1307229 九、發明說明： 【發明所屬之技彳标領域】 發明領域 本發明關於將網際網路協定IP(Internet Protocol)封包 5對映於實體頻道之無線電訊框的IP封包對映方法。 t先*前^^斗椅J 發明背景 弟 1 圖表示對習知 HSDPA(High Speed Downlink Packet Access)中的IP封包之實體頻道的對映方法。 10 於第1圖中’構成發送對象的IP封包，首先分割成 RLC(Radic) Link Control)層之再發送單位，即分割成一定尺 寸的RLC-PDU(Protocol Data Unit)(步驟S1)。又，RLC — PDU之前頭部分可附加包含有為了實現以RLC層再發送之 封包編號等控制資訊的標頭H(RLC標頭）。 15 其次’因應頻道狀態而集中能以1個訊框發送之rlc — PDU，且可產生MAC(Media Access Control) —PDU(步驟 丨 S2)。 其次’於此MAC — PDU之前頭部分可附加包含在mac 層之控制資訊的標頭MAC —H(MAC標頭），同時可附加為 20 了以MAC層進行再發送之誤差檢測碼CRC(Cyclic Redundancy Check)碼（步驟S3)。 接著，經附加MAC標頭及CRC碼之MAC ~~ PDU被進行 用以更正誤差的頻道編碼並構成頻道編碼區塊（Channel Coding Block)(步驟S4)。又，此頻道編碼區塊係在MAc芦 1307229 之再發送單位即ARQ( Am〇matic尺卬邮叫此州區塊。 頻道編碼區塊可對無線電訊框(Radio Frame)對映（步 驟S5)並發送。 又’申請人至本發明申請專利時未發現關聯本發明之 5先行技術文獻。爰此，未揭示先行技術文獻資訊。 c發明内容j 發明概要 發明所欲解決的課題 習知對於HSDPA中的IP封包的實體頻道的對映方法， 10如以上所述進行者，惟，存在有以下的問題點。 由於疋藉由第1 ’ RLC層、MAC層等多數層而對映實體 頻道之無線電訊框者，因此於中途可附加多數的標頭，因 此負擔而存在著無法達到通訊效率化的問題。第2圖表示習 知之對映IP封包之實體頻道之標頭等之位元數的例子，例 15如將正封包設為15〇〇位元，將RLC - PDU設為40位元，則 TP兀包刀割成38個RLC — PDU，而分別可附加16位元的 RLC標頭。又，可附加於MAC — PDU之標頭等如圖式所示 由於如此地可附加多的標頭等，故由於其負擔而無法連到 通訊效率化。 20 第2 ’用以最後性地維持品質之單位雖然為發送斜象即 IP封包，惟’發生資料誤差時用以維持品質之再發送，乃 以頻道編碼區塊(ARQ區塊）為單位的MAC層，以及以 — PDU為單位之上位的RLC層之二層來進行，而不與ip封包 直接關係。例如分割成10個RLC — PDU時，即使是其中9個 1307229 正常地接收，只要1個發生誤差的話，則汗封包整體呈無意 義者’且再發送控制未有效地作用。 第3,由於IP封包為可變長度，因此各種尺寸者為對映 的對象’惟’習知的方法分割成固定尺寸的RLC —pDu， 5並因應頻道狀態而能以1訊框發送的RLC — PDU被集中而 產生MAC-PDU者，而完全未考慮㈣包尺寸，因此係無 法達到充分效率化者。第3圖表示有線系統網際網路上之正 封包尺寸的分布例，第3圖（a)表示相對於汗封包尺寸(橫軸） 之封包數比例(縱軸）’第3圖(b)表示相對於1{>封包尺寸(橫軸) 1〇之封包數累積值（縱軸），具有4〇位元前後之極短尺寸者，以 及1500位元丽後之大尺寸者的二極化傾向（64位元以下約 50%，14GG位兀以上㈣％)。考制今後的無線電通訊 上，IP封包將成為主要流通量，期望能適切地對應如此寬 廣分布尺寸的IP封包。 15 本發明乃鑑於上述習知問題而提案者，其目的在於提 •供IP封包對映方法，而該作封包對映方法可刪減標頭所造 成之負擔而提昇效率，藉„>封包使再送單位設成接近者以 提昇α口貝控制，且忐適切地對應寬廣分布尺寸的IP封包。 解決課題的手段 20 為了解決上述課題’本發明如申請專利範圍第1項所記 載之IP封包對映方去，包含有：產生MAC —PDU步驟，該 步驟係IP封包比預定尺寸小的情形下就設成此尺寸，超過 預疋尺寸的ί月m尤進行分割而產生MAC —;附加步 驟該步“係將包含作封包分割資訊之標頭及碼附加 1307229 於月I]述MAC — PDU ;產生頻道編碼區塊步驟，該步驟係將 已附加前述標頭及CRC碼之MAC — PDU予以頻道編碼而產 生頻道編碼區塊；對映步驟，該步驟係將前述頻道編碼區 塊對映於無線電訊框。 又’如申請專利範圍第2項所記載之IP封包對映方法， 包含有：產生MAC —PDU步驟，該步驟係從Ip封包產生固 定尺寸之⑽或多數MAC —PDUW力口步.驟，該步驟係將包 含IP封包分割資訊之標頭及CRC碼附加於前述mac — PDU ;產生頻道編碼d塊H該步驟係將已附加前述伊 頭及CRC碼之厦-PDU予以頻道一而產生頻道編碼: 塊；對映步驟，該步驟係將前述頻道編碼區塊對映於無線 電訊框。 151307229 IX. Description of the Invention: [Technical Field of the Invention] Field of the Invention The present invention relates to an IP packet mapping method for mapping a Internet Protocol Packet 5 (IP) packet to a radio frame of a physical channel. t first * front ^ ^ bucket chair J Background of the invention 1 shows the mapping method of the physical channel of the IP packet in the conventional HSDPA (High Speed Downlink Packet Access). In Fig. 1, the IP packet constituting the transmission target is first divided into retransmission units of the RLC (Radic) Link Control layer, i.e., RLC-PDUs (Protocol Data Unit) divided into a certain size (step S1). Further, the header part H (RLC header) including the control information such as the packet number retransmitted by the RLC layer may be added to the head portion of the RLC_PDU. 15 Next, the rlc-PDU that can be sent in one frame is concentrated in response to the channel status, and a MAC (Media Access Control)-PDU can be generated (step 丨 S2). Secondly, the 'header' part of the PDU can be appended with the header MAC-H (MAC header) of the control information contained in the mac layer, and can be appended with the error detection code CRC (re-transmitted by the MAC layer). Redundancy Check code (step S3). Next, the MAC ~~ PDU to which the MAC header and the CRC code are attached is subjected to channel coding for correcting the error and constitutes a Channel Coding Block (step S4). Moreover, the channel coding block is a retransmission unit of the MAc A1307229, that is, the ARQ (Am〇matic ruler is called the state block. The channel coding block can be mapped to the radio frame (step S5). And the invention does not find the prior art document related to the present invention at the time of the applicant's patent application. Accordingly, the prior art document information is not disclosed. c SUMMARY OF THE INVENTION SUMMARY OF THE INVENTION The problem to be solved by the invention is known to HSDPA. The mapping method of the physical channel of the IP packet in the process, 10 is performed as described above, but there are the following problems. Since the mediation channel of the first 'RCC layer, MAC layer, etc. The radio framer can add a large number of headers in the middle, so there is a problem that communication efficiency cannot be achieved. Fig. 2 shows the number of bits of the header of the physical channel of the conventional IP packet. For example, if the positive packet is set to 15 bits and the RLC - PDU is set to 40 bits, the TP packet is cut into 38 RLC_PDUs, and the 16-bit RLC header can be added. Also, it can be attached to the MAC - PDU As shown in the figure, the number of headers and the like can be added as shown in the figure. Therefore, it is impossible to connect to the communication efficiency due to the burden. 20 The second unit for maintaining the quality last is the transmission oblique image, that is, IP. The packet is only used to maintain the quality of the retransmission when the data error occurs, and is performed by the MAC layer in the channel coding block (ARQ block) and the second layer of the RLC layer in the upper part of the PDU. It is not directly related to the ip packet. For example, when dividing into 10 RLC-PDUs, even if 9 of them 1307229 are normally received, as long as one error occurs, the sweat packet as a whole is meaningless' and the resend control is not valid. Thirdly, since the IP packet is of variable length, the various sizes of the object are mapped to a fixed-size RLC_pDu, 5 and can be framed according to the channel state. The transmitted RLC - the PDU is concentrated to generate the MAC-PDU, and the (4) packet size is not considered at all, so that it is not sufficient efficiency. Figure 3 shows the distribution of the positive packet size on the cable system Internet, 3rd Figure a) indicates the ratio of the number of packets relative to the size of the sweat pack (horizontal axis) (vertical axis) 'Fig. 3 (b) shows the cumulative value of the number of packets relative to the 1{> packet size (horizontal axis) 1〇 (vertical axis) ), the extremely short-term size of the four-dimensional position before and after, and the large-scale tendency of the large size of 1500-bit MN (about 50% below 64 bits, 14 GG above 四 (four)%). In radio communication, IP packets will become the main throughput, and it is expected that the IP packets of such a wide distribution size can be appropriately adapted. 