TY - JOUR
T1 - Real-time middleware for cyber-physical event processing
AU - Wang, Chao
AU - Gill, Christopher
AU - Lu, Chenyang
N1 - Publisher Copyright:
© 2019 Association for Computing Machinery.
PY - 2019/10
Y1 - 2019/10
N2 - Cyber-physical systems (CPS) involve tight integration of cyber (computation) and physical domains, and both the effectiveness and correctness of a cyber-physical system application may rely on successful enforcement of constraints such as bounded latency and temporal validity subject to physical conditions. For many such systems (e.g., edge computing in the Industrial Internet of Things), it is desirable to enforce such constraints within a common middleware service (e.g., during event processing). In this article, we introduce CPEP, a new real-time middleware for cyber-physical event processing, with (1) extensible support for complex event processing operations, (2) execution prioritization and sharing, (3) enforcement of time consistency with load shedding, and (4) efficient memory management and concurrent data processing.We present the design, implementation, and empirical evaluation of CPEP and show that it can (1) support complex operations needed by many applications, (2) schedule data processing according to consumers' priority levels, (3) enforce temporal validity, and (4) reduce processing delay and improve throughput of time-consistent events.
AB - Cyber-physical systems (CPS) involve tight integration of cyber (computation) and physical domains, and both the effectiveness and correctness of a cyber-physical system application may rely on successful enforcement of constraints such as bounded latency and temporal validity subject to physical conditions. For many such systems (e.g., edge computing in the Industrial Internet of Things), it is desirable to enforce such constraints within a common middleware service (e.g., during event processing). In this article, we introduce CPEP, a new real-time middleware for cyber-physical event processing, with (1) extensible support for complex event processing operations, (2) execution prioritization and sharing, (3) enforcement of time consistency with load shedding, and (4) efficient memory management and concurrent data processing.We present the design, implementation, and empirical evaluation of CPEP and show that it can (1) support complex operations needed by many applications, (2) schedule data processing according to consumers' priority levels, (3) enforce temporal validity, and (4) reduce processing delay and improve throughput of time-consistent events.
KW - Industrial Internet of Things
UR - http://www.scopus.com/inward/record.url?scp=85075744546&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85075744546&partnerID=8YFLogxK
U2 - 10.1145/3218816
DO - 10.1145/3218816
M3 - Article
AN - SCOPUS:85075744546
SN - 2378-962X
VL - 3
JO - ACM Transactions on Cyber-Physical Systems
JF - ACM Transactions on Cyber-Physical Systems
IS - 3
M1 - 29
ER -