Dealing Data From Sensors Network:
Continuous communication and collection of data string from data bases. They are for monitoring purpose such as climate, earthquakes, building structure. They are used to gather data using primitive networks. Cleaning of data techniques like noisy data, and handling of incomplete data due to missing values.
It has limited battery life and processing capabilities for special data handling and architectures for sensors data streams that may incorporate resource constraints. They are capable of contracting and maintaining the locally view its neighborhood.
Sensor stream only works on sample of population means only selected persons while traditional streaming works on entire population. Sensors are imprecise and very noisy while traditional streaming are exact and error free. Sensors have moderate size while they have huge size. Sensors used are very expensive whereas they are cost free. Sensors are for continuous domain while they use for frequent items and they are distinct.
Routing: it is an ability to respond the query of user to every node. It provides one or more routes to the every node and it uses tree based routing where one node is the root. TAG reduces the communication and increases the magnitude or reduction which in turn reduces the cost very much.
There are three approaches used to storing the data in sensor networks, external, local, and data centric storage. DCS stores the data by name and thus cost of communication from original data to new data is stored. This in turn reduces the cost of query in this method is reduced very much as the queries are stored. System has three types of queries streams like historical, snapshot and long running queries.
These sensors are used to build the decision making models. They are directly used in real time systems.
Download Paper Presentation for it on Dealing Data From Sensors Network .