A Dynamic Adaptive Threshold Transmission for XML Data on Networks

Abstract — Since XML became an official recommendation of the World Wide Web Consortium (W3C) in 1998, it is increasingly being used to transmit data on networks but is a verbose format and needs an efficient encoding to send relatively large amounts of data efficiently. This requirement is particularly important for wireless data communications. It is a common technical challenge for researchers in XML-driven networks to have good performance. One may employ a middleware to enhance performance by minimizing the impact of transmission time [1, 3]. Normally, to reduce the amount of data sent the XML documents are converted to a binary format using a compression routine such as Gzip. However while this would reduce the payload, it results in an increase in the CPU time as the XML document must be compressed before being sent and uncompressed when it is received. In this paper we extended our previous research results [2, 11-13] to an enabling technology, namely Dynamic Adaptive Threshold Transmission (DATT) for XML data on networks. We also show the experimental results obtained from our technique and that from the Network Adaptable Middleware (NAM) established by Ghandeharizadeh et al [1]. Experimental results show that our method is superior to the NAM method [1], supported by the fact that the time taken is 220.6 times better. We have introduced another way of determining when to