15 The present invention has been proposed in view of the above-mentioned conventional problems, and its purpose is to provide an IP packet mapping method. The method of packet encapsulation can reduce the burden caused by the header and improve the efficiency, and the packet is made to set the resending unit to be close to enhance the alpha-mouth control, and the IP packet corresponding to the wide distribution size is appropriately adapted. . Means for Solving the Problem In order to solve the above problem, the present invention is directed to the IP packet mapping method described in the first aspect of the patent application, and includes the step of generating a MAC-PDU, which is a case where the IP packet is smaller than a predetermined size. Set to this size, the ί月 m exceeding the pre-size is especially divided to generate the MAC-; the additional step is "This will include the header and code for the packet split information plus 1307229 in the month I] MAC-PDU; Generating a channel coding block step of channel coding the MAC-PDU to which the foregoing header and CRC code is appended to generate a channel coding block; and performing a mapping step of mapping the channel coding block to the radio The IP packet mapping method as described in item 2 of the patent application scope includes the step of generating a MAC-PDU, which is to generate a fixed size (10) or a majority MAC-PDU from the IP packet. In this step, the header and the CRC code including the IP packet splitting information are attached to the mac-PDU; the channel code d block H is generated, and the step is to add the channel of the aforementioned header and CRC code to the channel. Generating channel coding: block; enantiomeric step of the channel coding block based on the wireless telecommunication enantiomers frame 15.
又，如申請專利範圍第3項所記载之珊包對映方法， 包含有：產生MAC-PDU步驟，該步驟係從逆封包原原本 本地產生可變尺寸之MAC-PDU ;附加步驟，該步驟係將 包含ip封包分«訊之標頭及CRC碼附加於前述mac — PDU;產线道編碼區塊步驟，勢縣將已附加前述標 頭及CRC碼之MAC-刚予以頻道一而產生頻道編碼區 塊；對映步驟’該步驟係將前述頻道編碼區塊對映於無線 電訊框 又，如申請專利範圍第4項所記裁，於申請專利範圍第 1項所記載之㈣包對映方法中，其中前述預定尺寸可因應 接收品質及/或要求再發送紐頻“適合地設定。 又，如申請專利範圍第5項所記栽，於申請專利範圍第 20 1307229 1項所記載之封包對映方法中，其中前述IP封包超過預定尺 寸時之分割，係從前頭以前述預定尺寸部分分割前述IP封 包。 又，如申請專利範圍第6項所記載，於申請專利範圍第 5 1項所記載之封包對映方法中，前述IP封包超過預定尺寸時 之分割，係可將前述IP封包等分割成預定個數以成為前述 預定尺寸以下。 又，如申請專利範圍第7項所記載，於申請專利範圍第 1項所記載之封包對映方法中，其中包含有接收步驟，該步 10 驟係從接收側接收表示前述頻道編碼區塊之誤差程度之可 靠度；再發送步驟，該步驟係可靠度低時再發送多的位元， 可靠度高時僅再發送一部分位元。 又，如申請專利範圍第8項所記載，於申請專利範圍第 7項所記載之封包對映方法中，其中前述可靠度係可因應可 15 靠度而藉區分位階之NACK信號而接收者。 又，如申請專利範圍第9項所記載，於申請專利範圍第 1 1項所記載之封包對映方法中，其中包含有接收步驟，該步 驟係從接收側接收表示前述頻道編碼區塊之每一位元或每 一區塊之誤差程度的可靠度；再發送步驟，該步驟係因應 20 前述可靠度而僅再發送判斷為可靠度低而發生誤差的位元 或區塊。 又，如申請專利範圍第10項所記載，於申請專利範圍 第1項所記載之封包對映方法中，其中原原本本或已分割之 IP封包尺寸不滿足事前決定之一定值時或尺寸產生餘裕 1307229 時可進行與接著的IP封包連接、反覆編碼所構成之速率 匹配或插入虛擬位元。 又，如申請專利範圍第〖丨項所記載，於申請專利範圍 第1項所記載之封包對映方法中，#中依據前述CRC碼而來 5自接收歉要求再發送錢，可伴隨其他控㈣號且伴隨 已將ACK/ NACK信號予以CRC編碼之其他crc碼。 、又，如申請專利範圍第12項所記載之發送器，具有： 分割連接部，該分割連接部係進行Ip封包之分割及/或與 接著的IP封包的連接而產生Mac — pd^ ;封包編碼部， /封包編碼部係將包含Ip封包分g資訊之標頭及CRC碼附 加方、Θ述MAC-PDU者；頻道編碼部，該頻道編碼部係將 已附加則述標頭及CRC碼之M Ac — pDu予以編碼並產生頻 C、.扁碼區塊者，對映部，該對映部係將前述頻道編碼區塊 對映於無線電訊框者。 15發明效果 Φ 本黍明之1P封包對映方法，以使無RLC層之標頭而能 降低負擔並提昇效率。又，再發送單位躲封包而成為接 近者的狀態而能提昇品質控制的性能。而且，因應ιρ封包 尺寸並以若干圖案進行分割、連接的狀態，而能以良好效 率來傳送寬廣分布尺寸的IP封包。 圖式簡單說明 第1圖表示對習 映方法。 知HSDPA中的IP封包之實體頻道的對 第2圖表不對習知Ip封包之實體頻道對映中的標頭等 1307229 之位元數的例子。 第3圖（a)、（b)表示有線系統網際網路上之IP封包尺寸 之分布的例子。 第4圖（a)〜(c)表示本發明之IP封包對映方法的流程圖。 5 第5圖表示本發明之IP封包對映方法之圖案的分類。 第6圖表示無分割而為可變長度時之IP封包與MAC — PDU之關係的例子。 第7圖表示有分割且可變長度，等分割成一定長度以下 時之IP封包與MAC — PDU之關係的例子。 10 第8圖表示有分割且可變長度，以固定長度等分割（僅 一個可變長度）時之IP封包與MAC — PDU之關係的例子。 第9圖表示有分割且固定長度時之IP封包與MAC — PDU之關係的例子。 第10圖表示應用本發明之IP封包對映方法之通訊系統 15 的構成例。 第11圖（a)、（b)表示從IP封包經由MAC — PDU而產生頻 I 道編碼區塊的例子。 第12圖係適應調變解調頻道編碼(A M C)的概念圖。 第13圖表示資料調變方式與頻道編碼率之組合例。 20 第14圖（a)、（b)表示從頻道編碼區塊對映無線電訊框的 例子。 第15圖表示混合ARQ的處理。 第16圖（a)、（b)表示依據本發明之實施樣態所採用之 C R C碼所構成之要求再發送信號誤差檢測型混合A R Q的說 11 1307229 明圖。 第17圖（a)、（b)表示習知一般性的混合ARQ發生之要求 再發送信號誤差的例子。 第18圖表示習知R L C層所構成之要求再發送信號誤差 5 之補償的例子。 第19圖表示應用本發明之IP封包對映方法之通訊系統 的其他構成例。 第20圖⑻〜(c)表示NACK信號的例子。 第21圖表示合併整體可靠度表示之ACK/NACK信號 10 的例子。 第22圖表示區塊的例子。 第23圖⑻〜（c)表示混合ARQ之處理的類型。 第24圖表示解碼器的構成例。 第2 5圖表示使用軟判定輸出並檢測可靠度的例子。 15 第26圖表示使用在反覆解碼過程之碼反轉次數而檢測 可靠度的例子。 第27圖表示使用反覆解碼中的交叉熵而檢測可靠度時 之算出處理的例子。 第28圖表示使用接收SINR而檢測可靠度的例子。 20 【實施方式】 較佳實施例之詳細說明 以下說明本發明之最佳實施樣態。 第4圖表示本發明之IP封包對映方法的流程圖。 第4圖（a)表示不進行IP封包之分割而產生可變長度之 12 1307229 MAC — PDU的類型’且由以下步驟構成，即，從ip封包原 原本本地產生MAC —PDU的步驟(步驟S11)、將包含IP封包 分割資訊之標頭（包含在MAC層之控制資訊)及CRC碼（用以 進行以M A C層再發送的誤差檢測碼）附加於已產生之M A c 5 — PDU的步驟（步驟S12) '用以將已附加標頭及crc碼之 MAC — PDU予以誤差更正而頻道編碼化並產生頻道編碼區 塊的步驟(步驟S13)、將頻道編碼區塊對映於無線電訊框的 步驟（步驟S14)。又’在MAC — PDU之產生步驟（步驟S11) 中’ IP封包尺寸未滿足事前決定之一定值時(不獲得頻道編 10碼增益程度之尺寸時），有進行與不進行與接著的IP封包連 接（限於有連接發送緩衝器之IP封包時）、反覆編碼所構成之 速率匹配或插入虛擬位元情形。又，此等連接等處理可因 應無線電鍵接收態而進行控制。 第4圖（b)表示因應ip封包尺寸而進行分割並產生可變 15長度之MAC — pDU的類型，由以下步驟所構成，即，產生 MAC —PDU步驟，該步驟係IP封包比預定尺寸小的情形下 就没成此尺寸’超過預定尺寸的情形下就進行分割而產生 MAC —PDU者（步驟S21)、附加步驟’該步驟係將包含正封 包分割資訊之標頭及CRC碼附加於前述M AC — PDU者（步 20驟幻2)、產生頻道編碼區塊步驟，該步驟係將已附加前述 標頭及CRC碼之MAC — PDU予以頻道編碼而產生頻道編碼 區塊者（步驟S23)、對映步驟，該步驟係將前述頻道編碼區 塊對映於無線電訊框者（步驟S24)。又，在MAC —PDU之產 生步驟（步驟S21)中’原原本本或已分割之IP封包尺寸未滿 13 1307229 、事別决疋之-疋值時’有進行與不進行與接著的ιρ封包 、反復、，扁碼所構成之速率匹配或插人虛擬位元情形。 又’ ^等連接等處理可因應無線電鏈接收態而進行控制。 ,第4圖⑷表不經常產生固定尺寸之MAC — pDU的類 土 /由以下步驟所構成，即，產生隐d即步驟，該步 驟係產生固定尺寸之！個或多數嫩c—者（步驟如）、 附加步驟，該步驟係將包含Ip封包分割資訊之標頭及咖 碼附加於前述MAC-PDU者(步驟⑽、產生頻道編碼區塊 步驟，該步驟係將已附加前述標頭及CRC碼之Mac —pDu 10予以頻道編碼而產生頻道編碼區塊者（步驟S33)、對映步 驟’該步義將前述頻道編碼區塊對映於無線電訊框者(步 驟S34)。又，在MAC —PDU之產生步驟（步驟S31)中，已產 生之MAC-：PDU的尺寸比事前決定之一定值小日寺，有進行 與不進行與接著的IW包連接、反覆編瑪所構成之速率匹 15配或插入虛擬位元情形。又，此等連接等處理可因應無線 電鍵接收態而進行控制。 鲁 第5圖表示本發明之1p封包對映方法之圖案的分類，P1 〜P7相當於第4圖（a)的方法’ P8〜P2〇相當於第4圖（b)的方 法’ P8、P21〜P25相當於第4圖（c)的方法。又，p8〜p2〇更 20可區分為等分割成一定長度以下的情形P8〜P13、及以固定 長度分割的情形P8、P9、P14〜P20。 第6圖至第9圖表示上述圖案中代表性的者之正封包與 MAC — PDU之關係的例子。 第6圖表示無分割而為可變長度時(對應第4圖^)及第$ 14 1307229 圖之P1〜P7的方法）之IP封包與MAC— PDU之關係的例 子，且表示MAC — PDU之尺寸對IP封包之尺寸連動的情 形。又，IP封包之尺寸不滿足事前決定之一定值時，如暗 影(斜線部）所示，乃合併表示進行與接著的IP封包連接、反 5 覆編碼所構成之速率匹配或插入虛擬位元情形。 依據此方法，具有負擔最小的優點。又，再發送、頻 道編碼單位成為用以保證最終品質之jp封包，因此品質控 制變得簡單。相對於此’ IP封包達15〇〇位元那般非常大的 情形下，再發送單位會變得非常大，因此會有惡化再發送 10 效率的可能性。又’ MAC — PDU尺寸會大幅變動。又，如 將於後段所記載，以改良再發送之結構而能解決再發送所 造成效率惡化的問題。 第7圖表示有分割且為可變長度等分割成一定長度以 下時（對應第4圖（b)及第5圖之P8〜P13的方法）之IP封包與 15 MAC —PDU之關係的例子，且表示直至閾值尺寸STH，MAC 一PDU之尺寸對ip封包之尺寸連動，但是，超過閾值尺寸 STH時，等分割成一定長度以下的情形。又，ιρ封包之尺寸 不滿足事前決定之一定值時，如暗影所示，乃合併表示進 行與接著的IP封包連接、反覆編碼所構成之速率匹配或插 20 入虛擬位元情形。 依據此方法’乃不存在非常大尺寸的MAC —PDU，且 MAC — PDU尺寸變動變小，而負擔變大若干。 第8圖表示有分割而為可變長度且以固定長度分割（僅 1個可變長度）時(對應第4圖（b)及第5圖之P8、p9、pl4〜p2〇 15 1307229 的方法）之IP封包與MAC —PDU之關係的例子，且表示直至 閾值尺寸STH ’ MAC_PDU之尺寸對IP封包之尺寸連動，但 是，超過閾值尺寸STH時，從前頭側以固定長度分割，剩餘 部分成為可變長度的情形。又，原原本本或已分割之吓封 5包之尺寸不滿足事前決定之一定值時，如暗影所示，乃合 併表示進行與接著的IP封包連接、反覆編碼所構成之速率 匹配或插入虛擬位元情形。 依據此方法’乃不存在非常大尺寸的MAC — PDU，且 MAC —PDU尺寸變動變小，而負擔變大若干，而且可構成 10 小尺寸的MAC —PDU。 第9圖表示有分割且固定長度時(對應第4圖（c)及第5圖 之P8、P21〜P25的方法）之IP封包與mac — PDU之關係的例 子’且經常產生固定尺寸之MAC —PDU的情形。