在感知網路中使用合作偵測之傳輸策略

Abstract

合作式偵測是一個廣為人知的技術。其最大的目的在於蒐集資料以幫助感知無線電的傳送端，使其能更有效率、更準確地偵測到頻譜缺口。然而，一個成功的傳送不只在於傳送端要有頻譜缺口，在接收端也要能同時有頻譜缺口。這個傳送端與接收端的異質現象，在之前討論合作式通訊是一直被忽略的。利用這個特性，我們利用Poisson分佈以及隨機幾何來討論，在分散式感知網路中，時麼時候利用合作式感測器才是有用的。我們發現即使在最簡單的圓盤模型之下，能夠成功傳送資料的區域也不在是對稱的。我們也進一步的發現到，這所帶來的影響是，傳送的連線也不在是雙向的。這都跟之前的研究有極大的差異。另外，我們也將這個模型延伸到多個感知使用者中的競爭行為，是如何影響到最 後的可傳送區域。我們也發現在什麼樣的狀況下，使用合作式感測器不一定會增加最終的速率。基於這些觀察上，我們也發展出了如何針對不同的接受端，來選擇不同的合作式偵測器，來最佳化感知無線電網路。

Cooperative spectrum sensing is widely known to collect sensing information from cooperative sensors to assist cognitive radio transmitter to better judge the transmission opportunities, with appearance of hidden primary system terminals. However, it is usually ignored that a successful transmission can go through by requiring not only the spectrum availability at cognitive radio transmitter, but also spectrum availability at cognitive radio receiver. Using this special observation, we use Poisson distributed nodes and random geometric graph model to analyze the geometric property of the geographic region allowing CR transmission with the help of cooperative sensors. Contrary to intuition, we find that cooperative sensing is not always helpful and the region allowing CR transmission is no longer circular anymore. Using this geometric property, we further find that transmission link can remain bidirectional only under certain geometric conditions. And we further extend this model to discuss about the competition among the secondary users and the corresponding geographic transmission region. And we find the condition on which secondary users cannot improve their throughput by using cooperative sensors. Due to these observations, we develop a complete methodology so that a secondary transmitter can select helpful cooperative sensors corresponding to different receivers.