動態頻道分配(Dynamic Channel Assignment)是實現感知無線電(Cognitive Radio)概念中一項重要的工作。使用者量測所在位置的環境資訊,傳給系統端相關的資訊;系統端依據所獲得的頻譜監測資訊,找出空閒的無線電頻道,動態地分配頻道給使用者。由於使用者會移動,造成網路拓樸改變,造成原本使用相同頻道的使用者,因為移動至相鄰位置而產生干擾,降低傳輸效能,並需要更換新的通道。一旦發生通道衝突(Channel Collision)的情況,就是系統資源的浪費,系統就必須因應網路環境的變化重新計算以分配頻道。本論文利用預測使用者移動(Mobility Prediction)的資訊,來進行頻道的分配,盡可能降低通道衝突,服務最多的使用者,以增進頻譜的使用效率。透過對使用者移動歷史記錄的分析,預測使用者未來移動的位置。由於能得知未來的位置,當系統進行動態頻道分配時,不僅能在當下能使相鄰使用者不會使用相同頻道而產生干擾,也在可預測的期間內,避免干擾的發生。我們提出多種預測與頻道分配的演算法,說明系統架構和頻道分配與處理方式。我們對提出的頻道分配演算法進行模擬分析,驗證的確可減少系統重新分配頻道的機率,提升系統資源的有效利用。
Dynamic channel assignment is an important work in the concept of cognitive radio. Users measure the environmental information of their location and transmit the information to the system. System identifies idle channels and dynamically allocates channels to users according to the information of measurement. Because some users with same channel may move to the adjacent position in the future and interfere with each other, the transmission performance would decline and thus the communication would fail. When channel collisions happened, the system must reassignment channels to cope with the change of network environment. This paper presents some dynamic channel assignment algorithms based on mobility prediction to enhance the spectrum efficient and minimize channel collisions. System can predict the user’s future position according to their movement. If we can predict the future position of CR user, we can avoid channel collisions and interference of users in predictable period. We present channel assignment algorithms, illustrate system architecture and estimate the performance of these channel assignment approaches. Our simulation proves that the channel assignment algorithms can reduce the amount of system reassignment channels and enhance the utilization of system resources.