Compact Data Format for Advertising and Discovery in Ubiquitous Networks

In this paper, we describe a packet data size minimization method designed specifically for advertising and discovery in ubiquitous networks. The minimization is effective for achieving superior discovery performance characteristics such as discovery time and power consumption. The proposed method for data packet size minimization is based on indexing of advertisement text. In the method, dictionaries and indexed data are stored separately, i.e. dictionaries are stored on a server and indexed data is stored on ubiquitous wireless devices, and the same dictionaries are shared among all users. We evaluate an average packet data size and dictionary size for three indexing methods: regular indexing, category indexing and attribute indexing; and show that these methods achieve data packet sizes which are about two and three times smaller than raw data packets and zipped packet data sizes respectively. Also, we show that category indexing allows users to be less dependent on the infrastructu...

• Type: Conference
• Year: 2007
• Where: UCS
• Authors: Pavel Poupyrev, Yoshihiro Kawahara, Peter Davis, Hiroyuki Morikawa
• Keywords: Applied Computing, Data Packet Sizes, Packet Data Sizes, Size Minimization, UCS 2007