201014230 六、發明說明： 【發明所屬之技術領域】 本發明大體上係關於一種傳輸裝置、—種接收器裝置、 一種系統及一種在傳輸系統（例如但不限於無線區域網路） 中實行一傳輸端與複數個其他傳輸端之間的多使用者傳 '輸之方法。 【先前技術】 現今，由例如ΙΕΕΕ 802·Π規範所定義之無線區域網路 ^ (WLAN)幾乎無處不在。可用頻道輸送量之增大為一主要議 題，且研究已經集中於改進在實體層内之調變與編碼。藉 由結合高速信號星座（signal constellati〇n)使用正交分頻多 工（OFDM)，可以獲得至多54 Mbh/s。此巨大效能飛躍（即 使僅在非常有限的範圍内獲得）起因於〇FDM之固有特性， 這對尚位元速率系統尤其具有吸引力。在〇FDM中給定系 統頻寬被分成多個副頻道，也稱為副載波。取代透過一個 •(非常寬）頻道循序發送符號，並行發送多個符號。這導致更 長的符號持續時間，使得可明顯減小符號間干擾的影響， 因此不需要如高成本的等化之額外措施。 802.1 1標準強制要求所有站台或使用者終端機實施一分 散協調功能（DCF)，該功能為一避免衝突之載波感测多重存 取（CSMA/CA)的形式。CSMA/CA為一基於競爭之協定，其 確保所有站台在傳輸前先感測媒體。主要目的係為了避免 站台在同一時間傳輸，導致衝突及相應之重新傳輸。若欲 發送一訊框之一站台感測到媒體上的能量高於一特定臨限 140678.doc 201014230 值（這可能意謂著另一個站台之傳輸），則欲存取之站台在傳 輸該訊框之前將等待直到該媒體為閒置為止。該協定之衝 突避免態樣係關於使用認可（acknowledgement)，接收方站 台站台發送認可至發送方站台，以確認無錯誤地接收。雖 然存取媒體之此處理程序稍微較複雜，但可看作一會議， 在會議上每個人都很有禮貌，且每個人僅在沒有其他人發 言的時候才發言。另外，理解發言者所說内容之參與者點 頭以示同意。 分散協調功能（DCF)因其性質而支援非同步信號傳輸。非 _ 同步發信號之一區別因素為在資料攜帶訊框之間不存在時 序需求。例如，分散協調功能（0(：：17)協定對在任何時間框架 (timeframe)内或隨時及時遞送一系列資料訊框不做任何嘗 試。結果，在每個資料訊框傳輸之間存在一隨機量之延遲。 此同步形式對於網路應用係有效的，例如電子郵件、網頁 瀏覽及企業應用之VPN存取。 在一多輸入多輸出（ΜΙΜΟ)天線系統之使用中，可見進— .步之位元速率增長之一可能性。因此，已經提出一種新穎 ❹ 的媒體存取控制（MAC)協定機制，其支援根據基於ιεεε 802.1 1之標準之無線區域網路中的多使用者傳 輸。該所提出的新穎協定以如下方式延伸單使用者（su) ΜΙΜΟ的分散協調功能（DCF):對於一 MIM〇訊框内的封包 (其為在不同空間串流上同時發送之一組封包），不同的站台 可為目的地站台。 該標準包含傳輸前之一選用之請求發送（RTS)_清除發送 140678.doc 201014230 (CTS)交握。在資料傳輸前之一關聯程序中，各站台間相互 共用關於該等站台硬體效能之資訊。可使用延伸形式之 RTS與CTS控制訊框，交換有關所使用之天線元件的資訊。 延伸式RTS訊框（MIMO-RTS(M-RTS))與延伸式CTS訊框 (MIMO-CTS(M-CTS))可係基於IEEE 802·lla標準RTS與 CTS訊框之結構。為了支援多重天線，Μ-RTS訊框與M-CTS 訊框兩者皆具有一新欄位，例如一位元映射，其中每一位 元代表一個天線。因此，長度為一個位元組的一位元映射 可支援至多八個天線。當然，該位元映射欄位可為更長或 更短，其取決於一給定系統之行動台所支援的天線數量。 在Μ-RTS訊框中，此攔位可稱作提議的天線位元映射 (PAB)，且其可對提議用於後續傳輸之可用天線之選定子集 進行編碼。該訊框的接收器確認在Μ-CTS訊框之經確認天 線位元映射（CAB)欄位内哪些天線應為作用中的。亦廷伸 ACK訊框以支援單串流認可。更具體言之，MIMO-ACK (M-ACK)訊框可以有一個位元組長的位元映射欄位（其被 稱為經認可封包位元映射（APB))，用以個別確認接收到來 自不同串流的每個封包。其含有對每個空間串流的肯定認 可與否定認可。儘管一次傳輸多個封包，其仍然可以立即 認可。該等位元映射的長度（L)可以係任意的。 多使用者（MU)MIMO提高無線網路的頻譜效率，且對於 下一代無線區域網路（WLAN)系統將具有很高價值。在諸如 OFDM之通信系統及單載波區塊傳輸中使用循環首碼作為 防護週期，藉由將含有一頻道之線性捲積轉變成循環捲積 140678.doc 201014230 而簡化寬頻頻道中的頻域等化。然而’一般而言，例如， 在一上行鏈路多使用者傳輸方案中，為使頻域在存在干擾 下接近有效率，干擾使用者必須與所需的使用者同步，亦 即來自多使用者的封包應以小於OFDM的防護週期的時間 差到達接收器。本發明中，將在OFDM防護週期内之多上行 鏈路使用者之時間對準稱為多使用者同步。 然而，歸因於硬體限制（例如時脈精確性）、出於同步目 的之發信號之增加的PHY/MAC附加項（overhead)、及當修 改MAC協定以實現來自多使用者的同步傳輸時所面臨的回 溯相容問題中之至少一者，未來多使用者WLAN系統的 MAC協定將可能不支援多使用者同步。若未滿足多使用者 同步的要求，則一接收器需長時間的域濾波器以抑制干擾 使用者及等化所需的使用者，如於例如2004年《Wirel. Commun. Mob. Comput.》M. Breinholt、H. Jung 與厘· Zoltowski之「用於ΜΙΜΟ OFDM蜂巢式通信中之非同步干 擾抑制的空間-時間對準（Space-time alignment for asynchronous interference suppression in ΜΙΜΟ OFDM cellular communications)」中所描述。WLAN存取點於伺服 多使用者時可能無法負擔此種接收器複雜性。該等WLAN MAC協定不支援且很有可能將不支援多使用者同步。 【發明内容】 本發明之一目的為提供一種多使用者傳輸方案，其不需 要嚴格的同步，同時保持接收器複雜性合理。 此目的是藉由一種如技術方案1之接收器裝置、一種如技 140678.doc 201014230 ##案8之傳輸n裝置’及—種如技術方案⑺之方法予 現。 因此藉由提供—種簡單及高效的收發器方案，允許一 裝置飼服夕個非同步使用者，其中存取點伺服多使用 #m的多使用者同步需求，亦即來自多個使用者的 封包可以例如夫# ^ 大於OFDM的防護週期之時間差到達存取 ,。。更八體3之，建議將接收處理分成多使用者干擾消除 φ 〃單使用者等化。該單使用者等化可進-步分成可被實施 為傳輸濾波器之時域部分等化（頻道縮短），及於接收器側 處實施之-習知頻域等化。這大幅減輕了於作為一接故器 之存取點處的處理負荷’這是因為作為一傳輸器之每個使 用者對等化工作之部分有貢獻。 而且，已提礒一種支援上行鏈路]^11_]^1^1〇傳輸之诞八匸 協定，例如用於基於敝丨丨之紅篇系統。其增加一個 MU-MIMO傳輸建立階段，可於該階段期間在該存取點與多 Φ 4固使用者終端機之間交換一些控制訊框。這允許存取點在 實際多使用者傳輸前具有每個可行使用者的頻道狀態資訊 (fSI)，且也提供一種機制，其使存取點將以上遽波器待徵 单使用者等化或-些傳輸指令回饋給使用者。定義一增強 型MAC訊框，亦即MU_RTS。此增強型mac訊框與普通㈣ 訊框不同’因為增強型MAC訊框具有多接收者mac位址。 此實現-種將識別或位址的清單傳達到其他傳輸端的改進 方式。雖然所提議的增強型MAC訊框具有僅對]^1；裝置有意 義/可理解之特定欄位，但是增強型MAC訊框可以在舊式實 140678.doc 201014230 體層中傳輸且具有可為所有舊式裝置所理解之共同攔位。 因此，舊式裝置可以解碼位元，解譯共同欄位及初始化適 當的設定。增強型MAC訊框之解譯可以係一純粹的^^八^處 理程序，使彳于不需要從實體層得到更多資訊。此外，沒有 必要改變對相應的現有MAC訊框或舊式MAC訊框的解譯 規則。鑑於所有其他傳輸端可至少被所有其他傳輸端部分 解譯之事實，從實體層角度看其傳輸可看作係廣播傳輸。 因此’舊式裝置及程序僅需做很小的修改。 根據一第一態樣，藉由處理來自所有傳輸端之在傳輸請 求（例如Μ-RTS)中接收到的訓練序列，可在存取點（Ap)(亦 即接收器端）評估每個使用者終端機（亦即每個傳輸端）之傳 輸頻道。因此，在可有利地使用一初始建立階段提供之資 訊’而不需要額外的處理負荷。 根據可與該第一態樣結合之一第二態樣，提取一所需的 使用者終端機之信號的接收器處理可分成一多使用者干擾 消除階段及一單使用者等化階段。 根據可與以上第一態樣及第二態樣令之至少—者結合之 一第三態樣，該多使用者干擾消除可藉由從經評估的傳輸 頻道計算出的使用者特定干擾消除濾波器電路予以實施。 根據可與以上第一態樣至第三態樣中之至少—者結合之 一第四態樣，單使用者等化可分成對應於個別特徵之一時 域部分等化及於接收器裝置（例如A p)處實施之一頻域等 化。此分散等化方法導致該接收器側的一減少的處理負荷。 根據可與以上第一態樣至第四態樣中之至少—者結合之 I40678.doc -8- 201014230 -第五態樣’可基於經評估的傳輸頻道及基於多使用者干 擾消除m電4之-特徵決定該個別特冑。因為很容易 得到必需的資訊’所以可實施_順向傳輸濾波器的設计。201014230 VI. INSTRUCTIONS OF THE INVENTION: TECHNICAL FIELD OF THE INVENTION The present invention generally relates to a transmission device, a receiver device, a system, and a transmission in a transmission system such as, but not limited to, a wireless local area network. The method of multi-user transmission between the end and a plurality of other transmission ends. [Prior Art] Today, a wireless local area network (WLAN) defined by, for example, the 802. Π specification is almost ubiquitous. The increase in available channel throughput is a major issue, and research has focused on improving modulation and coding within the physical layer. Up to 54 Mbh/s can be obtained by using Orthogonal Frequency Division Multiplexing (OFDM) in conjunction with a high-speed signal constellation (signal constellati). This huge leap in performance (even within a very limited range) is due to the inherent characteristics of 〇FDM, which is especially attractive for still bit rate systems. The given system bandwidth in 〇FDM is divided into multiple subchannels, also known as subcarriers. Instead of sending symbols sequentially through a • (very wide) channel, multiple symbols are sent in parallel. This results in longer symbol durations, so that the effects of intersymbol interference can be significantly reduced, thus eliminating the need for additional measures such as high cost equalization. The 802.1 1 standard mandates that all stations or user terminals implement a Distributive Coordination Function (DCF), which is a form of carrier-sensing multiple access (CSMA/CA) that avoids collisions. CSMA/CA is a competition-based agreement that ensures that all stations sense the media before transmission. The main purpose is to prevent the stations from transmitting at the same time, resulting in conflicts and corresponding retransmissions. If one of the stations to send a frame senses that the energy on the media is higher than a certain threshold 140678.doc 201014230 (this may mean the transmission of another station), then the station to be accessed is transmitting the message. The box will wait until the media is idle. The conflict avoidance aspect of the agreement is about the use of