Time Synchronization Attacks in Sensor Networks
Michael Manzo
Department of Electrical Engineering
UC Berkeley
mpm@eecs.berkeley.edu
Tanya Roosta
Department of Electrical Engineering
UC Berkeley
roosta@eecs.berkeley.edu
Shankar Sastry
Department of Electrical Engineering
UC Berkeley
sastry@eecs.berkeley.edu
ABSTRACT
Time synchronization is a critical building block in distributed
wireless sensor networks. Because sensor nodes may be severely
resource-constrained, traditional time-synchronization protocols
cannot be used in sensor networks. Various time-synchronization
protocols tailored for such networks have been proposed to solve
this problem. However, none of these protocols have been
designed with security in mind. If an adversary were able to
compromise a node, he might prevent a network from effectively
executing certain applications, such as sensing or tracking an
object, or he might even disable the network by disrupting a
fundamental service such as a TDMA-based channel-sharing
scheme. In this paper we give a survey of the most common time
synchronization protocols and outline the possible attacks on each
protocol. In addition, we discuss how different sensor network
applications that are affected by time synchronization attacks, and
we propose some countermeasures for these attack.
Categories and Subject Descriptors
C.2.2 [Computer Systems]: Security
.
General Terms
Algorithms, Security, Design.
Keywords
Sensor Network, Security, Authentication, Time-synchronization,
Clock Drift and Skew.
1. INTRODUCTION
Ad hoc networks are infrastructureless, possibly multi-hop
wireless networks where every node can be either a host or a
router, forwarding packets to other nodes in the network. Some
applications of sensor networks are in providing health care for
elderly, surveillance, emergency disaster relief, and battlefield
intelligence gathering.
A sensor network consists of anything from a handful to very
many tiny wireless sensor nodes. One very popular type of nodes
are the motes developed primarily at U.C. Berkeley. These motes
have very constrained resources. An example of a mote is the
Telos mote, which runs on an 8 Mhz, 16-bit processor having
60KB program memory and 2KB of RAM. It uses a 2.4 Ghz radio
that can communicate at up to 250 Kbps. See figure 1. Given the
limited resources of these sensor nodes, it is a key technical
challenge to design secure services, such as time-synchronization.
Time synchronization protocols provide a mechanism for
synchronizing the local clocks of the nodes in a sensor network,
and it is used by many applications of sensor networks, such as
tracking and localization. In addition, some communication
protocols in sensor networks rely on time synchronization for
scheduling node services, such as TDMA-based channel sharing
protocols. There are several time synchronization protocols for the
internet, such as Network Time Protocol (NTP). However, given
the non-determinism in transmissions in sensor networks, NTP
cannot be directly used in wireless sensor networks.
Time-synchronization implementations have been developed
specifically for sensor networks. Three of the most prominent are
Reference Broadcast Synchronization (RBS) [3], Timing-sync
Protocol for Sensor Networks (TPSN) [5], and Flooding Time
Synchronization Protocol (FTSP) [10]. The current time
synchronization protocols are designed to maximize the energy
efficiency as well as the robustness of the protocol to changing
network conditions. However, none of these protocols were
designed with security as one the goals. Security is an important
issue in sensor networks, given their diverse and usually very
sensitive applications. For example, it is crucial to protect people's
privacy when sensors are used for elderly health care monitoring.
Sensor networks are usually unattended after deployment, and
their deployment location is untrusted. In addition, nodes
communicate using a radio channel, which makes all
communications susceptible to eavesdropping. Therefore, sensor-
network security can easily be breached either by passive attack,
such as eavesdropping, or active attacks, such as denial of service
attacks, which can be launched at, for example, the routing or the
physical layer.
To provide more secure wireless communications in sensor
networks, Karlof et. al proposed and implemented TinySec [8],
which uses symmetric private-key encryption to authenticate and
encrypt messages. If an adversary physically captures a node,
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.
SASN’05, November 7, 2005, Alexandria, Virginia, USA.
Copyright 2005 ACM 1-59593-227-5/05/0011...$5.00.
107