又，所產 生之MAC —PDU尺寸小於事前決定之一定值時，如暗影所Moreover, the method for generating a hill packet as described in claim 3 includes: a step of generating a MAC-PDU, the step of locally generating a variable size MAC-PDU from the reverse packet; an additional step, the step The system will include the header of the ip packet and the CRC code attached to the mac-PDU; the step of the production line coding block, the county will have the MAC of the header and the CRC code attached to it. Encoding block; mapping step 'This step is to map the above-mentioned channel coding block to the radio frame, and as shown in the fourth paragraph of the patent application scope, the (four) package is recorded in the first item of the patent application scope. In the method, the predetermined size may be appropriately set according to the receiving quality and/or the requirement to transmit the new frequency. In addition, as described in the fifth application of the patent application, the packet described in the patent application No. 20 1307229 In the mapping method, in the case where the IP packet exceeds a predetermined size, the IP packet is partially divided by the predetermined size from the front. Further, as described in claim 6 of the patent application, in the scope of claim 5 In the packet mapping method described above, when the IP packet exceeds a predetermined size, the IP packet or the like may be divided into a predetermined number to be equal to or less than the predetermined size. Further, as described in the seventh aspect of the patent application, The packet mapping method described in claim 1 includes a receiving step, and the step 10 receives a reliability indicating a degree of error of the channel coding block from the receiving side; and a sending step, the step When the reliability is low, more bits are transmitted, and when the reliability is high, only a part of the bits are transmitted. Further, as described in Item 8 of the patent application, in the packet mapping method described in Item 7 of the patent application scope In the above-mentioned reliability, the NACK signal can be received by the NACK signal according to the degree of the reliance. Further, as described in the ninth application of the patent application, the packet mapping method described in Item 1 of the patent application scope is described. Included therein is a receiving step of receiving, from the receiving side, a reliability indicating a degree of error of each bit or each block of the channel coding block; In this step, only the bit or block which is judged to be low in reliability and which causes an error is transmitted in response to the aforementioned reliability of 20. Further, as described in item 10 of the patent application, it is described in item 1 of the patent application scope. In the packet mapping method, when the original IP address of the original or the divided IP packet does not satisfy the certain value determined beforehand or the size of the size is 1307229, the rate matching with the subsequent IP packet connection and the reverse coding may be matched or the dummy bit inserted. In addition, as stated in the Scope of the Patent Application, in the packet mapping method described in item 1 of the patent application scope, ## according to the aforementioned CRC code, 5 accepts the request and then sends the money, which may be accompanied. Other control (4) numbers are accompanied by other crc codes that have CRC encoded the ACK/NACK signal. Further, the transmitter according to claim 12, further comprising: a split connection unit that performs division of the Ip packet and/or connection with the subsequent IP packet to generate a Mac-pd^; The coding unit/packet coding unit includes a header including a Ip packet and a CRC code adder, a MAC-PDU, and a channel coding unit, and the channel coding unit adds a header and a CRC code. The M Ac — pDu encodes and generates a frequency C, a flat code block, and an mapping unit that maps the aforementioned channel coding block to the radio frame. 15 Effects of the Invention Φ The 1P packet mapping method of the present invention can reduce the burden and improve efficiency without the header of the RLC layer. Further, the unit is again sent to hide the packet and becomes a state close to the player, thereby improving the performance of the quality control. Further, in accordance with the size of the packet and the division and connection in a plurality of patterns, the IP packet of a wide distribution size can be transmitted with good efficiency. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows the pairing method. Knowing the physical channel of the IP packet in HSDPA The second chart is not an example of the number of bits in the physical channel mapping of the conventional Ip packet, etc. 1307229. Figures 3(a) and (b) show examples of the distribution of IP packet sizes on the cable system internet. 4(a) to (c) are flowcharts showing the IP packet mapping method of the present invention. 5 Figure 5 shows the classification of the pattern of the IP packet mapping method of the present invention. Fig. 6 shows an example of the relationship between the IP packet and the MAC_PDU when the variable length is divided. Fig. 7 shows an example of the relationship between the IP packet and the MAC_PDU when the partition is divided into variable lengths and divided into a certain length or less. 10 Fig. 8 shows an example of the relationship between an IP packet and a MAC-PDU when divided and variable lengths are divided by a fixed length or the like (only one variable length). Fig. 9 shows an example of the relationship between the IP packet and the MAC-PDU when there is a divided and fixed length. Fig. 10 is a view showing an example of the configuration of a communication system 15 to which the IP packet mapping method of the present invention is applied. Fig. 11 (a) and (b) show an example in which a frequency channel coding block is generated from an IP packet via a MAC_PDU. Figure 12 is a conceptual diagram of adaptive modulation demodulation channel coding (A M C). Fig. 13 shows an example of a combination of a data modulation method and a channel coding rate. 20 Figure 14 (a), (b) show an example of mapping a radio frame from a channel coding block. Figure 15 shows the processing of hybrid ARQ. Fig. 16 (a) and (b) are diagrams showing the retransmission signal error detection type mixed A R Q which is constituted by the C R C code used in the embodiment of the present invention. Fig. 17 (a) and (b) show an example of a conventional general hybrid ARQ generation request for retransmission of a signal error. Fig. 18 shows an example of the compensation of the retransmission signal error 5 which is constituted by the conventional R L C layer. Fig. 19 is a view showing another configuration example of the communication system to which the IP packet mapping method of the present invention is applied. Fig. 20 (8) to (c) show examples of NACK signals. Fig. 21 shows an example of combining the ACK/NACK signal 10 of the overall reliability representation. Figure 22 shows an example of a block. Fig. 23 (8) to (c) show the types of processing of the hybrid ARQ. Fig. 24 shows an example of the configuration of the decoder. Fig. 25 shows an example in which the soft decision output is used and the reliability is detected. 