acknowledgement, and the receiving station station sends an acknowledgement to the sender's station to confirm receipt without error. Although this process of accessing the media is a bit more complicated, it can be seen as a meeting where everyone is polite and everyone speaks only when no one else speaks. In addition, participants who understand what the speaker said are nodded to show consent. The Decentralized Coordination Function (DCF) supports asynchronous signal transmission due to its nature. One of the distinguishing factors of non-synchronous signaling is that there is no timing requirement between the data carrying frames. For example, the decentralized coordination function (0(::17) protocol does not attempt to deliver a series of data frames in any timeframe or at any time. As a result, there is a random between each data frame transmission. This type of synchronization is effective for web applications such as email, web browsing, and enterprise application VPN access. In the use of a multi-input multi-output (ΜΙΜΟ) antenna system, it can be seen. One of the possibilities for bit rate growth. Therefore, a novel media access control (MAC) protocol mechanism has been proposed that supports multi-user transmission in a wireless local area network based on the standard ιεεε 802.1 1 . The proposed novel protocol extends the single user (su) 分散 Decentralized Coordination Function (DCF) by: for a packet within a MIM frame (which is a group of packets simultaneously transmitted on different spatial streams), The station can be the destination station. The standard includes the request to send (RTS) _ clear send 140678.doc 201014230 (CTS) handshake before transmission. Before data transmission In one of the associated procedures, each station shares information about the hardware performance of the stations. The extended form of RTS and CTS control frames can be used to exchange information about the antenna elements used. Extended RTS Frame (MIMO) -RTS (M-RTS) and extended CTS frames (MIMO-CTS (M-CTS)) can be based on the structure of IEEE 802.11a standard RTS and CTS frames. To support multiple antennas, Μ-RTS frame Both the M-CTS frame and the M-CTS frame have a new field, such as a one-bit mapping, where each bit represents an antenna. Therefore, a one-dimensional mapping of one byte in length can support up to eight antennas. Of course, the bit map field can be longer or shorter depending on the number of antennas supported by the mobile station of a given system. In the Μ-RTS frame, this block can be called the proposed antenna bit. Mapping (PAB), and it can encode a selected subset of available antennas proposed for subsequent transmissions. The receiver of the frame is confirmed in the Confirmed Antenna Bit Map (CAB) field of the Μ-CTS frame. Which antennas should be active. Also extend the ACK frame to support single stream approval More specifically, a MIMO-ACK (M-ACK) frame may have a bit-length bit mapping field (referred to as an Approved Packet Bit Map (APB)) for individual acknowledgement receipts from Each packet of a different stream, which contains a positive and negative acknowledgement for each spatial stream. Although multiple packets are transmitted at a time, they can be immediately recognized. The length (L) of the bit map can be arbitrary. Multi-user (MU) MIMO improves the spectral efficiency of wireless networks and will be of great value to next-generation wireless local area network (WLAN) systems. Using a cyclic first code as a guard period in a communication system such as OFDM and single-carrier block transmission, the frequency domain equalization in a wideband channel is simplified by converting a linear convolution containing a channel into a circular convolution 140678.doc 201014230 . However, in general, for example, in an uplink multi-user transmission scheme, in order to make the frequency domain close to efficient in the presence of interference, the interfering user must synchronize with the required user, that is, from multiple users. The packet should arrive at the receiver with a time difference less than the guard period of OFDM. In the present invention, the time alignment of multiple uplink users within the OFDM guard period is referred to as multi-user synchronization. However, due to hardware limitations (such as clock accuracy), increased PHY/MAC overhead for synchronization purposes, and when modifying MAC protocols for simultaneous transmission from multiple users At least one of the backward compatibility issues faced, the MAC protocol of future multi-user WLAN systems may not support multi-user synchronization. If the multi-user synchronization requirement is not met, a receiver needs a long-term domain filter to suppress interference users and users required for equalization, such as, for example, "Wirel. Commun. Mob. Comput." M. Breinholt, H. Jung and PCT Zoltowski, "Space-time alignment for asynchronous interference suppression in OFDM CDMA cellular communications" description. WLAN access points may not be able to afford such receiver complexity when servicing multiple users. These WLAN MAC protocols are not supported and will most likely not support multi-user synchronization. SUMMARY OF THE INVENTION It is an object of the present invention to provide a multi-user transmission scheme that does not require strict synchronization while maintaining a reasonable receiver complexity. This object is achieved by a receiver device as in claim 1, a transmission n device as in the case of 140678.doc 201014230 ##, and a method as in the technical solution (7). Therefore, by providing a simple and efficient transceiver solution, a device is allowed to serve a non-synchronized user, wherein the access point servo uses a multi-user synchronization requirement of #m, that is, from multiple users. The packet may arrive at the access, for example, by a time difference greater than the guard period of OFDM. . In addition, it is recommended to divide the receiving process into multi-user interference cancellation φ 〃 single user equalization. The single-user equalization can be further divided into a time domain partial equalization (channel shortening) that can be