Network information retrieval method controlled by energy and division probability parameters
K. Yamaoka, T. Yonemaru, A. Kobayashi, Y. Sakai · 2005
Conventional retrieval methods tend to either satisfy network administrators by reducing network traffic as flooding method, or to satisfy network users by increasing the speed of a successful search as sequential retrieval method, and these are conflicting demands. However, these retrieval methods do not consider the existence probability of target objects on a network. In this paper, we propose a new retrieval method that is flexibly adapted to the existence probability of target objects and to specified requirements of discovery probability, search time, and traffic level. This proposed method is a single query action based on an autonomous distributed network that freely changes the weight between network traffic and discovery probability according to retrieval requirements. The simulation results reveal that network traffic can be decreased by setting appropriate values for the energy and division probability of query messages with little reduction in the discovery probability of the retrieval target object.