15 Fig. 26 shows an example of detecting the reliability using the number of code inversions in the repeated decoding process. Fig. 27 is a diagram showing an example of calculation processing when the reliability is detected using the cross entropy in the repeated decoding. Fig. 28 shows an example of detecting reliability using the received SINR. [Embodiment] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described. Figure 4 is a flow chart showing the IP packet mapping method of the present invention. Fig. 4(a) shows a step of generating a variable length of 12 1307229 MAC - PDU type without dividing the IP packet and consisting of the steps of locally generating a MAC_PDU from the ip packet (step S11). And the step of attaching the header of the IP packet splitting information (the control information included in the MAC layer) and the CRC code (for performing the error detection code retransmitted by the MAC layer) to the generated MA c 5 — PDU (step S12) 'Step of correcting the error and correcting the channel and encoding the MAC-PDU of the attached header and the crc code to generate a channel coding block (step S13), and mapping the channel coding block to the radio frame (Step S14). Further, in the step of generating the MAC-PDU (step S11), when the IP packet size does not satisfy the predetermined value determined beforehand (when the size of the channel code 10 code gain is not obtained), there is an IP packet that is performed and not followed. Connection (limited to when there is an IP packet connected to the transmit buffer), rate matching by repeated encoding, or insertion of a dummy bit. Also, such connections and the like can be controlled in response to the radio key reception state. Figure 4(b) shows the type of MAC-pDU that is split according to the size of the ip packet and produces a variable length of 15, which consists of the following steps: a MAC-PDU step is generated, which is an IP packet smaller than a predetermined size. In the case where the size is not exceeded by the predetermined size, the MAC-PDU is generated (step S21), and the additional step 'this step is to add the header and CRC code including the positive packet split information to the foregoing. M AC — PDU 者 (Step 20 Fantasy 2), generating a channel coding block step, which is to channel-code the MAC-PDU with the aforementioned header and CRC code to generate a channel coding block (step S23) And mapping step, the step of mapping the aforementioned channel coding block to the radio frame (step S24). Further, in the MAC-PDU generation step (step S21), the original IP address of the original or divided IP packet size is less than 13 1307229, and the value of the 疋 疋 ' ' 有 有 有 有 有 有 有 有 有 有 有 有, the rate formed by the flat code matches or inserts a virtual bit. Further, the processing such as connection can be controlled in response to the reception state of the radio chain. Figure 4 (4) does not often produce a fixed-size MAC - pDU class / consists of the following steps, that is, the process of creating a hidden d, which produces a fixed size! Or a majority of the steps (steps), an additional step of attaching a header and a coffee code containing the Ip packet split information to the MAC-PDU (step (10), generating a channel coding block step, the step The channel-coded block is generated by channel-encoding the Mac-pDu 10 to which the foregoing header and CRC code is attached (step S33), and the mapping step is performed to map the channel coding block to the radio frame. (Step S34) Further, in the MAC-PDU generation step (Step S31), the size of the generated MAC-: PDU is smaller than the value determined beforehand, and the IW packet connection with and without subsequent connection is performed. In addition, the processing of such connections and the like can be controlled in response to the radio key receiving state. Lu 5 shows the pattern of the 1p packet mapping method of the present invention. The classification, P1 to P7 corresponds to the method of Fig. 4(a) 'P8 to P2〇 corresponds to the method of Fig. 4(b)' P8, P21 to P25 are equivalent to the method of Fig. 4(c). P8~p2〇20 can be divided into equal divisions into a certain length or less. P8 to P13 and cases P8, P9, and P14 to P20 divided by a fixed length. Fig. 6 to Fig. 9 show examples of the relationship between the positive packet and the MAC_PDU of a representative one of the above patterns. An example of the relationship between an IP packet and a MAC-PDU when there is no partition and is variable length (corresponding to FIG. 4) and the method of P1 to P7 of the $14 1307229 figure, and indicates the size of the MAC-PDU to the IP packet. The case of the size of the linkage. Moreover, when the size of the IP packet does not satisfy the predetermined value determined in advance, as shown by the shadow (hatched portion), the combination indicates that the rate matching with the subsequent IP packet connection and the inverse 5-layer coding is matched or the dummy bit is inserted. . According to this method, there is an advantage of having the least burden. Further, the retransmission and channel coding units become jp packets for guaranteeing the final quality, so quality control is simplified. In contrast to the case where the IP packet is as large as 15 bits, the resending unit becomes very large, so there is a possibility that the efficiency of resending 10 will be deteriorated. Also, the MAC-PDU size will vary greatly. Further, as will be described in the following paragraph, the problem of deterioration in efficiency caused by retransmission can be solved by improving the structure of retransmission. Fig. 7 is a view showing an example of the relationship between the IP packet and the 15 MAC-PDU when the variable length or the like is divided into a certain length or less (corresponding to the methods of Figs. 4(b) and 5P8 to P13). Further, it is shown that up to the threshold size STH, the size of the MAC-PDU is linked to the size of the ip packet, but when the threshold size STH is exceeded, it is divided into a certain length or less. Further, when the size of the ιρ packet does not satisfy the