implemented as a transmission filter, and a conventional frequency domain equalization implemented at the receiver side. This greatly reduces the processing load at the access point as a connector. This is because each user of the transmitter contributes to the portion of the peering operation. Moreover, a protocol for supporting the transmission of the uplink]^11_]^1^1〇 has been proposed, for example, for a red-based system based on 敝丨丨. It adds a MU-MIMO transmission setup phase during which some control frames can be exchanged between the access point and the multi-feet user terminal. This allows the access point to have channel state information (fSI) for each feasible user before the actual multi-user transmission, and also provides a mechanism for the access point to equalize the above-mentioned chopper to be used by the user or - Some transmission instructions are fed back to the user. Define an enhanced MAC frame, also known as MU_RTS. This enhanced mac frame is different from the normal (four) frame because the enhanced MAC frame has multiple receiver mac addresses. This implementation is an improved way of communicating a list of identities or addresses to other transports. Although the proposed enhanced MAC frame has a specific field that is meaningful/understandable only to the device, the enhanced MAC frame can be transmitted in the old layer 140678.doc 201014230 and has all the old devices. A common block of understanding. Therefore, legacy devices can decode bits, interpret common fields, and initialize appropriate settings. The interpretation of the enhanced MAC frame can be a purely ^^^^ process, so that no more information is needed from the physical layer. In addition, there is no need to change the interpretation rules for the corresponding existing MAC frame or legacy MAC frame. In view of the fact that all other transmissions can be interpreted at least by all other transmissions, their transmission can be considered as a broadcast transmission from the perspective of the physical layer. Therefore, old devices and programs need only minor modifications. According to a first aspect, each use can be evaluated at an access point (Ap) (i.e., at the receiver) by processing training sequences received from transmission requests (e.g., Μ-RTS) from all of the transmission terminals. The transmission channel of the terminal (ie, each transmission end). Therefore, it is advantageous to use the information provided during an initial setup phase without additional processing load. According to a second aspect that can be combined with the first aspect, the receiver processing for extracting a desired signal from the user terminal can be divided into a multi-user interference cancellation phase and a single user equalization phase. According to at least one of the above first aspect and the second aspect, the multi-user interference cancellation can be filtered by user-specific interference cancellation calculated from the evaluated transmission channel. The circuit is implemented. According to a fourth aspect that can be combined with at least one of the above first aspect to the third aspect, the single user equalization can be divided into one time domain portion corresponding to the individual feature and equalized to the receiver device (eg A p) is implemented at one frequency domain equalization. This decentralized equalization method results in a reduced processing load on the receiver side. According to at least one of the above first aspect to the fourth aspect, I40678.doc -8- 201014230 - the fifth aspect can be based on the evaluated transmission channel and based on multi-user interference cancellation m electricity 4 The feature determines the individual characteristics. Since the necessary information is easily obtained, the design of the _ forward transmission filter can be implemented.
根據可與以上第一態樣至第五態樣中之至少一者結合之 一第六態樣，可等決定該個別特徵，使得非零濾波器分接 頭的長度小於多使用者傳輸信號的防護週肖，或使得最小 化一第一組濾波器分接頭與一第二組濾波器分接頭的功率 比率。因此，可設計傳輸濾波器以滿足所需的性質。 隨附申請專利範圍定義了更多的有利的發展結果。 【實施方式】 現在，將參考附圖，基於各種實施例描述本發明。 下文，基於如圖1所示之一種多使用者多輸入多輸出（Mu ΜΙΜΟ)系統描述較佳實施例。 根據圖1，一MUMIMO存取點（ΑΡ)為Ν個站台提供Wlan 存取，該等站台包括各自單使用者傳輸（丁幻鏈1〇1至 10_N、各自使用者特定傳輸濾波器12-1至12-N，及各自天 線。該存取點（AP)包括用於提供不同編碼及/或調變方案的 N個不同單使用者接收（RX)鏈，經由各自使用者 特定濾波器電路22-1至22-N及複數個天線1至队而施加輸 入信號給該等單使用者接收（RX)鏈20-1至20-N。 一般而言，基於傳播頻道性質（亦即於接收器之天線陣列 處之空間相關性矩陣之結構）’兩種類型ΜΙΜΟ技術可用於 該ΑΡ與每個站台間的兩個方向上。在所接收信號之高相關 性之情況下，可應用不同的波束成形演算法，而在所接收 140678.doc 201014230 4號之出現低相關性之情況下’分集（DIV)及多工（Μυχ)方 法可給予更好效能。在Μυχ方案中，同時傳輸多串流，每 串抓使用#用天線。這增加了輸送量，其因數等於被 傳輸之D目。在DIV方案中，以一不同方式使用多個天 線。對於基本的DIV方案，傳輸器僅使用 —個天線。含有多 天線的接收器接收所傳輸信號的多重複本，則更使用一適 當的信號處理演算法獲得明顯更高的信雜比（snr^在結合 DIV與MUX的方案中，1多的傳輸天線係在作用中狀態， 但疋如同在所有DIV方案中，接收器仍然可以有比串流數目 更多的天線。多工技術存在，但是接收器得到的關於所傳 輸信號的資訊多於在純MUX情況下。 藉由以下實施例實施一簡易且高效的實體層收發器方 案，例如用於上行鏈路MU_MIM〇 WLAN系統，其中， 伺服多使用者而無需嚴格的多使用者同步，亦即來自多使 用者的封包可以大於〇FDM的防護週期之時間差到達存取 點。 所提4的收發器方案提供了 —種簡易方法以將接收處理 分成一多使用者干擾消除部分、步驟或階段及一單使用者 等化部分、步驟或階段。可在接收器側由使用者特定濾波 器電路22-1至22-N實施該多使用者干擾消除階段。該單使 用者等化階段可進一步分成可在傳輸器側由以上使用者特 定傳輸濾波器12-1至12-N予以實施之一時域部分等化（例 如頻道縮短）部分、步驟或階段，及可在接收器側由使用者 特定單使用者RX鍵20-1至20-N予以實施之一頻域ofdm等 140678.doc •10- 201014230 化步驟。這大幅減輕了於作為接收器之Ap處的處理負荷， 這是因為作為傳輸器之每個使用者對等化處理之部分有貢 獻0According to a sixth aspect that can be combined with at least one of the above first aspect to the fifth aspect, the individual feature can be determined such that the length of the non-zero filter tap is less than the protection of the multi-user transmission signal. Zhou Xiao, or to minimize the power ratio of a first set of filter taps to a second set of filter taps. Therefore, a transmission filter can be designed to meet the required properties. The scope of the accompanying patent application defines more favorable development results. [Embodiment] The present invention will now be described based on various embodiments with reference to the accompanying drawings. Hereinafter, a preferred embodiment will be described based on a multi-user multiple input multiple output (Mu) system as shown in FIG. According to FIG. 1, a MU MIMO access point (ΑΡ) provides Wlan access for each station, and the stations include respective single-user transmissions (Ding illusion chains 1〇1 to 10_N, respective user-specific transmission filters 12-1) Up to 12-N, and respective antennas. The