Time Synchronization for Wireless Sensor Networks - Adaptive Signal Processing in Wireless Communications - page 369

forwarding packets from a source to a destination. The feasibility of WSNs keeps growing
rapidly, and WSNs have been regarded as fundamental infrastructures for future ubiq-
uitous communications due to a variety of promising potential applications: monitoring
the health status of humans, animals, plants, and environment; control and instrumen-
tation of industrial machines and home appliances; homeland security; detection of
chemical and biological threats and leaks; etc. [1, 2].
Time synchronization is a procedure for providing a common notion of time across
a distributed system. It is crucial for WSNs in performing a number of fundamental
operations, such as:

• Data fusion: Data fusion is a main operation in all distributed networks for

processing and integrating in a meaningful way the collected data, and it
requires some or all nodes in the network to share a common timescale.
Power management: Energy efficiency is a key designing factor for WSNs •

since sensors are usually left unattended without any maintenance and bat-
tery replacement for their lifetimes after deployment. Most energy-saving
operations strongly depend on time synchronization. For instance, the duty
cycling (sleep and wake-up modes control) helps the nodes to save huge energy
resources by spending minimal power during the sleep mode. Thus, network-
wide synchronization is essential for efficient duty cycling, and its performance
is proportional to the synchronization accuracy.
Transmission scheduling: Many scheduling protocols require time synchroni-

•

zation. For example, the Time Division Multiple Access (TDMA) scheme, one
of the most popular communications schemes for distributed networks, is only
applicable to a synchronized network.

Moreover, many localization, security, and tracking protocols also demand the
nodes to time stamp their messages and sensing events. Therefore, time synchroniza-
tion appears as one of the most important research challenges in the design of energy-
efficient WSNs.
In general, synchronization is considered a critical problem for distributed wireless ad
hoc networks due to its decentralized nature and the timing uncertainties introduced by
the imperfections in hardware oscillators and message delays in physical and Medium
Access Control (MAC) layers. All these uncertainties cause the local clocks of different
nodes to drift away from each other over the course of a time interval. In the context
of the Internet (a kind of distributed network), time synchronization has been thor-
oughly studied and investigated. In the Internet, the Network Time Protocol (NTP) [3] is
employed ubiquitously due to its diverse advantages, such as scalability, robustness, and
self-configurability. Besides, NTP does not rely on GPS and is a software-based protocol.
However, NTP presents a number of challenges when applied to WSNs due to the unique
nature of sensor networks: limited power resources, wireless channel conditions, and
dynamic topology caused by mobility and failure. Therefore, different types of synchro-
nization schemes have to be explicitly designed for WSN applications to cope with these
challenges (see also the surveys in [ 4 - 9 ] for additional motivations in this direction).