predetermined value determined in advance, as indicated by the shadow, it is a combination of the rate matching with the subsequent IP packet connection and the reverse coding or the insertion of the dummy bit. According to this method, there is no very large-sized MAC-PDU, and the MAC-PDU size variation becomes small, and the burden becomes large. Fig. 8 shows a method of dividing into a variable length and dividing by a fixed length (only one variable length) (corresponding to P8, p9, pl4 to p2 〇 15 1307229 of Figs. 4(b) and 5) An example of the relationship between the IP packet and the MAC_PDU, and indicates that the size of the threshold size STH 'MAC_PDU is linked to the size of the IP packet. However, when the threshold size STH is exceeded, the length is divided from the front side by a fixed length, and the remaining portion becomes available. The case of variable length. Moreover, when the size of the original or divided scared 5 packs does not satisfy the certain value determined in advance, as indicated by the shadow, the merge indicates that the rate matching with the subsequent IP packet connection and the reverse coding is matched or the dummy bit is inserted. situation. According to this method, there is no very large-sized MAC-PDU, and the MAC-PDU size variation becomes small, and the burden becomes large, and 10 small-sized MAC-PDUs can be constructed. Figure 9 shows an example of the relationship between IP packets and mac-PDUs when there is a split and fixed length (corresponding to the methods of P8 and P21 to P25 in Figs. 4(c) and 5) and often generates a fixed size MAC. - The case of the PDU. Moreover, when the generated MAC-PDU size is smaller than a certain value determined beforehand, such as a shadow
15示’乃合併表示進行與接著的ip封包連接、反覆編碼所構 成之速率匹配或插入虛擬位元情形D 依據此方法，則MAC —PDU為一定尺寸以下，因此容 易對無線電訊框的匹配。相對於此，負擔變大。 其次，第10圖表示應用本發明之JP封包對映方法之通 20 訊糸統的構成例。 於第10圖中，發送器具有：暫時性地儲存發送對象之 IP封包的發送緩衝器11;擷取已儲存在該發送緩衝器丨丨之正 封包，並進行IP封包之分割及/或與接著的Ip封包的連接 而產生MAC —PDU的分割連接部】2 ;將包含Ip封包分割資 16 1307229 訊之標頭及CRC碼附加於已產生之MAC _ pDU的封包編碼 部13 ;將已附加標頭及CRC碼之MAC一pDU予以編碼並產 生頻道編碼區塊的頻道編碼部14 ;將頻道編碼區塊對映於 無線電訊框的對映部15 ;將已匹配之無線電訊框調變成無 5線電信號的調變部16。χ，發送器1〇更具有：因應從接收 器2〇發送而來的要求再發送信號而進行再發送控制的再發 送控制部17、因應從接收H2G發送而來之接收品質資訊及 要求再發送h號(頻度）而控制分割連接部丨2之分割尺寸及 有無連接的控制部18。控制部18於接收品質差的情形下或 要求再發送信號之頻度高的情形下，將分割尺寸設得小並 控制在不進行連接的方向的狀態，相對於此，於接收品質 良好的情形下或要求再發送信號之頻度低的情形下，將分 割尺寸設得大並控制在進行連接的方向的狀態，而能進行 更有效率地發送。 15 ^方面’接收㈣具有：將無線電信號解調的解調 部21、從經解調之信號來測量接收品質並發送至發送㈣ 側的接收品質測量部2 2、從經解調之信號來判斷有無再發 送並將要求再發送㈣發駐發送㈣的再發送控制部 23、從經解調之信號來再㈣封包的Ip封包再生部… 〇第11圖表示於第1〇圖之分割連接部12、封包編碼部 13、頻道編碼部14 ’從IP封包經由MAC —咖而產生頻遒 編碼區塊的例子。第表示㈣包大的情形，正封包 在分割成多數MAC-削的情形下被附加標頭及crc碼而 被設成頻道編碼區塊。第U_)表示Ip封包小的情形，多 17 1307229 數IP封包集中成1個MAC — PDU的情形下被附加標頭及 CRC碼而被設成頻道編碼區塊的狀態。 又’ MAC — PDU之尺寸除了連動於吓封包之尺寸的情 形或構成固定尺寸的情形之外，以控制部18(第1〇圖）所進行 5之MAC 一 PDU尺寸的控制上，能因應通訊狀態而使用可適 切地控制資料速率與資料誤差率之關係的適應調變解調頻 道編碼（AMC : Adaptive Modulation and channel Coding)。 此情形下’ MAC —PDU尺寸因應頻道狀態而變動，並改變 能以1訊框發送的封包數。第12圖係適應調變解調頻道編碼 10 (AMC)的概念圖，一旦設成從基地台1〇〇以相同的發送電力 進行發送者，則存在於基地台100近旁之使用者#1之終端 201之接收電力大且頻道狀態良好，因此選擇高資料速率的 調變方式（例如 16QAM(Quadrature Amplitude Modulation)) 及頻道編碼率（大）。又，存在於遠方之使用者#2之終端202 15 之接收電力小且頻道狀態差，因此選擇低資料速率而資料 誤差率低的調變方式（例如QPSK(Quadrature Phase Shift 1 Keying))及頻道編碼率（小）。第13圖表示資料調變方式與頻 道編碼率之組合例，隨著箭頭方向（朝下）而增大資料速率， 惟，資料誤差率增大。因此，頻道狀態愈好則適合應用位 2〇 於下側之資料調變方式與頻道編碼率的組合，頻道狀態愈 差則適合應用位於上側之資料調變方式與頻道編碼率的組 合。實際上，預先準備構成通道狀態之指標的SIR(Signalto Interference power Ratio)等與授與對應資料調變方式及頻 道編碼率之表單，並因應經測量之頻道狀態而參照表單， 18 1307229 而以切換至符合的資料調變方式及頻道編碼率的狀熊來實 現 AMC。 第Μ圖表示於第10圖之對映部15中’從頻道編碼區塊 對映無線電訊框的例子。第14圖（a)表示從多數頻道編碼區 5塊對無線電訊框進行對映的情形，第14圖（b)表示從丨個頻道 編碼區塊對多數無線電訊框進行對映的情形。又，無線電 訊框分為發送單位即塊(Chunk)，塊以頻率上的（〇FDM .15 indicates that the combination indicates that the rate matching with the subsequent ip packet connection and the reverse coding is matched or the virtual bit is inserted. According to this method, the MAC_PDU is below a certain size, so that the matching of the radio frame is facilitated. In contrast, the burden is increased. Next, Fig. 10 shows an example of the configuration of the JP packet mapping method to which the present invention is applied. In FIG. 10, the transmitter has a transmission buffer 11 for temporarily storing an IP packet of the transmission target, extracting a positive packet stored in the transmission buffer, and performing IP packet division and/or The subsequent Ip packet is connected to generate a MAC-PDU split connection unit 2; the header and CRC code including the Ip packet split 16 1307229 are appended to the generated packet coding unit 13 of the MAC_pDU; The MAC-pDU of the header and CRC code is encoded and the channel coding section 14 of the channel coding block is generated; the channel coding block is mapped to the mapping part 15 of the radio frame; the matched radio frame is converted to none. The modulation unit 16 of the 5-wire electrical signal. In other words, the transmitter 1 further includes a retransmission control unit 17 that performs retransmission control in response to a request for retransmission of a signal transmitted from the receiver 2, and reception quality information and retransmission in response to reception from the H2G reception. The h number (frequency) controls the division size of the divided connection unit 丨2 and the presence or absence of the connection control unit 18. When the reception quality is poor or when the frequency of requesting retransmission of the signal is high, the control unit 18 sets the division size to be small and controls the state in which the connection is not performed. In contrast, in the case where the reception quality is good. In the case where the frequency of resending the signal is required to be low, the division size is set large and the state in which the connection is made is controlled, and transmission can be performed more efficiently. 15 ^ Aspect 'receiving (4) having: a demodulation unit 21 that demodulates a radio signal, and a reception quality measurement unit that measures reception quality from the demodulated signal and transmits it to the transmission (four) side. 2. From the demodulated signal It is determined whether there is any retransmission and requests to resend (4) the retransmission control unit 23 of the transmission (4), and the Ip packet reproduction unit that is further (4) packetized from the demodulated signal. 