英语论文网

ge so that an agent can weigh up the costs and benefits that a particular situation holds. Importance caters for the significance of a particular situation to the thruster based upon time.

We merge the utility and importance of a situation into a single variable called weight, which is variable and increases or decreases with time. In our model we make use of trust agents that we suppose to reside on all network nodes.

Each agent operates independently and maintains its individual trust statistics. The duty of the agent is to gather data from all previous events in all states, filters it, assigns weights to each event and computes different trust levels based upon them.

Each node basically performs the following three functions:
Trust Derivation
Trust Quantification
Trust Computation

5.2.1 Trust Derivation

We compute the trust in our model based upon the information that one node can gather about the other nodes. Node gathers information about other nodes in the network in passive mode i.e. without requiring any special interrogation packets. Vital information regarding other nodes can be gathered by analyzing the received, forwarded and overheard packets. In passive mode, the possible events that can be recorded are:

i. Data packets forwarded

ii. Control packets forwarded

iii. Data packets received

iv. Control packets received

v. Data packets precision

vi. Control Packets precision

The information from these events is classified into one or more trust categories. Trust categories signify the specific aspect of trust that is relevant to a particular application. For example, we might trust a particular node for the category “data forwarding” but not for the category of “Accurate Data Reception”.

5.2.2 Trust Quantification

Secure routing protocols represent trust levels by either the presence of security or its absence. We don’t have others options regarding trust in routing. Trust in ad-hoc networks is always in a fluid state and is continuously changing due to the mobility of the nodes. As the period of interaction with any node may be brief, it is imperative that the trust be represented as a continual range to differentiate between nodes with comparable trust levels. The better idea would have to represent trust from –1 to +1 signifying a continuous range from complete distrust to complete trust. So the trust value would have to be stored in a floating point variable. But as we know that in ad hoc networks battery life (energy) is very precious. We can’t use much of floating point variables because floating point calculation is a processing overhead: which is undesirable. So instead we use an integer value to store our results and do integer calculations.

5.2.3 Trust Computation

Trust computation involves an assignment of weights (utility/importance factor) to the events that were monitored and quantified. The assignment is totally dependent on the type of application demanding the trust level and varies with state and time. All nodes dynamically assign these weights based upon their own criteria and circumstances. For example for a particular node at a certain time control packets may be more important than data packets. So control packets with be assigned more weight than data pa本论文由英语论文网提供整理，提供论文代写，英语论文代写，代写论文，代写英语论文，代写留学生论文，代写英文论文，留学生论文代写相关核心关键词搜索。