Information Technology Reference
In-Depth Information
100.00%
98.00%
96.00%
94.00%
Our model
92.00%
90.00%
Complain-only model
88.00%
86.00%
84.00%
82.00%
80.00%
1
51
101
151
201
251
301
351
401
451
Figure 4.6
Comparison of performance of the RCM and complain-only model.
4.4.5 evaluation
The test results have clearly revealed that our dynamic, reputation-based
trust model RCM, enhanced the system security level after a certain num-
ber of transactions were performed in the P2P community. An analysis of
each trend line, which describes the trends in the system security level,
revealed that the security of the P2P community improved with time. We
were also able to observe that, on account of its rapid self-adjusting property
as inferred from the trend line, our system performed well even when there
were many malicious nodes existing in the system. After having compared
the performance of our model with that of some existing models, we can
assert that our model affords significant advantages over the existing models
and that it is very effective in automatically enhancing system security by
using the concept of self-adjusting chain credibilities.
4.5 Conclusion
It could be concluded, from the above discussion, that our model success-
fully addresses the P2P reputation problem. Our model can locate a mali-
cious resource chain after each transaction and weaken it to prevent it from
