Abstract:

The servers or host that are being visited by mobile agents are often the targets of an attack. The mobile agents that have been affected and became malicious might steal resources, confidential data or utilizes the host as the basis for the new attacks. In the other hand, the host itself can be malicious; it may steal the resources and confidential data from the mobile agents as well. Accordingly, a lot of researches have been conducted in this area, and the techniques for protecting the agents and the hosts have been proposed. This abstract presents a new proposed technique for protecting the host from malicious mobile agents in a big range of smart building environments, where the studies of the existing techniques are necessary in order to fit this new proposed technique in its suitable environments. Now that smart building appliances are easily plugged into smart building networks, the questions of how to improve security management, especially authentication, identification, secure messaging, certification, trusted third party and resource control for such appliances have arise.

Related Work:
There are few existing techniques to protect and prevent the attack on the mobile agent from malicious host, and the attack on the host from malicious mobile agents, each technique has its own strength and drawback, and few of the techniques are: Sandboxing, Code signing, Proof-Carrying code, and Path History, and this new proposed technique is really related to Path History technique.

Proposed Technique:
The technique proposed in this research contains;
1.Shortest path Algorithm.
2.Probability theory or fuzzy logic for the proposed mathematical model.
3.A set of defined rules for data access, memory usage, computing time and authentication process based on the probability or a range of number from the fuzzy logic.
This technique will have a set of defined rules for data access, memory usage, computing time and authentication process. The rules will have a fix range of numbers or percentages to be set, to illustrate it for this Abstract, lets just assume the numbers to be: 95%, 70% and 50% for the illustration in how the technique works. These fix set of numbers will be defined later along with the progress and finding of this thesis.
These numbers, that have been stated above, later on will be used as the scores for the path traveled by the agents. If an agent traveled through the shortest path, the score of 95% will be given to this agent, while the 70% and 50% will be assigned to the agents travel through the second best shortest path and the third best shortest path, so, the further the path traveled by an agent, the smaller score to be assigned to it. These scores later on will be used as variables in the proposed mathematical model and Algorithm for this technique. The 95%, 70% and 50% numbers are assigned on the shortest path, second shortest path and the third shortest path accordingly.
Below is the general mathematical model, and this model is formed by using variables that affect the mobile agents :
F(X) =aD + bT + cN - SP
Expected findings:
1. To Find the constants {a,b,c} that affects Agents.
2. To Find the valid mathematical and security model.
3.To Find the valid algorithm for authentication for access control and malicious mobile agents’ prediction

Keywords:

smart home, mobile agents, security, shortest path, fuzzy logic controller.