access point (AP) includes N different single-user receive (RX) chains for providing different coding and/or modulation schemes, via respective user-specific filter circuits 22 -1 to 22-N and a plurality of antennas 1 to the team to apply an input signal to the single user receiving (RX) chains 20-1 to 20-N. Generally, based on the nature of the propagation channel (ie, at the receiver) The structure of the spatial correlation matrix at the antenna array) 'Two types of ΜΙΜΟ technology can be used in both directions between the ΑΡ and each station. Different beams can be applied in the case of high correlation of the received signals Forming the algorithm, and the 'diversity (DIV) and multiplex (Μυχ) methods can give better performance in the case of the low correlation of the received 140678.doc 201014230 No. 4. In the scheme, simultaneous transmission of multiple streams , each string grabs the use of # antenna. This increases the transport The factor is equal to the transmitted D. In the DIV scheme, multiple antennas are used in a different way. For the basic DIV scheme, the transmitter uses only one antenna. The receiver with multiple antennas receives more of the transmitted signals. Repeat this, then use a proper signal processing algorithm to obtain a significantly higher signal-to-noise ratio (snr^ in the combination of DIV and MUX, more than one transmission antenna is in the active state, but as in all DIVs) In the scheme, the receiver can still have more antennas than the number of streams. The multiplex technology exists, but the receiver obtains more information about the transmitted signal than in the case of pure MUX. The implementation of the following embodiment is simple and Efficient physical layer transceiver solution, for example, for uplink MU_MIM〇 WLAN system, wherein the server is multi-user without strict multi-user synchronization, that is, the packet from multiple users can be greater than the protection period of the FDM The time difference arrives at the access point. The proposed transceiver scheme provides an easy way to divide the reception process into a multi-user interference cancellation section, step or step. And a single user equalization portion, step or stage. The multi-user interference cancellation phase can be implemented by the user-specific filter circuits 22-1 to 22-N on the receiver side. The single-user equalization phase can be further Dividing into a time domain equalization (eg, channel shortening) portion, step or stage that can be implemented by the above user-specific transmission filters 12-1 to 12-N on the transmitter side, and can be user-specific on the receiver side The single-user RX keys 20-1 to 20-N are implemented in a frequency domain ofdm, etc. 140678.doc •10-201014230. This greatly reduces the processing load at the Ap as a receiver, because as a transmission Each user of the device contributes to the part of the equalization process.
作為一實施實例，現在參考圖丨考慮一多使用者上行鏈路 WLAN方案。存在N個站台或使用者終端機，且每個使用者 終端機有一個傳輸天線存在具有队個接收天線伺務N個使 用者之AP。從任意傳輸天線到任意接收天線的實體頻道 可模型化為例如含有Z個分接頭之一 FIR濾波器。 為不失一般性，現在更詳細地描述圖丨之上位使用者終端 機1之處理，亦即因此將從使用者終端機〗接收到的信號處 理為所需的信號’及將從其他N_i個使用者終端機接收到的 信號處理為干擾信號。於AP處的接收器使用其渡波器電路 22-1(可為一數位濾波器，例如一有限脈衝響應濾波器 (FIR))以抵銷使用者終端機丨之干擾使用者，該濾波器電路 的特徵為每個接收天線含有尨個分接頭之从⑴。可藉由使用 一矩陣^及一行向量w〇)來定義濾波器電路22-1的濾波操 作’其表示為如下： ^〇)= h{L-\) mAs an example, consider a multi-user uplink WLAN scheme with reference to FIG. There are N stations or user terminals, and each user terminal has one transmission antenna, and there are APs with N receiving users of the receiving antennas. The physical channel from any transmit antenna to any receive antenna can be modeled as, for example, a FIR filter containing one of the Z taps. For the sake of generality, the processing of the upper user terminal 1 is now described in more detail, that is, the signal received from the user terminal is thus processed as the desired signal 'and will be N_i from the other The signal received by the user terminal is processed as an interference signal. The receiver at the AP uses its waver circuit 22-1 (which may be a digital filter, such as a finite impulse response filter (FIR)) to offset the interfering user of the user terminal, the filter circuit It is characterized in that each receiving antenna contains a slave (1). The filtering operation of the filter circuit 22-1 can be defined by using a matrix ^ and a row vector w〇) which is expressed as follows: ^〇) = h{L-\) m
w(0) (1) (L+M -1 )xMNr h(L-l) 其中/2(/)-[4 (/)，.·” \⑺]，’ =〇,…乂_1為一列向量，其含有從 使用者終端機1至％個接收天線之實體頻道之/階分接項， 及W(W) = [«..，'»]X"，M-1為一行向量，其含有在 140678.doc •11 · 201014230 γ接收天線處的使用者終 ^ . 知機1之濾波器電路（也可被指定 為干擾4除濾波器」）的仍階 + ' 知機 了類似地疋義力⑷與祕(《)， 、 、 ’ n=Ll,…，N。 在一多使用者傳輸建立階 白奴期間，接收器藉由處 ==:的訓練序列來評估每個使用者的頻道： 下文之更洋細描述。具右f從士 、 斤有可能使用者之頻道狀n冑 訊（CSI)，即可例如藉由使用 狀〜貝 承解#式（2)之簡易強制歸 零方法計算特徵^ h(l)w(l) = 〇, (2) ~h(2)' 其中=: h(N) 為一干擾頻道矩陣，其含有干擾使用者終端機丨之使用者終 端機之頻道狀態資訊（CSI，s)。濾波器特徵，之解可以係以 上矩陣W之一空值空間向量（nuU space vect〇〇。空值空間 的存在要求P之列數至少比行數大i，其給定从咖，用於抵 銷使用者終端機1的干擾的每個接收天線所需的最小分接 頭數目，即： (3) M -(^-1)(1-1) + 1 minw(0) (1) (L+M -1 )xMNr h(Ll) where /2(/)-[4 (/),.·" \(7)], ' =〇,...乂_1 is a column of vectors , which contains the / order tap of the physical channel from the user terminal 1 to the % receiving antennas, and W(W) = [«.., '»] X", M-1 is a row of vectors containing In the 140678.doc •11 · 201014230 γ receiving antenna at the end of the user's filter circuit (also can be designated as interference 4 filter), the order + ' knows the machine similarly Force (4) and secret ("), , , 'n=Ll,...,N. During a multi-user transmission setup, the receiver evaluates each user's channel by a training sequence of ==:: A more detailed description below. With the right f from the squad, the user may have a channel-like information (CSI), which can be calculated, for example, by using the simple forced zeroing method of the type ~Bei Cheng solution# (2) ^ h(l) w(l) = 〇, (2) ~h(2)' where =: h(N) is an interfering channel matrix containing channel status information (CSI, s) that interferes with the user terminal of the user terminal ). Filter characteristics, the solution can be a null space vector of the above matrix W (nuU space vect〇〇. The existence of the null space requires that the number of columns of P be at least larger than the number of rows, which is given by the coffee, used to offset The minimum number of taps required for each receiving antenna of the user terminal 1 interference, ie: (3) M -(^-1)(1-1) + 1 min
Nr-N + \ 使用干擾頻道矩陣/z(I)及計算得到的濾波器特徵，可設 計該傳輸濾波器12-1之一特徵〆”，使得使用者終端機1的所 有有效頻道具有一些所需的性質’以減輕接收器處的等 化’例如使非空值分接頭之長度小於OFDM之防護週期及/ 140678.doc • 12- 201014230 或使某些分接頭與其他分接頭之功率比率最大化。 在以上實施例中’雖然該濾波器特徵尽(1)係關於一傳輸濾 波器’但疋同樣地可應用用於頻道縮短之一接收濾波技術 (如於例如1996年12月《IEEE Trans. Commim.》第44卷第12 號Ρ· Melsa、R. Younce與c. Rohrs之「用於離散多載頻調收 發器的脈衝響應縮短（ImpUlse reSp〇nse sh〇rtening f〇r discrete multitone transceiver)」中所描述），以計算濾波器 特徵〆”。如從此文獻中可搜集，一頻道縮短濾波器比該目 標脈衝響應短一位元。因此，在實施例中，具有特徵γ之 傳輸濾波器12-1所需的分接頭數目可粗略地選為相當於防 護週期長度。因此於ΑΡ處之接收器基於各自評估頻道矩陣 W及所計算的特徵一計算所有使用者終端機之傳輸濾波 器特徵g() ’《 = 1，··.，#。最後，於ΑΡ處之接收器可以使用一 提出的MU-CTS訊框以回饋決定的傳輸濾波器特徵給每 個使用者終端機。 然後在交換Μ-RTS與MU-CTS訊框後，可開始實際的多 使用者傳輸。 在圖1之方塊圖中，每個使用者終端機可以有一個習知的 單使用者傳輸鏈，其跟隨有在建立階段期間由ΑΡ通知之額 外的使用者特㈣輸渡波器12·1至12-Ν。該ΑΡ首先使用 使用者特定遽波器電路22W作為多使用者干擾消除 遽，器以提取—相關使用者終端機”，及然:後可以應用-習 知=使用者接收鏈處理以解調變一自使用者終端機”接收 的4 \ « l’.’.’TV。對各自單使用者接收鍵之處 140678.doc •13- 201014230 理可由同步、頻域頻道評估、OFDM解調變、頻域等化等組 成’因為對於每個使用者，多使用者干擾已經被一各自含 有特徵’之濾波器電路22-1至22-N之一抵銷，所以使總有 效頻道比防護週期短傳輸濾波器特徵容⑷。 接下來，提出一種MAC協定增強，其支援肘17訄1]^()傳輸 系統例如基於IEEE 802.1 1之WLANS之在MU傳輸系統之上 行鍵路方向之波束成形。其係基於一 mac機制，其中，一 共同接收器（例如圖1之AP)藉由廣播一傳輸（C4T)訊框之呼 叫至候選傳輸器（例如圖1之使用者終端機1至N)而初始化 一傳輸。被定址的候選傳輸器藉由發送M-Rts訊框作出反 應’以表明向共同接收器發送之意願，且用於接收器處之 頻道評估之訓練序列接在Μ-RTS訊框之後。作為一替代， 可在各自前置項中提供訓練序列。接收器評估來自每個傳 輸器的頻道’及根據其等之頻道實現來評定該等候選傳輸 器。該接收器也可以找到每個傳輸器之適合的傳輸波束成 形或濾波器向量’及可對Μ-RTS訊框回覆一 MU-CTS訊框， 其中該MU-CTS訊框可指示哪些傳輸器可藉由使用哪個傳 輸波束成形或濾波器向量存取該頻道。然後可以開始Mu ΜΙΜΟ傳輸。 因此，所提出的新機制為支援傳輸波束成形2MUMIM〇 傳輸提供了 一種上行鏈路頻道存取機制，其中一共同接收 器支援來自不同傳輸器的同步多封包接收。因此可增加該 系統的頻譜效率。該嵌入式傳輸波束成形機制提供多個傳 輸器之間之一良好的MUMIMO傳輸協調，使得空間串流間 140678.doc -14- 201014230 的干擾達到最小。此外，在頻道實現不適合一 MU ΜΙΜΟ傳 輸之情況下，該新機制給予在SU與MU ΜΙΜΟ傳輸模式間切 換之可能性。 更具體地說，在提議之用於上行鏈路方案之MU ΜΙΜΟ MAC機制中，C4T、Μ-RTS與MU-CTS訊框可用於存取一頻 道，及一 MU-ACK訊框可用於認可正確接收的封包。視情 況地，可藉由修改僅一使用者終端機子集之MUMIMO傳輸 之傳輸波束成形向量以建立下一 ΜΙΜΟ訊框而提供一適應 性MUMIMO傳輸，該使用者終端機子集之空間串流可被正 確接收。該決定可基於從例如一接收封包的錯誤校正碼（例 如一循環冗餘碼（CRC))獲致的資訊。 下文，基於圖2顯示之五個步驟更詳細地描述所提出的 MUMIMO上行鏈路MAC程序。 在第一步驟中，一共同接收器（例如圖1之AP)廣播一傳輸 (C4T)訊框之呼叫以在上行鏈路中初始化一 MU ΜΙΜΟ傳 輸。在C4T訊框中，指示所有具備上行鏈路MU ΜΙΜΟ能力 使用者終端機的位址，其可為一可變數量。或者，該ΑΡ可 決定僅輪詢具備上行鏈路MU ΜΙΜΟ能力使用者終端機之 一子集。 該C4T訊框也可攜帶一探測（sounding)請求及待探測之空 間維數之一指示。或者，待探測之空間維數可被標準化為 該AP之頻道評估能力，這可從該AP之HT能力欄位獲得，可 在信標訊框、關聯回應訊框等中傳輸。可設定該持續時間 欄位以涵蓋直至MU ΜΙΜΟ傳輸之起始的傳輸持續時間。如 140678.doc • 15· 201014230 以上所解說，可藉由將來自該使用者終端機的回應持續時 間、該MU-CTS訊框的持續時間及分離該等訊框之 SIFS/RIFS時間間隔相加來獲得此持續時間。因為在建構 C4T訊框時，指派用於MU ΜΙΜΟ傳輸之使用者終端機數量 係未知的，所以MU-CTS訊框的持續時間係未知的。對於 C4T持續時間欄位之計算，假設MU-CTS包含最大欄位數 量，並且使用的MCS相同於用於C4T訊框傳輸的MCS。注 意，在同一訊框内，必須傳達對於每個使用者終端機之CSI 報告之格式。用於傳達此資訊之機制可相同於用於下行鏈 路MU ΜΙΜΟ傳輸中的機制。該持續時間必須將CSI報告也 考慮在内。可定期傳輸該C4T訊框可。循環頻率則可取決於 具備上行鏈路MU ΜΙΜΟ能力ΑΡ之數量。可在信標訊框中傳 達該頻率給其他ΑΡ。 圖3顯示可在實施例中使用的含有多傳輸器位址攔位之 C4T訊框結構的一項實例。 在每個使用者終端機（傳輸器或傳輸端）已經接收到該呼 叫（例如該C4T訊框）後，在程序的第二步驟中，使用者終端 機藉由發送一 Μ-RTS封包或訊框作出回應，以指示其等向 該指示的接收器發送的意願。此處，「Μ-RTS訊框」代表 M-DCF RTS訊框且包含RTS訊框欄位及額外欄位，例如一 CSI欄位。該Μ-RTS訊框可由一 RTS訊框與一 CSI回饋訊框 之彙總代替。Nr-N + \ Using the interference channel matrix /z(I) and the calculated filter characteristics, one of the characteristics of the transmission filter 12-1 can be designed such that all active channels of the user terminal 1 have some The required property 'to mitigate the equalization at the receiver', for example, makes the length of the non-null tap smaller than the guard period of OFDM and / 140678.doc • 12- 201014230 or maximizes the power ratio of some taps to other taps In the above embodiment, 'although the filter feature (1) is related to a transmission filter', the same can be applied to one of the channel shortening receiving filtering techniques (for example, in IEEE Trans, for example, December 1996). Commim., Vol. 44, No. 12 Ρ · Melsa, R. Younce, and c. Rohrs, "Impulse response shortening for discrete multi-carrier tuning transceivers (ImpUlse reSp〇nse sh〇rtening f〇r discrete multitone transceiver (described in )) to calculate the filter characteristics 。". As can be gathered from this document, a channel shortening filter is one bit shorter than the target impulse response. Thus, in an embodiment, transmission filtering with characteristic gamma 12-1 The number of taps required can be roughly selected to be equivalent to the guard period length. Therefore, the receiver at the 计算 calculates the transmission filter characteristics g() of all user terminals based on the respective evaluation channel matrix W and the calculated characteristics. " = 1,··.,#. Finally, the receiver at ΑΡ can use a proposed MU-CTS frame to feed back the determined transmission filter characteristics to each user terminal. Then exchange Μ-RTS After the MU-CTS frame, the actual multi-user transmission can be started. In the block diagram of Figure 1, each user terminal can have a conventional single-user transmission chain that is followed during the setup phase. The additional user-specific (four) wave-transporters 12·1 to 12-Ν are notified by the user. The user first uses the user-specific chopper circuit 22W as a multi-user interference cancellation device to extract the relevant user terminal. ", and then: can be applied - conventional = user receive chain processing to demodulate the 4 \ « l'.'.'TV received from the user terminal. For each individual user to receive the key 140678.doc •13- 201014230 Step, frequency domain channel evaluation, OFDM demodulation, frequency domain equalization, etc. 'Because for each user, multi-user interference has been one of the filter circuits 22-1 to 22-N each containing a feature' Offset, so that the total effective channel is shorter than the guard period transmission filter feature (4). Next, a MAC protocol enhancement is proposed, which supports the transmission system of the elbow 17訄1]^(), for example, based on IEEE 802.1 1 WLANS. Beamforming of the uplink direction of the transmission system. It is based on a mac mechanism in which a common receiver (such as the AP of FIG. 1) broadcasts a C4T frame call to a candidate transmitter (eg, user terminals 1 through N of FIG. 1). Initialize a transfer. The addressed candidate transmitter responds by transmitting an M-Rts frame to indicate the willingness to transmit to the common receiver, and the training sequence for channel evaluation at the receiver is followed by the Μ-RTS frame. As an alternative, a training sequence can be provided in the respective predecessors. The receiver evaluates the channel from each of the transmitters' and evaluates the candidate transmitters based on their channel implementations. The receiver can also find a suitable transmit beamforming or filter vector for each transmitter and can reply to a MU-CTS frame for the Μ-RTS frame, wherein the MU-CTS frame can indicate which transmitters are available. The channel is accessed by which transmit beamforming or filter vector is used. Then you can start the Mu ΜΙΜΟ transfer. Therefore, the proposed new mechanism provides an uplink channel access mechanism for supporting transmit beamforming 2MUMIM(R) transmission, where a common receiver supports simultaneous multi-packet reception from different transmitters. This increases the spectral efficiency of the system. The embedded transmit beamforming mechanism provides good MUMIMO transmission coordination between multiple transmitters, minimizing interference between spatial streams 140678.doc -14- 201014230. In addition, the new mechanism gives the possibility of switching between the SU and MU ΜΙΜΟ transmission modes in case the channel implementation is not suitable for a MU transmission. More specifically, in the proposed MU ΜΙΜΟ MAC mechanism for the uplink scheme, the C4T, Μ-RTS and MU-CTS frames can be used to access a channel, and a MU-ACK frame can be used to recognize the correct one. Received packet. Optionally, an adaptive MU MIMO transmission can be provided by modifying a transmit beamforming vector of the MU MIMO transmission of only one subset of the user terminal to establish a next frame, the spatial stream of the subset of the user terminal Can be received correctly. The decision may be based on information obtained from, for example, an error correction code (e.g., a cyclic redundancy code (CRC)) that receives the packet. In the following, the proposed MU MIMO uplink MAC procedure is described in more detail based on the five steps shown in FIG. In a first step, a common receiver (e.g., the AP of Figure 1) broadcasts a transmission (C4T) frame call to initiate a MU transmission in the uplink. In the C4T frame, the address of all uplink MU ΜΙΜΟ capable user terminals is indicated, which can be a variable amount. Alternatively, the UI may decide to poll only a subset of the uplink MU ΜΙΜΟ capable user terminals. The C4T frame can also carry a sounding request and an indication of the spatial dimension to be detected. Alternatively, the spatial dimension to be detected can be normalized to the channel assessment capability of the AP, which can be obtained from the HT capability field of the AP, and can be transmitted in a beacon frame, an associated response frame, or the like. This duration field can be set to cover the transmission duration up to the beginning of the MU ΜΙΜΟ transmission. As explained above, the response duration from the user terminal, the duration of the MU-CTS frame, and the SIFS/RIFS interval separating the frames can be added. To get this duration. Because the number of user terminals assigned for MU transmission is unknown when constructing a C4T frame, the duration of the MU-CTS frame is unknown. For the calculation of the C4T duration field, it is assumed that the MU-CTS contains the maximum number of columns and the MCS used is the same as the MCS used for C4T frame transmission. Note that the format of the CSI report for each user terminal must be communicated within the same frame. The mechanism used to convey this information can be the same as the mechanism used in downlink MU ΜΙΜΟ transmission. This duration must also take into account the CSI report. The C4T frame can be transmitted periodically. The cycle frequency can then depend on the number of uplink MU capabilities. This frequency can be communicated to other devices in the beacon frame. Figure 3 shows an example of a C4T frame structure containing multiple transmitter address blocks that can be used in an embodiment. After each user terminal (transmitter or transmission end) has received the call (for example, the C4T frame), in the second step of the procedure, the user terminal transmits a Μ-RTS packet or message. The box responds to indicate its willingness to send to the receiver of the indication. Here, the "Μ-RTS frame" represents the M-DCF RTS frame and contains the RTS frame field and additional fields, such as a CSI field. The Μ-RTS frame can be replaced by a summary of an RTS frame and a CSI feedback frame.
藉由在C4T訊框中的清單中的傳輸器順序來隱含地決定 Μ-RTS訊框的順序。在一 SIFS時間間隔後傳輸第一M-RTS 140678.doc •16· 201014230 訊框，在各自RIFS時間間隔後傳輸隨後的訊框。攜帶m_rts 訊框之實體協定資料單元（ppDU)可以係一探測ppdu。持續 時間可以係兩個持續時間的和，其中第一持續時間始於 M-RTS傳輸完成後之一 SIFS時間間隔，直到MUMIMO傳輸 的起始’及該第二持續時間為在用於M_RTS訊框之撾以將 被用於傳輸搁置中資料之情況下的資訊訊框傳輸之持續時 間。從此持續時間欄位，該AP可獲知將由一使用者終端機 發送之資料量，及因此可恰當地設定MU-CTS訊框中的持續 時間欄位。 圖4顯示可在實施例中使用的m_rtS訊框結構。 在該AP(接收器或接收端）從候選使用者終端機接收 Μ-RTS訊框及評估使用者終端機的頻道實現後，於程序的 第三步驟中，對於一可能之MU ΜΙΜΟ傳輸評定使用者終端 機之頻道實現及為每個使用者終端機或空間串流尋找一合 適的傳輸波束成形向量。接著，藉由廣播一 MU-CTS訊框繼 續進行頻道保留’其中，MU-CTS訊框指示哪些傳輪器可藉 由使用哪些傳輸波束成形向量來存取頻道。 圖5顯示可在實施例中使用之含有τχ波束成形向量之增 強型MU-CTS訊框結構的實例。 或者’可使用更泛用的MU-CTS訊框，其不攜帶Τχ波東成 形向量。於是，可使用一彙總操縱訊框（經壓縮或未經壓縮） 來攜帶Τχ波束成形向量。也可在這些訊框（例如在hTc搁位） 中傳達待由受指派之使用者終端機在傳輸中使用的％“。 持續時間欄位可設定為最長的空間串流持續時間加上一 14067E.doc -17- 201014230 SIFS時間間隔與傳輸Μ-ACK訊框所需的時間。 在程序的第四步驟中，使用者終端機可藉由使用於 MU-CTS訊框中指示的Tx波束成形向量來存取頻道。 最後，在第五步驟中，於MU ΜΙΜΟ上行鏈路傳輸完成 後，該ΑΡ可傳輸一MU-ACK訊框，在MU-ACK訊框中認可 由受指派之使用者終端機同時傳輸之對封包的成功接收。 圖6顯示可在實施例中使用之一相應MU-ACK訊框結 構。可在經認可封包位元映射（ΑΡΒ)欄位中傳達此認可，該 經認可封包位元映射（ΑΡΒ)欄位的長度等於MU-CTS訊框令 Φ Tx位址的數目。可（例如）藉由將相對應於傳輸站台之位元 設定為「1」而認可一封包之成功接收。 運用以上機制，該接收器藉由找到適當的傳輸波束成形 向量，且此資訊饋送給傳輸器而初始化及協調上行鏈路中 之MU ΜΙΜΟ傳輸，因此，為MU ΜΙΜΟ上行鏈路傳輸提供 一種高效率頻道存取機制與一干擾避免技術。 下文中，提出一種機制以減小MU-DCF中的附加項。大多 數MU-DCF中的附加項係由於多個Μ-CTS與Μ-ACK訊框回 — 覆連同其等之SIFS時間間隔之先於每一訊框之一前置項而 產生。應用除分時多重存取（TDMA)以外之一多重存取方案 明顯地改良MU-DCF網路的效能。 在ΜΙΜΟ系統中，有可能空間多工該等訊框，但是不可假 設傳輸器處的頻道知識。在OFDM系統（例如IEEE 802.11a) 中，使用OFDM傳輸導致最小的硬體複雜性。然而，其他方 案（例如MC-CDMA或CDMA)可以有一類似的效果。 140678.doc 18 201014230 在OFMDA的情況下，藉由使用例如副載波的四分之一， 諸如Μ-CTS與M-ACK訊框之短封包較長較長四倍，這是因 為封包的大部分係前置項。取决於實體層，Μ-CTS與M-ACK 訊框較長數個符號。假設一大小為1024位元組的封包、用 於54 Mb/s資料封包的一實體層模式、一 36 Mb/s的實體層模 式（及如IEEE 802.1 la標準中之其他相關參數），傳輸窗在SU 模式下的持續時間為33 8微米，在MU模式（TDMA)下的持續 時間為578微米，在MU模式（OFDMA)下的持續時間為362 微米。 因此，提議減小在MU-DCF操作下之MU模式下傳輸 Μ-CTS與M-ACK訊框之所需時間。此使MU下的附加項幾乎 減小到一 SU系統的附加項，同時保留MU-MIMO傳輸的上 述優點。 代替在一 TDMA模式下傳輸Μ-CTS與M-ACK訊框，將所 有的副載波劃分為多個子集及將每個子集指派給一個必須 發送Μ-CTS或M-ACK訊框的使用者終端機。 可根據於MU-RTS訊框中傳達之位址清單中之接收器順 序決定有關副載波子集至使用者終端機之映射之資訊。因 此，可同時傳輸Μ-CTS與]VI-ACK訊框，使得可並行化SIFS 時間間隔與前置項，前置項在IEEE 802.11網路中先於每一 訊框且在MU-MIMO系統中係附加項的载波。取决於實體層 特徵，Μ-CTS與M-ACK訊框可僅以較長數個符號。 圖7顯示根據一實施例之發信號與處理之示意圖，其中一 使用者終端機或站台（STA)lO-i與一 AP20通信。 140678.doc -19· 201014230 在初始建立階段期間，在步驟S101中，該STA ΐο-i傳 輸含有一訓練序歹k — M(U)-RTS至AP 20。在步驟S102 中AP 20基於自所有使用者站台接收到的訓練序列來評估 5亥等使用者頻道，及在步驟S1G3中，基於獲取之頻道資訊， 為 S T A 1 λ · 1之干擾消除瀘、波器（例如圖i之遽波器電路 汁算使用者特定特徵。然後在步驟S1 04中，設計一使用者 特定傳輸錢n(例如圖i之傳輸濾波^ 12_i)，使得獲得— -有所需的質之總有效使用者頻道。在步驟S⑻中將 經設計的傳輸m的特徵回鎖至STA 1(M及在該處可用 於建立-使用者特定傳輸遽波器。最後在步驟S⑽中可 於在STA 1 〇-_ AP 20處得到的濾波器設定而開始 者傳輸。 用 總之，已描述了-種實行多使用者傳輸之傳輸器裝置、 接收器裝置、電腦程式產品及方法，其中，評估自多重傳 輸至| |收側之個別傳輸頻道，且基於該頻道評估 建立多使用者干擾消除。另外，用於單—使用者等化 特徵紅⑧#並由接收器料以用信號發送至所有傳輪側，❹ 二藉此分散傳輸側與接收器側間的接收處理。該頻道評估 w糸土於連同一傳輸請求—起接收的一訓練序列，例如 MU RTS讯框、MU_CFR訊框或以丁訊框，且可在一傳輪認 可中用n發送單使用者等化之特徵，例如mu cts等等。 上述實施例提供了具有此潛在_般概念的各種有利° 步增強與改良。 需注意的是’本發明不限於上述實施例，且可用於任何 140678.doc -20· 201014230 多使用者傳輸方案，不僅*MUMIMC^更具體地說，本發 明適用於所有類型的基於]^1]^0的WLAN，特別係“七⑶系 統。協定在單使用者（su)模式與多使用者（MU)模式下工 作。與Μ-DCF相比較，在存在多重連接的高度互連系統中 及AP下行鏈路中可預期效能改進。再者，本發明適用於所 有多使用者無線系統。預期促進無線網路中的多使用者訊 務，其中該訊務為非同步且以一單接收器為目的地，例如The order of the Μ-RTS frames is implicitly determined by the order of the transmitters in the list in the C4T frame. The first M-RTS 140678.doc •16·201014230 frame is transmitted after a SIFS interval, and subsequent frames are transmitted after the respective RIFS time interval. The physical protocol data unit (ppDU) carrying the m_rts frame can detect ppdu. The duration may be the sum of two durations, where the first duration begins at one of the SIFS intervals after the completion of the M-RTS transmission until the start of the MU MIMO transmission and the second duration is in the M_RTS frame The duration of the information frame transmission in the case of the Laos will be used to transmit the data in the hold. From this duration field, the AP can know the amount of data to be sent by a user terminal, and thus the duration field in the MU-CTS frame can be properly set. Figure 4 shows the m_rtS frame structure that can be used in the embodiment. After the AP (receiver or receiving end) receives the Μ-RTS frame from the candidate user terminal and evaluates the channel implementation of the user terminal, in the third step of the procedure, for a possible MU ΜΙΜΟ transmission evaluation The channel implementation of the terminal and the search for a suitable transmit beamforming vector for each user terminal or spatial stream. Next, channel reservation is continued by broadcasting a MU-CTS frame, wherein the MU-CTS frame indicates which of the transmitters can access the channel by using which transmit beamforming vectors. Fig. 5 shows an example of an enhanced MU-CTS frame structure containing a τχ beamforming vector which can be used in the embodiment. Or 'you can use the more general MU-CTS frame, which does not carry the Τχ波东成形 vector. Thus, a summary manipulating frame (compressed or uncompressed) can be used to carry the chirp beamforming vector. The %" to be used by the assigned user terminal in the transmission can also be communicated in these frames (eg in hTc). The duration field can be set to the longest space stream duration plus a 14067E .doc -17- 201014230 SIFS time interval and time required to transmit the Μ-ACK frame. In the fourth step of the procedure, the user terminal can use the Tx beamforming vector indicated in the MU-CTS frame. To access the channel. Finally, in the fifth step, after the MU ΜΙΜΟ uplink transmission is completed, the ΑΡ can transmit a MU-ACK frame, and the assigned user terminal is approved in the MU-ACK frame. Simultaneous reception of the transmitted packet at the same time. Figure 6 shows a corresponding MU-ACK frame structure that can be used in an embodiment. This recognition can be conveyed in the approved packet bit map (ΑΡΒ) field, the approved packet The length of the bit map (ΑΡΒ) field is equal to the number of MU-CTS frames Φ Tx addresses. The success of a packet can be recognized, for example, by setting the bit corresponding to the transmission station to "1". receive. Using the above mechanism, the receiver initializes and coordinates the MU ΜΙΜΟ transmission in the uplink by finding the appropriate transmit beamforming vector, which feeds the transmitter, thus providing a high efficiency for MU ΜΙΜΟ uplink transmission. Channel access mechanism and an interference avoidance technique. In the following, a mechanism is proposed to reduce the additional items in the MU-DCF. Most of the additional items in the MU-DCF are generated by a plurality of Μ-CTS and Μ-ACK frame backs along with their SIFS time intervals preceding a preamble of each frame. The application of a multiple access scheme other than Time Division Multiple Access (TDMA) significantly improves the performance of the MU-DCF network. In the ΜΙΜΟ system, it is possible to multiplex the frames in space, but it is not possible to assume channel knowledge at the transmitter. In OFDM systems (e.g., IEEE 802.11a), the use of OFDM transmission results in minimal hardware complexity. However, other schemes (such as MC-CDMA or CDMA) can have a similar effect. 