International Journal of Distributed Systems and Technologies, 2(3), 37-48, July-September 2011 37
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
Keywords: Causally Delivery, Group Communication, Heterogeneous, Hybrid Clock, Liner Clock, Physical
Clock, Unnecessarily Ordered Messages
Design and Implementation
of Hybrid Time (HT)
Group Communication
Protocol for Homogeneous
Broadcast Groups
Isamu Tsuneizumi, Seikei University, Japan
Ailixier Aikebaier, Seikei University, Japan
Makoto Ikeda, Seikei University, Japan
Tomoya Enokido, Risho University, Japan
Makoto Takizawa, Seikei University, Japan
ABSTRACT
To realize the cooperation of a group of multiple peer processes (peers), messages sent by peers must be
causally delivered to every peer. In a scalable group, it is necessary to reduce the communication overhead
to causally deliver messages. In this paper, the authors take advantage of the linear time (LT) and physical
time (PT) protocols, as the message length is O(n) for the number n of peers. However, some pairs are un-
necessarily ordered, that is, even if a pair of messages is ordered in the protocols, the messages may not be
causally ordered. The greater the number of messages that are unnecessarily ordered, the larger the overhead
is implied since the messages must be kept in a receipt queue if a message is lost or delayed. This paper dis-
cusses a hybrid time group communication (HT) protocol that reduces the number of messages unnecessarily
ordered. The HTprotocol is evaluated in terms of the number of unnecessarily ordered messages compared
with the PT and LT protocols. It is demonstrated that the number of unnecessarily ordered messages can be
reduced in the HT protocol compared with the LT and PT protocols.
INTRODUCTION
In peer-to-peer (P2P) information systems
(Schollmeier, 2001), a group of multiple peer
processes (peers) are cooperating to achieve
some objectives by exchanging messages with
each other. A P2P system is in nature fully
distributed with no centralized coordinator
and is scalable. Here, messages are required
to be causally delivered to peers (Nakamura
DOI: 10.4018/jdst.2011070103