r>
« back to results for ""
Below is a cache of http://www.cs.ucla.edu/~dabiri/papers/islped07.pdf. It's a snapshot of the page taken as our search engine crawled the Web.
The web site itself may have changed. You can check the current page or check for previous versions at the Internet Archive. Yahoo! is not affiliated with the authors of this page or responsible for its content.
Low Power Light-weight Embedded Systems* Low Power Light-weight Embedded Systems*
Majid Sarrafzadeh
1
, Foad Dabiri
1
, Roozbeh Jafari
2
, Tammara Massey
1
, Ani Nahapetan
1


Computer Science Department
1

University of California, Los Angeles
{majid, dabiri, tmassey, ani}@cs.ucla.edu

Electrical Eng. Dept. / University of Texas at Dallas
2

Electrical Eng. and Computer Sci. Dept. / UC Berkeley
rjafari@utdallas.edu

Abstract

Light-weight embedded systems are now gaining more popularity
due to the recent technological advances in fabrication that have
resulted in more powerful tiny processors with greater
communication capabilities that pose various scientific challenges
for researchers. Perhaps the most significant challenge is the
energy consumption concern and reliability, mainly due to the
small size of batteries. In this tutorial, we portray a brief
description of low-power, light-weight embedded systems, depict
several power profiling studies previously conducted, and present
several research challenges that require low-power consumption
in embedded systems. For each challenge, we highlight how low-
power designs may enhance the overall performance of the
system. Finally, we present a several techniques that minimize the
power consumption in such systems.
Categories and Subject Descriptors
B.8.2 [Performance and Reliability]: Performance Analysis and
Design Aids; C.3 [Special-purpose and Application-based
Systems]: Real-time and embedded systems
General Terms
Design, Performance
Keywords
Light-weight embedded systems, sensor networks, power
optimization
1.
Introduction
Light-weight embedded systems, recently introduced due to the
advancement of fabrication of powerful tiny processors, have the
ability to revolutionize capture, processing and actuation in
several collaborative and networked systems. The new class of
tiny embedded systems has been widely utilized in several
domains from medical monitoring applications [1][2] to
collaborative object tracking systems [3]. Many systems similar to
the aforementioned applications require low-profile, mobile and
cost-effective devices. The physical size and the cost-
effectiveness immediately deduce several constraints in
processing power and communication bandwidth. In addition, it
enforces restriction on the size of batteries. These unique
limitations require rethinking and reinventing the design process
in particularly light-weight embedded systems. Predominantly,
power issues become a major concern in the design phase due to
the unique properties of such systems. Optimization of power
consumption in light-weight embedded systems is no longer just
an objective function that is to be minimized. Power optimization
is a tight constraint that must be accommodated to deliver a
practical system.
2.
System definition
An embedded system is a special-purpose system in which a
computer is entirely encapsulated by the gadget it controls. Unlike
a general-purpose computer, an embedded system performs pre-
defined tasks, usually with very specific requirements and
constraints [4]. Since the system is dedicated to a specific task,
designers can optimize it, reducing the size and cost of the
product. Light-weight embedded systems are often referred to low-
profile, small size, unobtrusive and portable processing elements
with limited power resources. Such systems typically
incorporate sensing, processing and communications and are often
manufactured to be simple and cost-effective. These low profile
systems usually have limited computational capabilities, memory
(storage), speed and I/O interfaces. Despite their low complexity,
computationally intensive tasks impede light-weight embedded
systems from being deployed in collaborative networks in large
quantities. While their sensing capabilities allow for a seamless
integration into the physical world, their processor architecture
designs yield notable advantages such as reconfigurability and
adaptability with various applications and environments.
Consumers, on the other hand, constantly demand thinner, smaller
and lighter systems with smaller batteries in which the battery
life is enhanced to meet their lifestyle. Improving the performance
of battery life, however, has been always a major scientific
challenge for researchers. Due to its criticality, the battery life
becomes an objective as opposed to being a constraint in
traditional systems. In order to optimize the power consumption in
such systems, researchers must understand the major sources of
power consumption. Therefore, we present a power profiling
study on light-weight embedded systems in the next section. In
section 4, we portray a number of challenges with respect to
power optimization in such systems. In sections 5 and 6, we
discuss two problems that we selected in regards to low-power
light-weight embedded system design in details. Finally, Section 7
concludes the paper.
3.
Power profiling
In this section, we present a few widely used embedded systems
deployed in several monitoring and mobile applications. Motes
from CrossBow [5] are among popular candidates due to their tiny
size and on-chip sensing. For applications where interaction with
users is desired, Pocket PCs have also been broadly deployed [6].
Pocket PCs, in addition to their portability, incorporate relatively
* Second to fifth author names appear in alphabetical order.

Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
ISLPED06, October 46, 2006, Tegernsee, Germany.
Copyright 2006 ACM 1-59593-462-6/06/0010...$5.00. large displays, and communicate using commonly accepted
wireless communication protocols such as Bluetooth and 802.11
Wi-Fi. These communication capabilities facilitate real-time and
remote monitoring. Several power profiling studies have been
conducted on motes [7] [8]. Pocket PCs behave differently mainly
to due to their architecture dissimilarities with motes. Several
studies are conducted on power profiling of handheld devices [9]
[10] including Pocket PCs.
4.
Challenges
In this chapter, we describe several challenges for low-power
embedded system design. For each challenge, we begin with a
general definition, followed by low-power solutions proposed for
traditional systems and then we depict new techniques suitable for
light-weight embedded systems suggested by researchers.
4.1
Scheduling for power management
Task scheduling on single or multiple processing elements is
considered as one of the most common methods to achieve lower
power consumption. In particular, in light-weight embedded
systems, scheduling saves power by shutting down devices when
they are not operating. Processing elements in embedded systems
usually serve different requests at different times. Ordering task
execution adjusts the lengths of idle periods and exploits the
opportunities for power management [11][12][13]. Several
approaches have been proposed for task scheduling on low-power
embedded systems that consider highly constrained energy source
and environmental sources [14][15][16][17].
4.2
Software power optimization
Software constitutes a major component of today's systems, and
its role is projected to continue to grow. In traditional processors,
instruction level analysis of a processor aids in developing power
consumption models of software execution in processors.
Software power evaluation also gives the designers the ability to
optimize their programs in terms of power. Common techniques
include code compression and coding [18][19][20]. Similar
approaches have been applied to embedded systems with tiny
processors. Light-weight embedded systems are highly
constrained in terms of the memory size available to them. Most
work on code compression thus far has focused mainly on the
memory optimization. However, code compression has a
significant effect on energy consumption. Since a compressed
code is smaller in size, fewer accesses to the main memory is
required resulting in less energy consumption. Meanwhile,
reductions in memory accesses result in the reduction