Complex event processing in internet of things

With the Internet of Things (IoT), the amount of raw data generated by sensors has increased. In order to extract meaningful information from this high volume and speed data, the data must be processed in real time. In this context, Complex event processing (CEP) is promising technology that used for a real time data processing with extracting valuable information from raw data. However, there are open problems about CEP system such as defining automatically CEP rules and CEP prediction. In this thesis, DL based solutions offered to solve these problems with using IoT data. Firstly, a model which has clustering and rule mining phase is proposed to extract CEP rules from IoT data. Experimental results show that the developed model shows excellent performance with determining the scope of the cluster. Secondly, we propose a generalized framework for automatic CEP rule extraction with using DL methods. The proposed framework has two phases which names labeling and automatic rule extraction. In this context, we compare several DL methods with each other and regression-based methods to evaluate the proposed framework in smart city scenario with using air pollution gases by reconstruction error and prediction metrics. The result shows that the Dl based framework success rate increase when the number of attributes increases as well. Finally, we focus on the predictive CEP problem. In this context, we propose a novel deep learning based predictive CEP system for air pollution dataset in IoT framework. DL methods and SVR are compared in terms of the prediction performance. The predicted data is processed in the CEP engine and the performance of the each algorithm in CEP engine is shown in detail. We examine proposed system in terms of end-to-end network delay and measure its impact on the network. The result demonstrate that predicted system with Dl methods show excellent performance while guaranteeing minimum end-to-end network delay.