H.-W. Tseng, M.-H. Chuang, Y.-H. Chou, Y.-H. Lee, Y.-G. Jan, Y.-K. Wang, and C.-W. Su (Taiwan)
OFDMA (Orthogonal Frequency Division Multiplexing Access), Genetic Algorithm(GA), Subchannelization, Scheduling
In OFDMA (Orthogonal Frequency Division Multiplexing Access) communication when it serves many users it becomes an important issue of how to design a scheduler to optimally serve these users, in other words we need to design a scheduler to efficiently allot these users to the limited available and less sub-carriers. Usually it is dubbed as a NP-hard problem for this kind of scheduling problem when we try to search for a solution that is resulting from many possible permutation and combinations of system parameters. The Genetic Algorithm is the algorithm developed from the animation of the natural evolution principle that it helps us to fast and effectively solve the NP problems by utilizing the characteristic of multiple-point searching and following the principle of the survival of the fittest. In this paper we will devise and develop improved and realizable hardware architecture for the Genetic Algorithm based architecture, and to implement this kind of architecture to find the optimal scheduler so as to improve the transmission efficiency and to reduce the operating cost. Finally we will list the possible limitations in the hardware implementation, the number of chromosomes formed in the Genetic Algorithm, its convergent efficiency when it is in the process of simulation analysis in the hardware realization, and also set up a suitable architecture for the hardware implementation.
Important Links:
Go Back
