Distributed Data Streams

Abstract We consider a scenario where server is wirelessly connected to countless sensor nodes that continuously measure data. The task of the monitor sensors’ More precisely, at each time step calculates function defined over current measurements sensors. Since sensors only have small computational power and tight battery constraints, communication between should be as possible, i.e., we aim minimizing total number messages transferred. There are various conceivable problems for setting above. demonstrate our approaches on following three: In Top- k -Value Monitoring Problem, aims identifying largest values. -Position Problem shifts identify observing these Finally, Count Distinct obliges determine distinct values currently observed. For all three problems, not present communication-efficient protocols one step, also show how it can exploited if input similar consecutive steps reduce long term. Thereby, utilize different techniques – involving sampling, dynamic data structures, filter-based approaches, combinations them their potential limits in broad described