行動合作網通之無線訊號與訊息處理技術研究-子計畫二：合作型感知通訊之機會式傳輸技術

Abstract

由於最近研究結果顯示，核照頻譜90%的時間處於閒置狀態，也因許多通訊設備皆使用ISM頻譜，不難想像想像不久之將來，無線通訊將面臨頻譜匱乏的窘境。感知無線是解決此問題的技術。感知技術可以讓感知使用者在不干擾使用者的前提之下進行傳輸。其中最常被討論的就是交錯法(Interweave method)，或稱為先聽後說(Listen before talk, LBT)，感知使用者必須先偵測是否有其他使用者正在使用該頻段，以避免擾亂正在進行的傳輸。此模式可應用在自我組態(self-configuring) 的無線網路，例如物聯網。 LBT有其優點，但也有其缺點。缺點是感知使用者必須等到無使用者佔用頻帶時方可使用，頻譜效能必然降低。相較於前述方法，機會式空間頻譜分享更能提升頻譜效能。拜多根天線所賜，感知傳輸端可調整能量傳輸方向以避免造能對其他使用者的干擾。然而，這種方法仍然仰賴精確的通道資訊，事實上，通道資訊都無法如此精確。本計畫中，我們將針對單一使用者以及多重使用者進行在MIMO以及MIMO-OFDM的感知系統中之穩健機會式傳輸技術之研究。藉由提出較不保守的錯誤模型，所提出的方法將會在位元錯誤率上比之前的方法有更好的效能。我們也將提出方法以改善「機率限制最佳方法」的缺點。即便如此，此所提的穩健傳輸效能也只是次佳的，這是因為假設所有的通道資源包含中繼點都是同等重要。這樣的假設並不實際，因為所有的通道包含中繼點都是會變化的。因此，針對這一點我們將分別對直接傳輸以及中繼系統傳輸提出穩健資源分配的解決方法。我們將研究在通道不確定的前提之下，資源如何做到最佳分配。此問題包含決定通道數目、中繼點數，以及通道是如何備使用的。以下是未來兩年工作項目第一年 在感知無線MIMO/MIMO-OFDM系統上開發穩健機會傳輸演算法 a. 單一使用者 b. 多重使用者 將演算法以MATLAB實現第二年 在感知無線MIMO/MIMO-OFDM系統上開發單模以及多模穩健資源分配演算法 a. 直接傳輸系統 b. 中繼協傳系統 i. 單一中繼站 ii. 虛擬天線陣列 將演算法以MATLAB實現

Recent studies have indicated that spectrum under utilization problem is a serious one, with many licensed frequency bands being unused 90 percent of the time. Moreover, with the flurry of communication devices operating in the ISM band and with a large amount being introduced, it will not be hard to envision in the future, devices designed for this band will find it difficult to find available spectrum for transmission. Cognitive radio has recently been proposed to alleviate these problems by allowing the cognitive radios to opportunistically transmit without causing noticeable interference to other users. Several different cognitive transmission techniques have already been proposed, with most attention given to “listen before talk” (LBT) or interweave method, which requires accurate detection of users already occupying the “spectrum” (frequency spectrum, spatial spectrum, temporal spectrum, etc.) so that the cognitive user does not stumble across existing users. It has been envisioned that such scenario will be applicable to self-configuring wireless network, such as Internet of Things. Despite its advantages, LBT is not an efficient method for cognitive transmission as the cognitive users have to wait for an available channel before transmission. Opportunistic spatial sharing has recently been proposed to solve this spectral inefficiency problem by allowing cognitive radio terminals to transmit simultaneously with other users without causing noticeable interference to the primary users. This is achieved by exploiting design freedom made available by multiple antennas at the cognitive transmitters such that signal emitted by them is steered away from other users. However, this can only be achieved when the accurate channel state information or channel distribution information are available, which is not realistic. In this work, robust opportunistic transmission techniques for single-user/multiuser MIMO and MIMO-OFDM cognitive systems will be investigated to tackle such problem. The proposed technique will outperform previously proposed worst case performance methods in literature in terms of BER by treating the channel mismatch less conservatively. The proposed approach will also aim to overcome the shortcoming which exists in probability constrained optimization approach. Performance of the proposed robust transmission scheme will only achieve suboptimal performance as it assumes all resources, both channel and relay nodes, are of equal importance. This assumption is unrealistic because the quality of the channel and relays undergoes constant variation. Thus, we intend to tackle the robust resource allocation problem for direct transmission only and relay-assisted systems. Our work will investigate the problem of how these resources should be best allocated for optimal transmission under the premise that uncertainty in the channels exists. This problem will include determination of the number of (virtual) channels (spatial, spectral, temporal) and relay nodes, and which (virtual) channels and relays nodes are used. Below is a summary of our proposed work: 1st year: #15; Devise robust opportunistic transmission algorithm for MIMO/MIMO-OFDM cognitive radio networks. – Single-user systems. – Multiuser systems. #15; Implementation of such algorithms in MATLAB. 2nd year: #15; Devise unimode and multimode robust resource allocation algorithms for MIMO/MIMO-OFDM cognitive radio systems. – Direct transmission only systems. – Relay-assisted systems. #3; Single relay selection. #3; Virtual antenna array. #15; Implementation of such algorithms in MATLAB.