Information Technology Reference
In-Depth Information
Application Support Network
Data Analysis
and Sensor
Configuration
WBSN
Personal Servers
(cell phones)
Data Centers
Energy Optimization
Energy-aware resource
management
Decision
Making
System
(Reputation
System)
Knowledge
Storage and
Inference
System
Energy-aware resource
configuration
...
Power Model
FIGURE 12.1
Overview of the proposed energy analysis and optimization system.
• Sensing infrastructure: Global energy optimization requires a clear
understanding of the current state of the network, the characteris-
tics of the different resources, and of the analysis to be performed.
Therefore, additional hardware (HW) or software (SW) sensors
should be added to the system to obtain insight.
• Data analysis and sensor configuration: Not every sensor has the
same importance to understand the power consumption character-
istics of the different components. After a careful analysis, the sens-
ing infrastructure has to be configured to provide only the relevant
data at the required rate for the power model to be useful and to
minimize the energy overhead.
• Storage and inference system: The data provided by the sensing
infrastructure has to be stored and statistically analyzed in search
of recurrent behaviors that could lead to simple but accurate enough
power models to be used for proactive optimizations. Although the
data provided by the sensors is low level, simple inference tech-
niques can be used to raise the level of abstraction, for example,
Search WWH ::
Custom Search