimal resource allocation technique suitable to this environment.
to allocate system resources in a manner that optimizes a given system
performance measure. However, this optimization is often dependent on system
parameters whose values are subject to uncertainty. Thus, an important research
problem arises when system resources must be allocated given uncertainty in
system parameters. Robustness can be defined as the degree to which a system
can function correctly in the presence of parameter values different from
those assumed. In this research, we define mathematical models of robustness in
both static and dynamic stochastic environments. In addition, we model
dynamic environments where estimates of system parameter values are provided as
point estimates where these estimates are known to deviate substantially from
their actual values.
The main contributions of this research are (1) mathematical models of
robustness suitable for dynamic environments based on single estimates of
system parameters (2) a mathematical model of robustness applicable to
environments where the uncertainty in system parameters can be modeled
stochastically, (3) a demonstration of the use of this metric to design
resource allocation heuristics in a static environment, (4) a mathematical
model of robustness in a stochastic dynamic environment, (5) we demonstrate the
utility of this dynamic robustness metric through the design of resource
allocation heuristics, (6) the derivation of a robustness metric for
evaluating resource allocation decisions in an overlay network along with
a near optimal resource allocation technique suitable to this environment.
Adviser: HJ Siegel
Co-Adviser: A. A. Maciejewski
Non-ECE Member: Pat Burns
Member 3: A. A. Maciejewski
Addional Members: N/A
Co-Adviser: A. A. Maciejewski
Non-ECE Member: Pat Burns
Member 3: A. A. Maciejewski
Addional Members: N/A
Publications:
Program of Study: