Hierarchical Trust Management for Wireless Sensor Networks and Its Application to Trust-Based Routing

1
Hierarchical Trust Management for Wireless Sensor
Networks and Its Application to Trust-Based
Routing

• Fenye Bao, Ing-Ray Chen, Moonjeong Chang
• Presented by Scott Hackman
• 03 November 2011

2
Introduction

• Cluster-based approach to creating a system for
  wireless routing better than shortest-distance
  and flood-based routing.
• Utilizes Social Networking and Quality of Service
  (QoS) techniques to model the behaviors of nodes
  to determine their reliability.
• Highly scalable due to being a cluster-based
  model.

3
Wireless Sensor Network

• A Wireless Sensor Network (WSN) refers to a
  distributed network of autonomous sensors, each
  operating independently for the greater good of
  the network.
• A WSN is inherently unstable due to the
  independence of the Sensor Nodes (SN) and their
  different operating characteristics, including
  malicious and selfish activity.
• The WSN must take input from its SNs, evaluate
  their input, and determine the overall picture
  for what is happening across its network.

4
Sensor Node

• A SN monitors physical or environmental
  conditions, such as temperature, sound,
  vibration, pressure, motion, or pollutants.
• A SN is can transmit, or forward information
  through multi-hop routing.
• SNs have very limited resources
• Energy
• Memory
• Computational Power
• May be susceptible to malicious attacks when
  their energy is low.

5
Cluster Head

• A Cluster Head (CH) is a node that has been
  elected to take charge of a group of SNs.
• A CH receives direct input from each of its SNs.
• A CH is responsible for reporting to all of the
  other CHs in the system.
• CHs use more energy than SNs.

6
Abnormal Node Behavior

• Malicious Node
• A node may be captured by the enemy at any point
  and start passing erroneous information or drop
  packets.
• A node is more likely to become malicious if it
  has low energy or if it is surrounded by
  malicious nodes.
• Selfish Node
• A node may become selfish if its energy becomes
  low relative to its neighbors.
• Selfish can be thought of as efficient. If a
  node recognizes that its battery level is low and
  its neighbors have sufficient energy, it may
  start dropping packets so its neighbors pick up
  more of the burden.
• The challenge becomes How do we create a model
  such that malicious and selfish nodes can be
  identified and the WSN can adjust to these
  conditions to achieve a near-optimal performance?

7
System Model

• First, how do we determine which nodes are SNs
  and which nodes are CHs?
• HEED (Hybrid energy-efficient, distributed) The
  CHs must have higher energy and have relative
  proximity. This will allow for higher energy
  consumption as well as optimal communications.
• SNs will collect data and evaluate their peers.
  That information will be passed to their
  respective CHs.
• The CHs will collect the SNs data and collect
  their own peer-to-peer (P2P) data from other CHs.
• CHs will pass their data to a CH Commander for
  evaluation.

8
How Does Trust Factor In?

• Once the hierarchy is established, the
  evaluations completed by each node follow a trust
  scheme that allows for direct and indirect
  trust-based reporting.
• Trust is established by evaluating directly, and
  indirectly, four different factors
• Energy
• Measures competence
• Less susceptible to malicious attacks
• Unselfishness
• Measures cooperativeness
• Honesty
• Whether or not the node is compromised based on
  intrusion detection capabilities in the system
  based on software-based code attestation
• Intimacy
• Relative degree of interaction experiences
  between two nodes

9
Evaluation Process

• A weighted evaluation is performed and all four
  categories are factored into one, overall trust
  score
• Tij(t) denotes the trust that node i has toward
  node j at time t.

10
Peer-to-Peer Trust Evaluation

• P2P Trust Evaluation is performed between SNs and
  between CHs.
• When node i evaluates its trust toward node j, it
  snoops, or overhears enough data to provide
  direct observation. (It is assumed, notationally,
  that i and j are direct neighbors.)
• When i evaluates a node that is beyond its
  communication range, we refer to the node as node
  k.
• Node i cannot directly evaluate k, so it must
  rely on the information passed to it by some node
  j and multiply that evaluation by a weight that
  correlates to is trust toward j.

