| 摘要: | 感知無線電網路(Cognitive radio network)提供了解決頻譜使用率不佳的環境，它讓那些未持照使用者(unlicensed user)也可以在不影響持照使用者(licensed user)的方式下共同使用頻譜。在此環境下也存在一般網路環境會碰到的因素，整個網路傳輸品質會隨著地理位置或是空間不同而改變，再加上空閒頻譜會隨著持照使用者未使用才會出現，因此整個網路環境是變動的。該如何有效的在這動態的環境中利用這閒置的頻譜成為一個重要的議題，在目前許多方法會假設在網路中會有一個資訊整合中心(Fusion center)，它除了可以整合環境的資訊之外也可以採用統一決策的方式來有效率分配這空閒的頻譜給未持照使用者，也可以隨著動態的環境做調整，但如果網路中不存在這樣的腳色來分配頻譜，未持執照的使用者將變成各自決策使用的通道，頻譜容易陷入競爭而沒有妥善利用，造成使用率仍不佳。本論文在考慮動態的通道狀況與分散式的通道決策方式下，提出一個通道決策方法讓這些未持執照的使用者能彼此協調通道的使用，進而提升頻譜使用率，在模擬的結果中顯示提出的方法能夠提升頻譜的使用率來增加系統的效能並且系統中未持照的使用者能之間能夠在換手造成的 overhead 方面維持一定的公平性，然而在動態調整參數下能夠以不降低系統效能的狀況下降低多餘的換手次數來降低 overhead。Cognitive Radio network provides an environment which is able to solve the problem of the inefficient utilization of the spectrum. It helps the unlicensed users to make use of the spectrum together, under the condition of having no influence to licensed users. In Cognitive Radio network, the transmission quality will change due to the different location or space in the whole network, which is similar to the general network environment. Moreover, the idle spectrum would appear only in the case that the licensed user does not make use of it, which burdens the changeable network environment. Thus, how to make efficient use of the idle spectrum becomes an important issue in the dynamic environment. In many methods presently, it is assumed that there is Fusion center in the network, which is able to integrate the environment information, allocate idle spectrum to unlicensed users efficiently in the centralized decision-making method, and adjust it according to the dynamic environment. However, if there is no Fusion center to allocation the spectrum, then unlicensed users make their own decisions of the usage of channels independently. Under such condition, the spectrum utilization would be inefficient because of the contention.In this paper, a channel decision-making method considering the dynamic channel condition and the distributed channel decision-making is proposed, which is able to make the unlicensed users coordinate the channel utilization and improve the spectrum efficiency. In the simulation, the result shows that the proposed method can improve the system performance by improving the spectrum utilization, and maintain certain fairness in the aspect of overhead resulting from the handover between unlicensed users. However, under the dynamic adjustment of parameters, the overhead would be decreased due to the fewer numbers of unnecessary handover while keeping the system performance. |