〇 FIG. 11 shows the split connection in the first diagram. The part 12, the packet coding unit 13, and the channel coding unit 14' generate an example of a frequency coding block from the IP packet via the MAC. The first representation (4) is a case where the packet is large, and the positive packet is set as a channel coding block by adding a header and a crc code in the case of division into a plurality of MAC-cuts. The U_) indicates a state in which the Ip packet is small, and a plurality of 17 1307229 IP packets are grouped into one MAC-PDU, and the header and the CRC code are added to be set as the channel coding block. In addition, the size of the 'MAC-PDU' can be controlled by the control unit 18 (Fig. 1), except for the case of the size of the scary packet or the fixed size. The state uses an adaptive modulation and channel coding (AMC) that can appropriately control the relationship between the data rate and the data error rate. In this case, the 'MAC_PDU size' varies depending on the channel status and changes the number of packets that can be sent in the 1-frame. Fig. 12 is a conceptual diagram of the adaptive modulation and demodulation channel coding 10 (AMC). Once it is set to transmit from the base station 1 with the same transmission power, the user #1 exists in the vicinity of the base station 100. Since the reception power of the terminal 201 is large and the channel state is good, a modulation method of a high data rate (for example, 16QAM (Quadrature Amplitude Modulation)) and a channel coding rate (large) are selected. Moreover, the terminal 202 15 of the remote user #2 has a small received power and a poor channel state, so a modulation method with a low data rate and a low data error rate (for example, QPSK (Quadrature Phase Shift 1 Keying)) and a channel are selected. The coding rate (small). Fig. 13 shows an example of a combination of the data modulation method and the channel coding rate. As the direction of the arrow (downward) increases the data rate, the data error rate increases. Therefore, the better the channel state is, the better the combination of the data modulation mode and the channel coding rate of the lower side is applied, and the worse the channel state is suitable for the combination of the data modulation mode and the channel coding rate located on the upper side. In fact, a SIR (Signalto Interference power Ratio) or the like that constitutes an indicator of the channel state is prepared in advance, and a form for imparting a corresponding data modulation method and a channel coding rate is applied, and the form is referred to in accordance with the measured channel state, 18 1307229 The AMC is implemented to meet the data modulation method and the channel coding rate. The figure is shown in the example of the mapping of the radio channel from the channel coding block in the mapping section 15 of Fig. 10. Fig. 14(a) shows a case where a radio frame is mapped from a plurality of channel coding areas, and Fig. 14(b) shows a case where a plurality of radio frames are mapped from one channel coding block. Also, the radio frame is divided into a transmission unit, that is, a block (Chunk), and the block is on a frequency (〇FDM.
Orthogonal Frequency Division Multiplexing 等）、石馬上的 (CDMA : Code Division Multiple Access 等）、時間上的 10 (TDMA : Time Division Multiple Access 等）、空間上的 (ΜΙΜΟ : Multiple Input Multiple Output)或此等性質的組合 來區分。 其次，第15圖〜18圖係說明依據第1〇圖之發送器1〇之 再發送控制部17及接收器20之再發送控制部23所進行之再 15 發送控制的說明圖。 第15圖表示混合ARQ之處理，在發送器10側進行授與 CRC位元（步驟S101)及誤差更正編碼（步驟si〇2)，在接收器 20側於誤差更正解碼（步驟S201)後進行使用CRC位元的誤 差檢測（步驟S202)’有誤差的情形下對發送器10側進行再發 20送要求，無誤差的話結束發送(接收)(步驟S203)。 又，第16圖表示依據本發明之實施樣態所採用之CRC 碼所構成之要求再發送信號誤差檢測型混合ARQ的說明 圖’如第16圖⑻所示’從接收器20側來的要求再發送信號 與其他控制符號（在適應調變解調頻道編碼上所必要的接 19 1307229 收SIR等）一同伴隨已將ack/ NACK信號予以CRC編碼之 其他CRC碼(CRC~bits)的狀態，降低ACK/NACK信號本 身的誤差。如此一來，如第16圖（1))所示將從發送器1〇側發 送封# k的結果’在接收器20側檢測誤差並將NACK信號發 5 送至發送器10側的情形下，或是此N A C K信號本身誤差而被 當作ACK信號並被接收的情形下，可藉CRC碼而辨識原來 為NACK信號，並適切地再發送已發生誤差的封包#k。 第17圖係為了與本實施樣態作比較而表示習知一般性 的混合ARQ發生之要求再發送信號誤差的例子，如第17圖 1〇 (a)所示將原來的ACL信號辦識為NACK信號時，會再發送 原來被正常地接收的封包#k而降低了功效。又，如第17圖 (b)所示將原來的NACL信號辦識為ACK信號時，會將未正 常地接收的封&#k作成正常地接收，因此會產生封包的缺 漏而惡化了接收品質。而且，習知技術如第18圖所示，以 藉上位的RLC層而進行再發送未正常到達之封包的狀態來 應對，惟，相對於MAC層之再發送以l〇msec程度的時間來 進行，RLC之再發送則需要i〇〇msec程度的時間，故無法匆、 视延遲時間。對於此點問題，以採用依據CRC碼所構成之 要求再發送信號誤差檢測型混合ARQ的本實施樣態，則能 2〇正確地辨識ACK/NACK，即使不使用RLC層亦能進行高品 質的發送。 如上所述’本實施樣態以無RLC層之標頭的狀態而能 降低負擔並提昇效率。又，再發送單位藉吓封包而構成接 近者的狀態能提幵品質控制的性能。而且，.以因應ip封勺 1307229 之尺寸並以若干圖案進行分割及連接的狀態，能適切地對 應尺寸範圍分布廣的ip封包。 其次，說明解決不分割IP封包而進行對映時（對應第4 圖（a)、第5圖之P1〜P7、第6圖的方法）因再發送所造成效率 5 惡化問題的手法。又，此手法於不分割IP封包而進行對映 時特別有效，亦能應用於其他圖案。 即，一般的混合ARQ於檢測出誤差時以將NACK信號 返回至發送側的狀態而以頻道編碼區塊(ARQ區塊）的單位 進行再發送，於不分割IP封包而作為頻道編碼區塊時，若 10 是IP封包長的情形下會因再發送而降低效率。而於包含誤 差的情形下，有整體誤差情形的話，亦存在著僅一部分誤 差的情形，而總是整體再發送的狀態不具效率。因此，以 將表示整體誤差之程度的可靠度或每一位元/每一區塊之 可靠度附加於NACK信號的狀態下，採用執行必要最小限度 15 之再發送的手法。 第19圖表示應用本發明之IP封包對映方法之通訊系統 1 的其他構成例。第19圖中的構成與第10圖所示的構成約相 同，而於接收器20中，設置從解調部21之解調過程來檢測 每一位元的可靠度，並因應需要而將此等位元予以平均化 20 並產生表示每一區塊或整體誤差程度的可靠度，對於再發 送控制部23附加此等位元的資訊並送回NACK信號的可靠 度檢測部25之點，以及於發送器10被附加依據再發送控制 部17被附加N A C K信號的可靠度而進行必要最小限度之再 發送的功能之點不同。 21 1307229 第20圖表示NACK信號的例子，第20圖（a)表示NACK 信號之辨識資訊且表示可靠度之資訊，而該可靠度表示 ARQ區塊整體的誤差程度，第20圖（b)表示NACK信號之辨 識資訊且表示可靠度之資訊，而該可靠度表示每一位元的 5 誤差程度，第20圖⑷表示NACK信號之辨識資訊且表示可 靠度之資訊，而該可靠度表示每一區塊的誤差程度。又， 第20圖（b)、（c)的情形係能將可靠度資訊作為可推定可靠度 低而發生誤差位置的指定資訊。 第21圖表示對應第20圖之合併整體可靠度表示之ACK 10 /NACK信號的例子，表示將無誤差的ACK信號設為1種類 的ACK(O)，將表示有誤差的ACK信號設為高可靠度之 NACK(O)至NACK(6)之7階段的例子。又，可考慮控制效率 等情形而任意地決定將NACK信號設成幾個階段。 又，於第20圖（b)、（c)，表示每一位元可靠度之圖（a) 15 中回授的資訊量變多，因此表示每一位元可靠度之圖（b)有 利。第22圖表示區塊的例子，且表示1個封包分成4個區塊 1#〜4#的例子。 接收了伴隨上述可靠度資訊之N A C K信號之發送器10 的再發送控制部17於整體可靠度資訊的情形下，在可靠度 20 低時再發送多的位元，在可靠度高時僅發送一部分位元。 又，於混合ARQ進行與以不同圖案進行了打孔之封包的封 包合成時，藉變更要追加之冗餘度大小的情形就能對應。 第23圖表示混合ARQ處理的類型，具有如第23圖（a)所示， 於封包pl有解調誤差時廢棄封包pl，而接受再發送同内容 22 1307229 之封包p2並再次進行解調處理的類型，以及如第23圖（b)及 (c)所示，於封包pi有解調誤差時不廢棄該封包pi而先保持 著，乃將此封包pi與接受了再發送之封包p2予以封包合成 而產生封包p3，並對此封包p3進行解調處理的類型。第23 5 圖（b)為接受再發送相同封包者，乃藉封包合成而改善接收 SIR的類型。又，第23圖（c)為接受再發送已進行在不同圖案 打孔之封包者，乃藉封包合成而改善編碼增益的類型。上 述變更要追加之冗餘度大小的情形為第23圖（c)的類型。 又，接收了伴隨上述可靠度資訊之NACK信號之發送 10 器10的再發送控制部17於每一位元或每一區塊可靠度資訊 的情形下，僅再發送判斷為可靠度低而發生誤差之位元或 區塊。 第24圖〜第28圖表示從解調過程產生可靠度的手法。 第24圖表示進行反覆解瑪之解碼器的構成例，y〇〜y2 15 為經多工分離的接收信號(解碼器之輸入信號），且y〇為僅由 信號系列之資訊位元構成的信號，y!為信號系列經回歸性 地組織疊入編碼的信號，y2為信號系列經交錯並再經回歸 性地組織豐入編碼的信號。信號y 〇及y 1藉軟輸入軟輸出解碼 器DEC1而解碼，其輸出信號與信號y0藉交錯器I而被交錯 20 者，以及信號y2藉軟輸入軟輸出解碼器DEC2而解碼，其輸 出信號藉去交錯器DI而被去交錯者，加上軟輸入軟輸出解 碼器DEC1的輸入信號。軟輸入軟輸出解碼器DEC1、DEC2 之輸出信號藉辨識器Dl、D2而被硬判定而成為輸出信號。 第25圖表示使用解碼器之軟判定輸出並檢測可靠度的 23 1307229 例子，能採用將包含於封包之每一位元的軟判定輸出（似然） 作為可靠度的指標。即，第24圖所示之反覆解碼器係具有 連接多數軟輸入軟輸出解碼器DEC1、DEC2的構成，且於 最後段以將軟輸出予以硬判定而再生信號系列者，且係構 5 成接近硬判定之閾值位準（0位準），絕對值小的el、e2誤差 的或然率高而可靠度低者。 第26圖表示使用在反覆解碼過程之碼反轉次數而檢測 可靠度的例子，包含於封包之每一位元的軟判定輸出在反 覆過程中，可採用將碼反轉之次數作為可靠度之指標。即， 10 以反覆數i與i+Ι碼反轉中的el、e2，其解碼結果不穩定且 誤差的或然率高，因此可靠度低。 第27圖表示使用反覆解碼中的交叉摘（cross entropy)而 檢測可靠度時之算出處理的例子，能使用第24圖中所示的 信號值來計算。此情形下，交叉摘TCE愈大則解瑪結果不穩 15 定且誤差的或然率高，因此可靠度低。 第28圖表示使用接收SINR而檢測可靠度的例子，包含 於封包之每一接收SINR愈小則誤差的或然率高，因此可靠 度低。 又，亦可將封包内最低的對數似然比（LLR : Log 20 Likelihood Ratio)作為可靠度指標來使用。 以上藉本發明之最佳實施樣態說明了本發明。於此展 示特定的具體例說明了本發明，惟，只要是不脫離申請專 利範圍所定義之本發明廣泛的宗旨及範圍，當然可對此等 具體例加以各種修正及變更。即，不應解釋為以具體例之 24 1307229 詳細内容及附加的圖式來限定本發明。 c圖式簡單說明3 第1圖表示對習知HSDPA中的IP封包之實體頻道的對 映方法。 5 第2圖表示對習知IP封包之實體頻道對映中的標頭等 之位元數的例子。 第3圖（a)、（b)表示有線系統網際網路上之IP封包尺寸 之分布的例子。 