【摘要】：目前世界各國主流的頻譜分配策略是靜態(tài)分配方式,大多采用授權(quán)許可制度。然而許多授權(quán)用戶,并非一直占用授權(quán)頻段,許多頻段處于空閑狀態(tài),這直接導(dǎo)致了的頻譜利用率低下。在這樣的情況下,Joseph Mitola于1999年在軟件無線電的基礎(chǔ)上提出了認(rèn)知無線電(CR,Cognitive Radio)的概念,闡述了機會式頻譜接入的方法。然而,要實現(xiàn)動態(tài)頻譜接入,首先要解決的問題就是如何有效地檢測頻譜空穴,避免對主用戶的干擾。本文對認(rèn)知無線電網(wǎng)絡(luò)中的頻譜檢測技術(shù)進行研究,并且分別在單用戶及多用戶的模型下提出了控制干擾的最大化利用頻譜資源的算法,通過對其數(shù)學(xué)模型分析和推導(dǎo),給出了下面的科研成果:1.提出了一種自適應(yīng)調(diào)整能量檢測器門限的算法。固定的能量判決門限是傳統(tǒng)能量檢測器的特征之一,雖然能保證主用戶所受干擾限制在一定范圍之內(nèi),但未深入考慮頻譜的效率問題。自適應(yīng)門限的能量檢測技術(shù)通過動態(tài)地調(diào)整能量判決門限,在保證主用戶所受干擾滿足限制條件的同時,最大化頻譜利用率。2.提出了基于冪集置信度(PSB,Power Set Belief)的聯(lián)合檢測技術(shù)。傳統(tǒng)的聯(lián)合檢測技術(shù)中,K/N聯(lián)合檢測缺乏對每個用戶表現(xiàn)的置信度的認(rèn)識。基于冪集置信度的聯(lián)合檢測技術(shù)從Dempster-Shafer證據(jù)理論出發(fā),根據(jù)每個用戶的置信度進行有效地數(shù)據(jù)融合。3.提出了基于Adaboost的聯(lián)合檢測技術(shù)。Adaboost算法常用于圖像識別,人工智能等領(lǐng)域。結(jié)合了Adaboost的聯(lián)合檢測技術(shù),可以通過訓(xùn)練測試,達到十分優(yōu)異的檢測性能。
[Abstract]:At present, the main spectrum allocation strategy in the world is static allocation, mostly using authorized license system. However, many authorized users do not always occupy the authorized band, many bands are idle, which directly leads to low spectrum utilization. In this case, Joseph Mitola in 1999 in the software radio base. In this paper, the concept of cognitive radio (CR) is proposed, and the method of opportunistic spectrum access is described. However, to realize dynamic spectrum access, the first problem to be solved is how to effectively detect spectrum holes and avoid interference to primary users. Based on the analysis and derivation of its mathematical model, the following scientific research results are given: 1. An adaptive algorithm for adjusting the threshold of energy detector is proposed. Fixed threshold of energy decision is the characteristic of traditional energy detector. Firstly, although the interference of the primary user is limited to a certain range, the efficiency of the spectrum is not considered deeply. The adaptive threshold energy detection technique can maximize the spectrum utilization while ensuring that the interference of the primary user meets the limitation conditions by dynamically adjusting the energy decision threshold. In the traditional joint detection technology, K/N joint detection lacks the knowledge of the confidence of each user's performance. Based on the power set confidence, the joint detection technology proceeds from Dempster-Shafer evidence theory and carries on the effective data fusion according to the confidence of each user. Adaboost joint detection technology. Adaboost algorithm is often used in image recognition, artificial intelligence and other fields. Combined with Adaboost joint detection technology, can be trained and tested to achieve very good detection performance.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TN925

【參考文獻】

相關(guān)期刊論文 前2條

1 馬勇,丁曉青;基于層次型支持向量機的人臉檢測[J];清華大學(xué)學(xué)報(自然科學(xué)版);2003年01期

2 葉航軍,白雪生,徐光yP;基于支持向量機的人臉姿態(tài)判定[J];清華大學(xué)學(xué)報(自然科學(xué)版);2003年01期

，

本文編號：2191643