22-01-2013, 12:51 PM
Trusted Management for Mobile Computing Entities Enabled by Fuzzy Logic Approach
Trusted Management.pdf (Size: 126.37 KB / Downloads: 31)
Abstract:
In the mobile computing environments, such as the
mobile agent system, the management for the distributed mobile
entities is complicated. Some characteristic information in this
environment is uncertain and could not be presented by accurate
value. The fuzzy logic provides the technology to deal with this
condition. This paper presents a method for trusted management
of entities in mobile computing environment. A model system
based on fuzzy logic rules is designed. The mechanism of the
system is described and the experiment results are analyzed.
INTRODUCTION
Recently, the mobile computing systems, such mobile
agent and ad hoc system, are researched wildly. The
large-scale entities collaboration is an important aspect of the
mobile computing system and other mobile computing modes.
How to organize, manage and assign these distributed entities
are important and complicate. And another problem, security,
is becoming focus for us in recently years according to the
quickly development of mobile computing systems. Then,
how to manage the mobile entities efficiently and archive the
security requirements is important for system designers.
In a large mobile computing system, the missions should
be assigned to the distributed entities to archive the using
efficiency as high as possible, and at the meantime, the user
should choose the entities that they trust. As we know the one
that we trust may not be the one that should fulfill the high
efficiency requirement, on the other hand, the ones that could
be used to archive the high efficiency may no be the one that
qualify the trust requirement of the users. So, there should be
a balance between the efficiency and the trust.
The mechanism of model system
In our model system, there are distributed entity sites that
work together. Each mission submitted to the system should
be divided and allotted to these sites. We give each site a
4-tuple to describe their characteristics. The 4-tuple is S = (Q,
R, F, C), presenting the entity quantity, reliability, job
finish-rate, the cost for using the entity. The reliability is
about each site’s stability and failure rate of the hardware and
software. The quantity presents each site’s entity that could be
used, such as the computing ability, storage volume. Job
finish-rate presents the proportion of the job finished
successfully to all the jobs that assigned to one site. The user
should pay cost for using the entities in one site.
THE EXPERIMENT AND RESULTS
To verify the effectiveness of this fuzzy logic rule-based
entity manage model system, we simulated the entities on five
sites, with 50, 100 and 300 mobile computing entities. The
quantity of the entity is set as storage and computation ability.
The initial trusted degree is set randomly from 0.4-0.8. Each
mobile computing entity just needs the computation or the
storage entity, not both. Each site has the computation and
storage entity with the cost respectively. We set some faults
on the site randomly. If a mobile entity moves to these sites on
the fault time, the entity will fail. We compare the assign in
two conditions: one has the fuzzy rules, another has not. The
simulation execution results are presented in table 3.
CONCLUSIONS
This paper presents a method for the entity management
based on the fuzzy logic in the mobile computing
environment. Efficient fuzzy reasoning is a critical issue for
the entity management. In the proposed model system, we
used the fuzzy rules to derive the fuzzy information the
trusted degree, and manage to archive the high efficient for
using the mobile entities. The experiment result presents the
availability of this method. The proposed approach could be
used in different mobile computing environments.