An efficient divide-and-conquer approach for big data analytics in machine-to-machine communication

摘要: Machine-to-Machine (M2M) communication relies on the physical objects (e.g., satellites, sensors, and so forth) interconnected with each other, creating mesh of machines producing massive volume data about large geographical area living non-living environment). Thus, M2M is an ideal example Big Data. On contrary, platforms that handle Data might perform poorly or not according to goals their operator (in term cost, database utilization, quality, processing computational efficiency, analysis feature extraction applications). Therefore, address aforementioned needs, we propose a new effective, memory efficient system architecture for in M2M, which, unlike other previous proposals, does require whole set be processed (including raw sets), kept main memory. Our designed exploits divide-and-conquer approach block-wise vertical representation follows particular petitionary strategy, which formalizes problem applications. The goes from servers, where first transformed into several blocks can quickly processed, then it classifies reorganizes these same source. In addition, are aggregated sequential manner based machine ID, equally partitions using fusion algorithm. Finally, results stored server helps users making decision. feasibility efficiency proposed implemented Hadoop single node setup UBUNTU 14.04 LTS core?i5 3.2GHz processor 4GB show efficiently extract various features (such as River) satellite data.