Final published version
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
}
TY - GEN
T1 - Optimizing push/pull envelopes for energy-efficient cloud-sensor systems
AU - Xu, Y.
AU - Helal, Sumi
AU - Thai, M.T.
AU - Schmalz, M.
AU - SIGSIM, ACM
PY - 2011
Y1 - 2011
N2 - Unlike traditional distributed systems, where the resources/needs of computation and communication dominate the performance equation, sensor-based systems (SBS) raise new metrics and requirements for sensors as well as for computing and communication. This includes sensing latency and energy consumption. In this paper, we present a performance model for SBS based on a three-tier architecture that uses edge devices to connect massive-scale networks of sensors to the cloud. In this architecture, which we call Cloud, Edge, and Beneath (CEB), initial processing of sensor data occurs in- and near-network, in order to achieve system sentience and energy efficiency. To optimize CEB performance, we propose the concept of optimal push/pull envelope (PPE). PPE dynamically and minimally adjusts the base push and pull rates for each sensor, according to the relative characteristics of sensor requests (demand side from the Cloud) and sensor data change (supply side from Beneath). We demonstrate the CEB architecture and its push/pull envelope optimization algorithm in an experimental evaluation that measures energy savings and sentience efficiency over a wide range of practical constraints. In addition, from the experiments we demonstrate that by combining PPE optimization algorithm with lazy sampling algorithm, we can achieve further energy saving. Copyright 2011 ACM.
AB - Unlike traditional distributed systems, where the resources/needs of computation and communication dominate the performance equation, sensor-based systems (SBS) raise new metrics and requirements for sensors as well as for computing and communication. This includes sensing latency and energy consumption. In this paper, we present a performance model for SBS based on a three-tier architecture that uses edge devices to connect massive-scale networks of sensors to the cloud. In this architecture, which we call Cloud, Edge, and Beneath (CEB), initial processing of sensor data occurs in- and near-network, in order to achieve system sentience and energy efficiency. To optimize CEB performance, we propose the concept of optimal push/pull envelope (PPE). PPE dynamically and minimally adjusts the base push and pull rates for each sensor, according to the relative characteristics of sensor requests (demand side from the Cloud) and sensor data change (supply side from Beneath). We demonstrate the CEB architecture and its push/pull envelope optimization algorithm in an experimental evaluation that measures energy savings and sentience efficiency over a wide range of practical constraints. In addition, from the experiments we demonstrate that by combining PPE optimization algorithm with lazy sampling algorithm, we can achieve further energy saving. Copyright 2011 ACM.
KW - Cloud computing
KW - Energy efficiency
KW - Optimization
KW - Performance
KW - Pervasive computing
KW - Push pull envelope
KW - Sentience efficiency
KW - Demand-side
KW - Distributed systems
KW - Energy efficient
KW - Experimental evaluation
KW - Optimization algorithms
KW - Performance Model
KW - Push pull
KW - Sampling algorithm
KW - Sensor based systems
KW - Sensor data
KW - Three-tier architecture
KW - Algorithms
KW - Communication
KW - Data handling
KW - Energy utilization
KW - Network architecture
KW - Sensors
KW - Ubiquitous computing
U2 - 10.1145/2068897.2068904
DO - 10.1145/2068897.2068904
M3 - Conference contribution/Paper
SN - 9781450308984
SP - 17
EP - 26
BT - MSWiM '11 Proceedings of the 14th ACM international conference on Modeling, analysis and simulation of wireless and mobile systems
PB - ACM
CY - New York
ER -