姚其楨 鄧海

摘? ?要：隨著無線通信業(yè)務(wù)的不斷增長和智能終端的日益多樣化，無線頻譜資源變得愈加緊缺，認(rèn)知無線電作為一項(xiàng)能夠有效解決頻譜利用率過低的技術(shù)，近年來引起了學(xué)術(shù)界和工程界的廣泛關(guān)注。在認(rèn)知無線電網(wǎng)絡(luò)中，次用戶以合作頻譜感知的方式智能判斷主用戶的工作狀態(tài)，以便在授權(quán)頻譜的空閑時(shí)段內(nèi)接入并加以使用。將全雙工通信技術(shù)引入到傳統(tǒng)的認(rèn)知無線電網(wǎng)絡(luò)中，實(shí)現(xiàn)了頻譜感知和數(shù)據(jù)傳輸?shù)耐?，既能確保數(shù)據(jù)傳輸?shù)倪B續(xù)性，又能減少對主用戶的干擾，對提升無線頻譜利用率具有重要作用。對于多信道全雙工認(rèn)知無線電網(wǎng)絡(luò)的信道分配，建立了次級網(wǎng)絡(luò)吞吐量最大化模型，并通過懲罰函數(shù)法，將混合整數(shù)非線性規(guī)劃問題轉(zhuǎn)換成不帶約束條件的非線性規(guī)劃問題，提出了基于遺傳算法的全雙工認(rèn)知無線電網(wǎng)絡(luò)信道分配策略。仿真實(shí)驗(yàn)結(jié)果表明，該算法能在綜合考慮算法復(fù)雜度和性能的情況下，完成信道分配并保證次級網(wǎng)絡(luò)吞吐量的最大化。

關(guān)鍵詞：全雙工認(rèn)知無線電網(wǎng)絡(luò);合作頻譜感知;遺傳算法;信道分配

中圖分類號：TP393.2? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 文獻(xiàn)標(biāo)識碼：A

Research on Channel Assignment Strategy Based on

Genetic Algorithm in Full-duplex Cognitive Radio Networks

YAO Qi-zhen DENG Hai

（College of Computer Science and Technology，Nanjing University of Aeronautics and Astronautics，Nanjing，Jiangsu 210016，China）

Abstract：With the continuous growth of wireless communication services and the growing diversity of intelligent terminals，wireless spectrum resources are becoming more and more scarce. Cognitive radio has attracted extensive attention in the academic and engineering fields as a technology that can effectively improve the spectrum utilization in recent years. In cognitive radio networks （CRNs），the secondary users（SUs） intelligently judge the state of the primary user（PU） in a cooperative spectrum sensing（CSS） manner to access and use the idle time of the licensed spectrum. Full-duplex（FD） communication is introduced to conventional CRNs，in order to realize the synchronization of spectrum sensing and data transmission which not only ensures the continuity of data transmission，but also reduces interference to PU and plays an important role in improving the utilization of the wireless spectrum. A secondary networks throughput maximization（SNTM） model is established in this paper for the channel assignment in multi-channel FD-CRNs. The mixed-integer nonlinear programming（MINLP） problem is transferred to a nonlinear programming（NLP） without constraint problem by penalty function. Moreover，a channel assignment based on genetic algorithm in FD-CRNs is proposed in this paper. The simulation reveals that the algorithm can complete channel assignment and ensure the maximum throughput of the secondary network while considering the complexity and performance of the algorithm.

Key words：full-duplex cognitive radio networks; cooperative spectrum sensing; genetic algorithm; channel assignment