11
Peer-to-Peer Trust Evaluation

• This relationship is represented as
  follows
• ? and a represent weights associated with trust
  decay. X represents one of the trust components.

12
Peer-to-Peer Trust Factors

• - This measures the level of
  interaction experiences. It is computed by the
  number of interactions between node i and j over
  the maximum number of interactions between node i
  and any neighbor node over the time period 0,
  t.
• - This refers to the
  belief of node i that node j is not compromised
  base on node is direct observations toward node
  j. It can be a binary quantity, 0 or 1, based on
  the result of Intrusion Detection System (IDS)
  deployed on node i about whether or not node j is
  compromised at time t.

13
Peer-to-Peer Trust Factors

• - This indicates the percentage of
  node js remaining energy that node i directly
  observes at time t. Node i can overhear or even
  monitor node js packet transmission activities
  over the time period 0, t to estimate this
  value.
• - This provides
  the degree of unselfishness of node j as
  evaluated by node i based on direct observations
  over 0, t. Node i can apply overhearing and
  snooping techniques to detect selfish behaviors
  from node j.

14
Other Parameters Defined

• a - Weight that represents a more instantaneous
  evaluation, since the higher a, the more weight
  is given to time t.
• ß Represents the impact of indirect
  recommendations.
• ?
• These parameters are used to adjust the trust
  decay over time. Lower factors cause a dampening
  effect that puts more weight on past events. This
  reduced high rates of change and may stabilize a
  system that receives sporadic, erroneous
  readings.

15
CH-to-SN Trust Evaluation

• Once all calculations are complete for a given
  time period t, the CH applies statistical
  analysis principles to all Tij(t) values received
  to perform CH-to-SN trust evaluation toward node
  j.
• CH can also detect any outliers in the cluster to
  see if any good-mouthing or bad-mouthing is
  occurring.
• The CH can exclude a sensor or reroute with the
  information it obtains.

16
Performance Model

• To create an objective model for comparison, a
  stochastic Petri net model is used.
• The Petri new model essentially computes the same
  values, but takes away the trust aspect. All
  values are known by the model at all times and
  routing data is computed accordingly.
• The underlying data of this model is used by the
  trust-based simulation, but each component can
  only see the data as defined by the initial
  conditions. Hence, best-case scenario, the
  trust-based approach can only perform as well as
  the objective Petri-net model.

17
Petri Net Model
18
Petri Net Model - Energy

• Energy represents the remaining energy in a node.
• A token will be expended from Energy when
  T_ENERGY triggers.
• Energy consumption rates

19
Petri Net Model - Selfishness

• A node may become selfish to save energy.
• An unselfish node may decide whether it will be
  selfish or not upon every time interval Ts
  according to its remaining energy and the number
  of unselfish neighbors.
• A selfish node may become redeemed based on trust
  evaluation.

20
Petri Net Model - Honesty

• A node becomes compromised when T_COMPRO fires
  and places a token in CN.

21
Subjective Trust Evaluation

• If j is a selfish node (a/c), compromised node
  (b/c) or normal node (c/c)

22
Objective Trust Evaluation
23
Trust Evaluation
24
Trust Evaluation
25
Geographic Routing
26
Geographic Routing
27
Geographic Routing
28
Geographic Routing
29
Conclusion

• This model presents a very practical framework
  that allows for highly reliable transmissions
  with reduced overhead.
• Social networking and QoS methods allow peers to
  quantitatively rate their peers, drastically
  reducing the work needed to be done by the
  cluster head.
• This model remains highly scalable because of its
  hierarchical nature.
• Possible Future Work Apply a genetic algorithm
  to this model and train it off of real-world data
  to achieve optimal weighting factors.