RTPD/MACSC: a novel approach for effective pervasive information retrieval

Abstract

We propose the novel RTPD/MACSC dynamic cache size tuning approach to shorten the information retrieval roundtrip time by a SFF (small-form-factor) client in a pervasive computing environment. By itself the MACSC (model for adaptive cache size control) does not compensate for the ill effects due to the Internet traffic. In the proposed approach, however, the MACSC is enhanced with the real-time traffic pattern detection (RTPD) capability. With this support the MACSC tuner strives to maintain the given hit ratio as the minimum and reduces the amount of cyber foraging. In the process the MACSC component self-tunes its core point-estimate (PE) calculation to neutralize any ill effects of the traffic pattern currently detected and identified by the supporting RTPD capability.