Study on Distributed Complex Event Processing in
Internet of Things based on Query Plan*
Lingyun Yuan, Dongdong Xu, Guili Ge, Mingli Zhu
College of Information Science and Technology
Yunnan Normal University
Kunming, China
blues520@sina.com, xudongdong.ynnuedu@163.com
Abstract—Complex event processing is an efficient
method in data stream processing of Internet of things, but
more of these methods are referred to a single complex
event or a small quantity of events. Aiming at this problem,
a distributed complex event processing architecture for
Internet of things is presented in this paper, in which a
distributed query plan of complex event process structure
based on directed acyclic graph (DAG) is given, moreover,
a distributed query-plan complex-event-processing
algorithm based on directed acyclic graph is proposed. The
complex tasks are decomposed into several simple sub-
tasks which are processed in parallel with the
corresponding operator nodes, to realize distributed
processing and to improve the efficiency of processing and
execution. The simulation results indicate that our method
is more efficient in lower RAM consumption, processing
time, and others, and the efficiency of data stream
processing for Internet of things is improved.
Keywords—Internet of things; complex event processing; query
plan; distributed parallel processing
I. INTRODUCTION
As a new generation network technology, IOT which is
known as the third wave of information after the computer and
the Internet, has been widely admired by people from all
walks of life and is gradually applied to health care, intelligent
transportation, logistics transportation and other fields [1].The
IOT data are always uncertain, magnanimous and
heterogeneous, so how to quickly find the required data [2]
and how to excavate the implicit internal relations among data
and how to extract and abstract meaningful events have been a
matter that needs to be studied urgently.
As one of the IOT technologies, complex event processing
(CEP) technology has been widely used in all kinds of fields.
Researchers have won some valuable achievements in the
field. The most typical complex event processing system is
SASE [3], which is proposed by Eugene from University of
California, Berkeley. It is a query plan-based complex event
detection technology that can share event operator according
to the modeling characteristics of query plan. But SASE uses
active instance stack (AIS) to store intermediate results,
causing a large number of backtracking during detection.
Aiming at the problem, WANG [4] proposes a distributed
CEP architecture and a high-performance CEP method
(HPCEP) based on SASE. However, other problems present
that the use of large sliding windows for the constraint of data
out of date needs to traverse in each AIS and the time cost is
high.
In order to realize queries to complex events and
processing of queries, the research in WANG [5] has made
some extension based on Petri nets. But it only directs at
studying query processing and optimization technology of
single complex event and lack of researches on multiple
complex ones.
SONG [6] proposes an extended sub-query sharing
approach, which realizes query sharing of multiple RFID
complex events by analyzing the rich semantics of queries.
But the approach is only applicable for the condition where
the amount of complex events is small, and if the amount
increases, the memory space of system would be exceeded.
To cure the above problems, a distributed complex event
processing architecture for Internet of things is presented in
this paper, and an architecture based on directed acyclic graph
(DAG) for distributed query plan complex event processing is
also described. Meanwhile, based on the above, a Distributed
Query Plan Complex Event Processing Algorithm based on
Directed Acyclic Graph (DQCEP) is put forward, which
decomposes complex task into several simple sub-tasks, and
allots them to the corresponding operator nodes to process
concurrently to realize distributed processing and to improve
the efficiency of processing and execution. Simulation results
show that, compares with SASE, the proposed DQCEP not
only can process events with high efficiency and reliability,
but also can process massive and real-time IOT event streams
timely and efficiently.
The remainder of this paper is organized as follows. In
section II, a distributed IOT complex event processing system
architecture is presented. In section III, we give a complex
event processing model and algorithm description based on
DAG, and in section IV simulation experiments and analysis
are introduced. Section V conclusions the paper.
II. DISTRIBUTED IOT COMPLEX EVENT PROCESSING
SYSTEM ARCHITECTURE
A distributed complex event processing system
architecture for IOT is constructed in this section. The
architecture consists of event preprocessing engine, complex
event processing engine, the CEP client and the query
interface, as shown in Fig.1. The event preprocessing engine is
mainly in charge of preprocessing the IOT events stream,
This work is supported by the National Natural Science Foundation of
China (Grant No.61262071), MOE (Ministry of Education in China) Project
of Humanities and Social Sciences (Grant No.13YJCZH233), the Applied
Basic Research Plan on the Project in Yunnan Province (Grant No.2013) and
the Yunnan Normal University Graduate Student Scientific Research
Innovation Fund Project.
978-1-4799-8729-0/15/$31.00 © 2015 IEEE
The 5th Annual IEEE International Conference on
Cyber Technology in Automation, Control and Intelligent Systems
June 8-12, 2015, Shenyang, China