140678.doc 18 201014230 In the case of OFMDA, by using, for example, a quarter of the subcarriers, short packets such as Μ-CTS and M-ACK frames are four times longer and longer because of the majority of the packets. The predecessor. Depending on the physical layer, the Μ-CTS and M-ACK frames are longer symbols. Suppose a packet of size 1024 bytes, a physical layer mode for 54 Mb/s data packets, a physical layer mode of 36 Mb/s (and other related parameters such as the IEEE 802.1 la standard), transmission window The duration in SU mode is 33 8 microns, the duration in MU mode (TDMA) is 578 microns, and the duration in MU mode (OFDMA) is 362 microns. Therefore, it is proposed to reduce the time required to transmit the Μ-CTS and M-ACK frames in the MU mode under MU-DCF operation. This allows the additional items under the MU to be reduced to almost an add-on to the SU system while preserving the above advantages of MU-MIMO transmission. Instead of transmitting the Μ-CTS and M-ACK frames in a TDMA mode, all subcarriers are divided into multiple subsets and each subset is assigned to a use that must transmit a Μ-CTS or M-ACK frame. Terminal. Information about the mapping of subcarrier subsets to user terminals can be determined in accordance with the order of the receivers in the address list communicated in the MU-RTS frame. Therefore, the Μ-CTS and ]VI-ACK frames can be transmitted simultaneously, so that the SIFS time interval and the preamble can be parallelized. The preamble is preceded by each frame in the IEEE 802.11 network and in the MU-MIMO system. The carrier of the additional item. Depending on the physical layer characteristics, the Μ-CTS and M-ACK frames can only be a long number of symbols. Figure 7 shows a schematic diagram of signaling and processing in accordance with an embodiment in which a user terminal or station (STA) 10-i communicates with an AP 20. 140678.doc -19· 201014230 During the initial setup phase, in step S101, the STA ΐο-i transmits a training sequence —k — M(U)-RTS to AP 20. In step S102, the AP 20 evaluates the user channel such as 5H based on the training sequence received from all the user stations, and in step S1G3, based on the acquired channel information, cancels the interference of the STA 1 λ · 1 The device (for example, the chopper circuit of Figure i calculates user-specific features. Then in step S1 04, a user-specific transmission money n (for example, transmission filter ^ 12_i of Figure i) is designed so that - The total effective user channel of the quality. The feature of the designed transmission m is locked back to STA 1 in step S(8) (where M can be used to establish a user-specific transmission chopper. Finally in step S(10) The transmitter is transmitted at the STA 1 〇-_ AP 20 and is initially transmitted. In summary, a transmitter device, a receiver device, a computer program product and a method for implementing multi-user transmission have been described, wherein Evaluate individual transmission channels from multiple transmissions to | | and establish multi-user interference cancellation based on the channel evaluation. In addition, for single-user equalization feature red 8# and signaled by the receiver to all The wheel side, ❹ two, thereby disperses the receiving process between the transmitting side and the receiver side. The channel evaluates a training sequence, such as a MU RTS frame, a MU_CFR frame, or a The frame can be used to send a single user equalization feature, such as mu cts, etc. in a pass authorization. The above embodiments provide various advantageous step enhancements and improvements with this potential concept. The present invention is not limited to the above embodiment, and can be applied to any 140678.doc -20· 201014230 multi-user transmission scheme, not only *MUMIMC^ more specifically, the present invention is applicable to all types based on ^^1]^ 0 WLAN, especially "seven (3) system. The agreement works in single-user (su) mode and multi-user (MU) mode. Compared with Μ-DCF, in the highly interconnected system with multiple connections and AP Performance improvements are expected in the downlink. Furthermore, the present invention is applicable to all multi-user wireless systems. It is expected to facilitate multi-user services in wireless networks where the traffic is asynchronous and is a single receiver. Destination, for example
在一上行鏈路方案中的一Ap或提供一外部網路存取之一閘 道0 熟悉此項技術者從對圖式、揭不]^及附加請求項的研 九可理解及實現對所揭示實施例之各種變更。在請求項 中，單詞「包括」不排除其他元件或步驟，及不定冠詞「一」 不排除複數個元件或步驟…單處理器或其他單元可完成 圖1與5及請求項中所述之-些項目的功能。某些方法被敍 逸於互異的附屬請求項中之純粹事實不表示此等方法之組 口不可被有利使用…種用於控制處理器以實行所主張的 =法特徵的電腦程式可儲存/分散於—合適媒體例如一光 學記憶體媒體或與其#麻# 又興再他硬體一起或作為其他硬體之部分之 一固態媒體上’但也可以其 h > 』以具他形式分散，例如經由網際網 路或其他有線或無線的電信系統。 ’、項中的任何參考符號不應該理解為 【圖式簡單說明】 圖1顯示根據一實施例的容蚀田本 ^ 夕使用者MIM〇傳輸系統的示意 万塊圖； 140678.doc -21· 201014230 圖2顯不根據—第五實施例的MU ΜΙΜΟ上行鏈路機制； 圖3顯不一種可用於實施例中之C4T訊框結構； 圖4顯示一種可用於實施例中之m_rts訊框結構； 圖5顯示一種可用於實施例中之含有傳輸（Τχ)波束成形 向量之MU-CTS訊框結構； 圖6顯示一種可用於實施例中之MU-ACK訊框結構；及 圖7顯示根據一實施例之發信號及處理之示意圖。 【主要元件符號說明】 10-1至10-Ν 單使用者傳輸鏈 12-1 至12-Ν 使用者特定傳輸濾波器 20-1至20-Ν 單使用者接收鏈 22-1至22-Ν 使用者特定濾波器電路 1至队 天線 140678.doc 22·One of the uplinks in an uplink scheme or one of the external network access gateways 0 is familiar to the person skilled in the art from the diagram, the disclosure and the additional claims. Various changes to the embodiments are disclosed. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" does not exclude a plurality of elements or steps. A single processor or other unit can be accomplished as described in Figures 1 and 5 and in the claims. The function of some projects. The mere fact that certain methods are invoked in mutually exclusive sub-claims does not imply that the group of such methods is not advantageously used... a computer program for controlling the processor to implement the claimed features can be stored/ Disperse in a suitable medium such as an optical memory medium or on a solid medium with its #麻# yet another piece of hardware or as part of another hardware 'but can also be h > For example via an internet or other wired or wireless telecommunications system. 'Any reference symbol in the item should not be understood as a brief description of the drawing. FIG. 1 shows a schematic diagram of a MIM〇 transmission system of a user-preserving field in accordance with an embodiment; 140678.doc -21· 201014230 FIG. 2 is not based on the MU ΜΙΜΟ uplink mechanism of the fifth embodiment; FIG. 3 shows a C4T frame structure that can be used in the embodiment; FIG. 4 shows an m_rts frame structure that can be used in the embodiment; 5 shows an MU-CTS frame structure including a transmission (Τχ) beamforming vector which can be used in the embodiment; FIG. 6 shows a MU-ACK frame structure usable in the embodiment; and FIG. 7 shows an implementation according to an embodiment. A schematic diagram of signaling and processing. [Main component symbol description] 10-1 to 10-Ν Single user transmission chain 12-1 to 12-Ν User-specific transmission filters 20-1 to 20-Ν Single user reception chain 22-1 to 22-Ν User specific filter circuit 1 to team antenna 140678.doc 22·