第4圖（a)〜(c)表示本發明之IP封包對映方法的流程圖。 10 第5圖表示本發明之IP封包對映方法之圖案的分類。 第6圖表示無分割而為可變長度時之IP封包與MAC — PDU之關係的例子。 第7圖表示有分割且可變長度，等分割成一定長度以下 時之IP封包與MAC —PDU之關係的例子。 15 第8圖表示有分割且可變長度，以固定長度等分割（僅 一個可變長度）時之IP封包與MAC — PDU之關係的例子。 第9圖表示有分割且固定長度時之IP封包與MAC — PDU之關係的例子。 第10圖表示應用本發明之IP封包對映方法之通訊系統 20 的構成例。 第11圖（a)、（b)表示從IP封包經由MAC —PDU而產生頻 道編碼區塊的例子。 第12圖係適應調變解調頻道編碼(A M C)的概念圖。 第13圖表示資料調變方式與頻道編碼率之組合例。 25 1307229 第14圖⑻、（b)表示從頻道編碼區塊對映無線電訊框的 例子。 第15圖表示混合ARQ的處理。 第關⑻、⑻表示依據本發明之實施樣態所採用之 5 CRC碼所構叙要求再發送㈣㉟差檢測型混合arq的說 明圖。 第17圖⑻、⑼表示習知—般性的混合ARQ發生之要求 再發送信號誤差的例子。 第I8圖表示習知RLC層所構成之要求再發送信號誤差 10 之補償的例子。 第19圖表示應用本發明之IP封包對映方法之通訊系統 的其他構成例。 苐20圖（a)7 (c)表不NACK信號的例子。 第21圖表示合併整體可靠度表示之ACK/NACK信號 I5的例子。 第22圖表示區塊的例子。 第23圖⑻〜（c)表示混合ARQ之處理的類型。 第24圖表示解碼器的構成例。 第25圖表示使用軟判定輸出並檢測可靠度的例子。 20 第26圖表示使用在反覆解碼過程之碼反轉次數而檢測 可靠度的例子。 第27圖表示使用反覆解碼中的交叉熵而檢測可靠度時 之算出處理的例子。 第28圖表示使用接收SINR而檢測可靠度的例子。 26 1307229 【主要元件符號說明】 ίο…發送器 11…發送緩衝器 12…分割連接部 13…封包編碼部 14…頻道編碼部 15…對映部 16…調變部 17…再發送控制部 18…控制部 20…接收器 21…解調部 22…接收品質測量部 23…再發送控制部 24…IP封包再生部 25…可靠度檢測部 100…基地台 201…終端 202…終端 S1〜S5…步驟 S11〜S14…步驟 S21〜S24…步驟 S31〜S34…步驟 P1〜P25…封包 27Orthogonal Frequency Division Multiplexing, etc., Shimadzu (CDMA: Code Division Multiple Access, etc.), time 10 (TDMA: Time Division Multiple Access, etc.), spatial (ΜΙΜΟ: Multiple Input Multiple Output), or the like Combine to distinguish. Next, Fig. 15 to Fig. 18 are explanatory views for explaining the retransmission control by the retransmission control unit 17 and the retransmission control unit 23 of the receiver 20 in accordance with the transmitter 1 of Fig. 1 . Fig. 15 shows the processing of the hybrid ARQ, the CRC bit is transmitted on the transmitter 10 side (step S101) and the error correction coding (step si 〇 2), and the error correction decoding (step S201) is performed on the receiver 20 side. Error detection using the CRC bit (step S202) 'When there is an error, a request for retransmission 20 is sent to the transmitter 10 side, and if there is no error, transmission (reception) is ended (step S203). Further, Fig. 16 is a diagram showing the retransmission signal error detection type hybrid ARQ composed of the CRC code used in the embodiment of the present invention, as shown in Fig. 16 (8), from the receiver 20 side. The retransmission signal is accompanied by other control symbols (such as the SIR required to accommodate the modulation and demodulation channel coding, 19 1307229, SIR, etc.) along with the other CRC codes (CRC~bits) in which the ack/NACK signal has been CRC encoded. Reduce the error of the ACK/NACK signal itself. As a result, as shown in FIG. 16 (1), the result of transmitting the seal #k from the transmitter 1 side is detected in the case where the error is detected on the receiver 20 side and the NACK signal is sent to the transmitter 10 side. Or, in the case where the NACK signal itself is regarded as an ACK signal and received, the original NACK signal can be identified by the CRC code, and the packet #k in which the error has occurred is appropriately retransmitted. Figure 17 is a diagram showing an example of a conventional general hybrid ARQ requesting retransmission signal error in comparison with the present embodiment, as shown in Fig. 17 (a) of Fig. 17 to identify the original ACL signal as When the NACK signal is transmitted, the packet #k that was originally received normally is transmitted, and the effect is lowered. Further, when the original NACL signal is recognized as an ACK signal as shown in FIG. 17(b), the normally-received &#k is normally received, and thus the packet is missing and the reception is deteriorated. quality. Further, as shown in FIG. 18, the conventional technique is to deal with the state in which the packet that has not arrived normally is retransmitted by the upper RLC layer, but the retransmission with respect to the MAC layer is performed for a period of about 10 msec. The retransmission of the RLC requires a time of i〇〇msec, so it is impossible to rush and view the delay time. For this problem, the present embodiment can re-transmit the signal error detection type hybrid ARQ according to the requirement of the CRC code, so that ACK/NACK can be correctly recognized, and high quality can be performed even without using the RLC layer. send. As described above, the present embodiment can reduce the burden and improve the efficiency in the state without the header of the RLC layer. Moreover, the state in which the retransmission unit borrows the packet to form a nearby person can improve the performance of quality control. Further, in the state in which the ip sealing spoon 1307229 is divided and connected in a plurality of patterns, the ip packet having a wide range of size can be appropriately matched. Next, a description will be given of a method for solving the problem that the efficiency 5 is deteriorated due to retransmission when the mapping is performed without dividing the IP packet (corresponding to the methods in Figs. 4(a) and 5D to P1 to P7 and Fig. 6). Moreover, this method is particularly effective when it is mapped without dividing the IP packet, and can be applied to other patterns. That is, when the error is detected, the general hybrid ARQ retransmits in the channel coding block (ARQ block) by returning the NACK signal to the transmission side, and does not divide the IP packet as the channel coding block. If 10 is a long IP packet, it will reduce efficiency due to resending. In the case of an error, if there is an overall error situation, there is also a case of only a part of the error, and the state of the overall retransmission is not efficient. Therefore, in a state in which the reliability indicating the degree of the overall error or the reliability of each bit/block is added to the NACK signal, the method of performing the retransmission of the necessary minimum 15 is employed. Fig. 19 is a view showing another configuration example of the communication system 1 to which the IP packet mapping method of the present invention is applied. The configuration in Fig. 19 is about the same as the configuration shown in Fig. 10, and in the receiver 20, the demodulation process from the demodulation unit 21 is set to detect the reliability of each bit, and this is required as needed. The equalizers are averaged 20 and generate a reliability indicating the degree of error of each block or the whole, and the point at which the retransmission control unit 23 adds the information of the bits and returns the reliability detecting unit 25 of the NACK signal, and The transmitter 10 is different in the function of performing the minimum necessary retransmission according to the reliability of the retransmission control unit 17 to which the NACK signal is added. 