Published online Nov 2. Author information Article notes Copyright and License information Disclaimer. Received Sep 23; Accepted Oct Abstract Efficient matching of incoming events of data streams to persistent queries is fundamental to event stream processing systems in wireless sensor networks.
Keywords: complex event processing, data streams, adaptive strategy, parallel processing, queue theory, probability theory. Introduction Recently, there has been an increasing interest in wireless sensor networks, which require continuously processing flowing data from geographically-distributed sources to achieve timely responses to complex queries, such as data stream processing DSP systems [ 1 , 2 , 3 ] and complex event processing CEP systems [ 4 , 5 , 6 ]. Related Work In CEP systems, the operators are demanded to be highly scalable under high event stream rates. Preliminaries 3. Event Model An event that represents an instance and is atomic is an occurrence of interest at a point in time.
System Model 4. Parallelization Model In this section, we propose a parallelization model that can be utilized for pattern operators, which is shown in Figure 1. Open in a separate window. Figure 1. Event Splitting Policies In this section, the event splitting policies are given, which can be utilized for processing pattern operators in parallel. Adaptive Parallel Processing Strategy In this section, an adaptive parallel processing strategy APPS is proposed to estimate and select the optimal event splitting policy, which can suit the most recent workload conditions such that the selected policy has the least expected waiting time for processing the coming events.
Figure 2. Table 1 Notation. Degrees of Parallelization The aim of this stage is to decide the degrees of parallelization for pattern operators in the CEP system to be used for processing data streams. Expected Size of the Batch Partition For further parallel processing, the input stream I E j needs to be divided into batch partitions.
Event Processing Time Collection The aim of this stage is to collect the processing time of the events from the last event type matched by the pattern operator, which are used in the on-line estimation step to estimate various distributional properties of the processing time distribution. Trade-Off between the Estimation Accuracy and the Processing Time Figure 3 depicts an example of obtaining an appropriate policy for processing the further coming events. Figure 3. Example of obtaining an appropriate policy for processing the coming events.
On-Line Selection of Event Splitting Policies This stage is pretty critical in the proposed adaptive parallel processing strategy, which can estimate and decide the appropriate policy on-line. Experimental Evaluation Based on the parallelization model in Figure 1 , we implemented the experiments on the StreamBase [ 12 ] system for query q 1. Figure 4.
Figure 5. Comparing the methods under the variation of time window sizes. Figure 6. Conclusions In this paper, we started off with identifying the general problems of adaptive parallel processing with respect to pattern operators in CEP systems. Author Contributions F. Conflicts of Interest The authors declare no conflict of interest.
References 1. Lee I. A scalable and adaptive video streaming framework over multiple paths. Tools Appl.
Ding J. Perceptual quality based error control for scalable on-demand streaming in next-generation wireless networks. Jang J.
An effective handling of secure data stream in IoT. Soft Comput. Chen M. Recommendation-aware smartphone sensing system. Boubeta-Puig J. A model-driven approach for facilitating user-friendly design of complex event patterns. Expert Syst. Complex event processing modeling by prioritized colored Petri nets. IEEE Access. Cugola G. Processing flows of information: From data stream to complex event processing. ACM Comput. CSUR ; 44 Oracle CEP. CEP for Hospital. Kim K. Xiao F. Efficient processing of multiple nested event pattern queries over multi-dimensional event streams based on a triaxial hierarchical model.
- Understanding atomic and composite patterns for big data solutions – IBM Developer.
- Big Data Architecture in Data Processing and Data Access?
- Oppression and Liberty (Routledge Classics);
- Product Description & Reviews.
Parallel processing of continuous queries over data streams. Parallel Databases.
Han W. Parallelizing query optimization. VLDB Endow. Hirzel M. Johnson T. Liu B.
- Diversification of Agriculture in Eastern India.
- 1. Introduction!
- Innovations in Power Systems Reliability.
- Wind Dancer;
- Semiconductor Research Corporation - SRC!
- That Dream Shall Have a Name: Native Americans Rewriting America?
- Understanding atomic and composite patterns for big data solutions!
Chaiken R. Upadhyaya P. Dispatching stream operators in parallel execution of continuous queries. Brenna L. Akdere M. Plan-based complex event detection across distributed sources. Nested pattern queries processing optimization over multi-dimensional event streams; Proceedings of the 37th Annual Computer Software and Applications Conference; Kyoto, Japan. Carney D. Economical and fault-tolerant load balancing in distributed stream processing systems. Suhothayan S.
Parallelizing stateful operators in a distributed stream processing system: How, should you and how much? Balkesen C. Brito A. Schneider S. De Matteis T.
Memory access pattern - Wikipedia
May not contain Access Codes or Supplements. May be ex-library. Buy with confidence, excellent customer service!. Mary E. Publisher: Kluwer Academic Publishers , This specific ISBN edition is currently not available.