Affiliations: [a] Department of Computer Science and Engineering, Vivekanadha College of Engineering for Women, Elayampalayam, Tiruchengode – TK, Tamilnadu, India
| [b] Department of Information Technology, Vivekanadha College of Engineering for Women, Elayampalayam, Tiruchengode – TK, Tamilnadu, India
| [c] Department of Computer Science and Engineering, SACS MAVMM Engineering College, Madurai, TamilNadu, India
| [d] Department of Computer Science and Engineering, Muthayammal Engineering College, Rasipuram, Tamilnadu, India
Correspondence:
[*]
Corresponding author. M. Sayeekumar, Department of Information Technology, Vivekanadha College of Engineering for Women, Elayampalayam, Tiruchengode – TK, Namakkal District, Tamilnadu – 637 205, India. E-mail: sayee.academic@gmail.com.
Abstract: The main challenge of problem lies in the perception of Cognitive Radio technology is to discover licensed empty spectrum pattern. The efficient model is needed for allocation among licensed and unlicensed users in wireless spectrum to improve the extraction rate and collision rate. To discover the spectrum hole in spectrum paging bands, stirred by FP mining technique proposed an efficient enumeration approach, namely Constraint Based Frequent Periodic Pattern Mining (CBFPP). The proposed algorithm uses TRIE-like data structure with data mining constraints. CBFPP algorithm predicts periodic spectrum occupancy holes in the paging bands. It is shown that CBFPP has a high prediction accuracy with reasonable time complexity. Experiment with synthetic and real data validate higher prediction accuracy and with reasonable time complexities. The unlicensed user utilizes the predicted spectrum pattern in spectrum usage of channel without significant interference to licensed users.
Keywords: Cognitive radio, data mining, frequent pattern, spectrum occupancy
prediction
