Abstract—Voice frequency pattern is asymmetric just like nature. Although voice frequency lies between 0 ~ 3.5 KHz, a person talking some specific word in different times is always consisting of a very narrow range of frequencies which are varying person to person. Voice frequency of the selective words used by the subscriber at the beginning of conversation like Hello, Good Morning, Namaskar etc is taken as first entity for authentication purpose. We propose second entity as probability of salutation or greeting word from subscriber’s talking habit (set of salutation words) whiles the subscriber starts a call. Third entity is chosen as probable location (place) of the subscriber from the selective locations at the time of initializing a call i.e. distance between the subscriber and the network. These three entities such as probability of particular range of frequencies for the salutation word, particular salutation or greeting word, location at the time of starting a call are used with most frequently, more frequently and less frequently by the subscriber like uncertainty in Artificial Intelligence (AI). Now different relative grades are assigned for most frequently, more frequently and less frequently used parameters and the grades are modified according to the assigned weightage of the relative grades. Then Fuzzy operations are performed on modified relative grades (sets). We invent a Fuzzy Rule (condition). If the results obtained from fuzzy operations are satisfied by the invented fuzzy rule, the subscriber (MS) and the network (MSC or PDSN) is mutually authenticated in 3-G mobile communications.

Index Terms—AAA server, Artificial Intelligence, Fuzzy set, Fuzzy operation, Subscriber (MS), Network (MSC or PDSN), SIM, Salutation word, USIM.

Pijush Kanti Bhattacharjee is an Assistant Professor in the Department of Electronics and Communication Engineering, Bengal Institute of Technology and Management, Santiniketan, WB, India. He was an Ex Asssitant Director in the Department of Telecommunications (DoT), Government of India, India. He is a member of IACSIT, IAEng, CSTA, IE. (phone: +91-33-25954148; email: pijushbhatta_6@hotmail.com).
Chandan Koner is an Assistant Professor in the Department of Computer Science and Engineering, Bengal Institute of Technology and Management, Pin-731236, WB, India. He is pursuing PhD course. (phone: +91-9434535556; email: chandan_durgapur@yahoo.com).
Chandan Tilak Bhunia is a Director, Bengal Institute of Technology and Management, Santiniketan, Pin-731236, India. He is a Senior Member of IEEE, FIE (I) and FIETE. (phone: +91-9434033157; email: ctbhunia@vsnl.com).
Ujjwal Maulik is currently a Professor in the Department of Computer Science and Technology, Jadavpur University, Kolkata, India. (phone: +91-33-24131766; email: ujjwal_maulik@yahoo.com).

Cite: Pijush Kanti Bhattacharjee, Chandan Koner, Chandan Tilak Bhunia and Ujjwal Maulik, "Artificial Intelligence Based Authentication Technique With Three Entities In 3-G Mobile Communications,"  International  Journal  of
Computer and Electrical Engineering vol. 2, no. 2, pp. 339-344, 2010.