Downloads: 125 | Views: 164
M.Tech / M.E / PhD Thesis | Computer Science & Engineering | India | Volume 5 Issue 6, June 2016
Knowledge Extraction from Wireless Sensor Networks
Abstract: As of late, information administration and handling for remote sensor systems are turned into a theme of dynamic exploration in a few fields of software engineering, for example, circulated framework, database frameworks and information mining. WSN creates substantial measure of information streams. Removing data from the information is the most critical undertaking in WSN. Numerous techniques are accessible to extricate the data from WSN. These strategies produce the affiliation rules utilizing recurrence designs. It makes vast number of tenets a large portion of which are non-educational and neglects to reflect genuine connection among the sensor information. This paper shows new sort of behavioral example called as a related sensor design. Related sensor design catches the co-events and worldly connection of the sensor information. To catch such examples a conservative tree structure, called related sensor design tree (ASP-tree) and mining calculation are proposed. At the point when information streams moves through the system there is plausibility of loss of essentialness for the present time of the old data. ASP tree is further improved to SWASP-tree to catch centrality of late information by receiving sliding perception window and overhauling tree structure in like manner.
Keywords: Wireless sensor networks, data mining, knowledge discovery, behavioral patterns, sensor data stream
Edition: Volume 5 Issue 6, June 2016,
Pages: 1907 - 1910