Summary: Local Optimization of Global Objectives:
Competitive Distributed Deadlock Resolution and
Resource Allocation
Baruch Awerbuch \Lambda Yossi Azar y
Abstract
The work is motivated by deadlock resolution and
resource allocation problems, occurring in distributed
server­client architectures. We consider a very general
setting which includes, as special cases, distributed band­
width management in communication networks, as well as
variations of classical problems in distributed computing
and communication networking such as deadlock resolu­
tion and ``dining philosophers''.
In the current paper, we exhibit first local solutions
with globally­optimum performance guarantees. An ap­
plication of our method is distributed bandwidth man­
agement in communication networks. In this setting,
deadlock resolution (and maximum fractional indepen­
dent set) corresponds to admission control maximizing
network throughput. Job scheduling (and minimum frac­

Source: Azar, Yossi - School of Computer Science, Tel Aviv University

Collections: Computer Technologies and Information Sciences