無線蜂巢網路之高效率D2D射頻選擇機制設計

Abstract

隨著行動設備的流行普及，越來越多使用者可以透過智慧型裝備存取無線網路，因此行動網路數據的需求量也日與遽增，cisco也預測未來行動網路數據需求會有爆炸性的成長，其中又以影片串流的資料佔據所有行動網路數據高達約70\%，屆時將會對基地台造成非常大的負荷量，因此我們希望利用點對點傳輸針對基地台做出資料卸載的行為，以減輕基地台傳輸資料的負擔，點對點傳輸(Device-to-device (D2D))，可允許兩台裝置不透過基地台直接連結傳輸資料，我們針對LTE以及Wi-Fi兩種傳輸介面作探討，直覺來看，我們會希望使用Wi-Fi當為D2D傳輸介面來減輕基地台負擔，原因是Wi-F傳輸不會耗費基地台資源，但我們發現在某些情況下使用LTE作為D2D傳輸的介面其實對基地台而言是較節省的。因此我們希望透過適當選擇不同的點對點傳輸介面與傳輸的設備來達到讓基地台花費最少資源的目的。另一方面，每個使用者觀看影片的服務品質需要被滿足，但Wi-Fi較難精準的計算出傳輸的速率，在這部份我們利用了實際測量以及其他有關預測無線干擾環境下的傳輸速率模型來評估速率。我們提出了動態規劃演算法來解決這個問題，實驗結果驗證了我們的發現：只透過Wi-Fi來做D2D傳輸對基地台而言並非最節省資源的手段，也顯示出我們的方法可以為基地台節省顯著比例的資源。

Cisco has predicted that there will be an explosive growth of the mobile traffic data. In addition, the mobile video data demand will occupy nearly 70\% of the whole mobile traffic demand. Thus, eNB is going to face to heavy traffic load. In this thesis, we$’$d like to use D2D technique to do data offloading for cellular networks. The concept of Device-to-device (D2D) is to allow direct transmission between two devices without an infrastructure station$’$s help. Here, we target on two different interface, Wi-Fi and LTE of D2D transmission link, namely Wi-Fi direct and LTE direct. Intuitively, transmission through Wi-Fi direct can help to relieve loading on BS. However, we find that transnission through LTE direct may saved more resource than through Wi-Fi direct. We aim to minimize the using LTE resource block by the help of D2D relay and by selecting the different interface of transmission links properly. Also, the QoS of each user should be guaranteed. As we know, it$’$s hard to correctly evaluate the data rate of a Wi-Fi transmission link. So the capacity of a Wi-Fi transmission link under certain interference environment will be evaluated by our real measurement and interference models proposed in previous related works. We propose a dynamic program algorithm to solve it. The result shows that a large proportion of resource blocks can be saved. A significant data offloading can be done for eNB in cellular networks. And the result also proves our observation: transmission through Wi-Fi direct may not save most resource.