Proc. SPIE. 5284, Wireless Communications and Networks
KEYWORDS: Internet, Surface plasmons, Detection and tracking algorithms, Multiplexing, Data transmission, Personal digital assistants, Telecommunications, Global system for mobile communications, Communication engineering, General packet radio service
Transmission scheduling algorithm for WWW traffic in GPRS needs to be optimized according to characteristics of Temporary Block Flow, TCP/IP and mobile access behavior. We analyze the radio resource allocation problem in GPRS and design a new scheduling algorithm FIFO-with-flow. It is simple and can adapt to temporal flows, and allocate bandwidth among temporal flows fairly. Based on the characteristics of mobile user access behavior, we determine the wireless WWW traffic model and its parameters reflecting TCP congest control. We simulate and compare the performance of FIFO-with-flow and FIFO. The FIFO-with-flow can support more WWW users than FIFO.
CC/PP (Composite Capability/Preference Profile) is a mechanism for content negotiation and adaptation between users and servers. The heterogeneity in communication devices and difference between application situations make it difficult to guarantee the compatibility of CC/PP documents. This increases CC/PP parsing time, and decreases service quality. The compatibility and efficiency are two contrary requirements. In this paper, we compare different approaches of CC/PP resolution. We introduce our CC/PP Parser design that is based on the UAProf CC/PP Schema and employs Java Factory Mode and XML SAX parser. We try to solve both the compatibility and efficiency problems. The parsing speed is 3-5 times that with RDF Parser. The coordination between server-adaptation and proxy-adaptation via CC/PP is simple and effective. The proxy and server can focus on adaptations optimized for free and paid services respectively, and system performance can be improved. Our system is configurable with an XML setup file. The existing approaches for CC/PP security and privacy protection decrease service quality. We suggest performing P3P comparison in the proxy. Our P3P approach increases service performance and can be integrated into the coordination model of server-adaptation and proxy-adaptation.