Information Technology Reference
In-Depth Information
in terms of trust aggregation accuracy, convergence rate, and message over-
head.
6.10 Trust Establishment in Wireless Sensor Networks
As discussed in Chapter 2, wireless sensor networks (WSNs) are consid-
ered practical P2P distributed processing platforms for many applications
(e.g., battle-field data processing). Trust establishment has long been a chal-
lenging problem in WSNs due to high security requirements and strict re-
source constraints in WSNs. Recent approaches based on randomized key
pre-distribution schemes mainly focus on key allocations supported by the
pre-deployment estimations of the post-deployment information items. Un-
fortunately, such information items may be unavailable or may change over
time. The performance of the resulting network may be unstable and unable
to react to the change of topology due to sensor node dynamics.
In the following, we present a detailed survey of recently proposed tech-
niques for trust establishment in WSNs.
6.10.1 Symmetric Key-Based Approaches
Generally, these schemes are further divided into two categories: determin-
istic and probabilistic key pre-distribution schemes.
6.10.1.1
Deterministic Key Pre-Distribution Schemes
A deterministic scheme assigns keys to each node intentionally so that
these keys are used for specific purposes. Since each key is selected care-
fully, it is expected that the number of keys stored for each node will not be
large; otherwise, the cost of key assignment will be very high if the network
is deployed in a large scale. Thus, this scheme usually focuses on the com-
munication between a fixed number of entities such as node-to-base station,
node-to-gateway, and gateway-to-base station. A simple example is the mas-
ter key approach in which each node communicates with a single common
key [Kwok, 2007].
Jolly et al. [Jolly et al., 2003] proposed a lightweight key management
protocol. They observed that sensor-to-sensor secure communication is not
always necessary in some applications. Thus, very few keys (typically two)
are enough to establish secure connections between nodes and base station
as well as cluster gateways. Since there is no sensor-to-sensor communication,
the first key is used to talk with the base station while the second one is used
to talk with the cluster gateway. These two keys can be computed e ciently
and distributed to sensor nodes before deployment. Their objective is to pro-
Search WWH ::




Custom Search