21 1307229 Fig. 20 shows an example of a NACK signal, and Fig. 20(a) shows identification information of a NACK signal and indicates reliability information, and the reliability indicates the degree of error of the entire ARQ block, and Fig. 20(b) shows The identification information of the NACK signal and the information indicating the reliability, and the reliability indicates the degree of error of each bit, and the figure 20 (4) indicates the identification information of the NACK signal and indicates the reliability information, and the reliability indicates each The degree of error in the block. Further, in the case of (b) and (c) of Fig. 20, the reliability information can be used as the designation information indicating that the error position is low and the error position is generated. Fig. 21 is a view showing an example of the ACK 10 /NACK signal corresponding to the combined overall reliability shown in Fig. 20, showing that the error-free ACK signal is of one type of ACK (0), and that the ACK signal indicating error is set high. An example of 7 stages of reliability NACK(O) to NACK(6). Further, the NACK signal can be arbitrarily determined in several stages in consideration of the control efficiency and the like. Further, in Fig. 20 (b) and (c), the amount of information to be fed back in the graph (a) 15 of each bit reliability is increased, and therefore the graph (b) indicating the reliability of each bit is advantageous. Fig. 22 shows an example of a block, and shows an example in which one packet is divided into four blocks 1# to 4#. In the case of the overall reliability information, the retransmission control unit 17 of the transmitter 10 that has received the NACK signal of the reliability information transmits a plurality of bits when the reliability 20 is low, and only transmits a part when the reliability is high. Bit. Further, when the hybrid ARQ performs packet synthesis of packets punched with different patterns, it is possible to cope with the case where the degree of redundancy to be added is changed. Figure 23 shows the type of hybrid ARQ processing, as shown in Fig. 23(a), discarding the packet pl when the packet pl has a demodulation error, and accepting the retransmission of the packet p2 of the same content 22 1307229 and performing demodulation again. The type, and as shown in Fig. 23 (b) and (c), when the packet pi has a demodulation error, the packet pi is not discarded and is held first, and the packet pi and the packet p2 that has received the retransmission are given. The packet is synthesized to generate a packet p3, and the packet p3 is demodulated. Figure 23 (b) shows that the same packet is accepted and retransmitted, and the type of receiving SIR is improved by packet synthesis. Further, Fig. 23(c) shows a type in which the packet that has been punctured in a different pattern is accepted for retransmission, and the coding gain is improved by packet synthesis. The case where the degree of redundancy to be added is changed as described above is the type of Fig. 23(c). Further, in the case where the retransmission control unit 17 of the transmission 10 device 10 that has received the NACK signal of the reliability information is in each bit or each block reliability information, only the retransmission is judged to be low in reliability. The bit or block of error. Figures 24 to 28 show the method of generating reliability from the demodulation process. Fig. 24 is a view showing an example of a configuration of a decoder for performing inverse decoding, where y 〇 y y 2 15 is a multiplexed received signal (an input signal of a decoder), and y 〇 is composed only of information bits of a signal series. The signal, y!, is a repetitively organized signal that is superimposed and encoded by the signal series, and y2 is a signal that is interleaved and regressively encoded by the signal series. The signals y 〇 and y 1 are decoded by the soft input soft output decoder DEC1, the output signal and the signal y0 are interleaved by the interleaver I, and the signal y2 is decoded by the soft input soft output decoder DEC2, and the output signal thereof The deinterlacer is borrowed by the interleaver DI, and the input signal of the soft input soft output decoder DEC1 is added. The output signals of the soft input soft output decoders DEC1 and DEC2 are hard-determined by the identifiers D1 and D2 to become output signals. Fig. 25 shows an example of using the soft decision output of the decoder and detecting the reliability. In the example of 1 1307229, the soft decision output (likelihood) included in each bit of the packet can be used as an index of reliability. That is, the flip decoder shown in Fig. 24 has a configuration in which a plurality of soft input soft output decoders DEC1 and DEC2 are connected, and in the last stage, the soft output is hard-determined to reproduce the signal series, and the system is close to 5 The threshold level of the hard decision (0 level), the absolute value of the el, e2 error is high or the reliability is low. Figure 26 shows an example of detecting the reliability using the number of code inversions in the repeated decoding process. The soft decision output included in each bit of the packet can be used in the process of repetitively using the number of times of code inversion as reliability. index. That is, 10 is el and e2 in the inverse of the number i and the i+ weight, and the decoding result is unstable and the probability of the error is high, so the reliability is low. Fig. 27 shows an example of calculation processing when the reliability is detected by cross entropy in the reverse decoding, and can be calculated using the signal value shown in Fig. 24. In this case, the larger the cross-extraction TCE is, the worse the solution result is, and the probability of error is high, so the reliability is low. Fig. 28 shows an example of detecting reliability using the received SINR, and the smaller the received SINR included in the packet, the higher the probability of error, and therefore the reliability is low. Further, the lowest log likelihood ratio (LLR: Log 20 Likelihood Ratio) in the packet can also be used as a reliability index. The invention has been described above by way of a preferred embodiment of the invention. The present invention has been described with reference to the specific embodiments thereof, and various modifications and changes can be made to these specific examples without departing from the spirit and scope of the invention. That is, the invention should not be construed as limiting the details of the specific examples and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows a mapping method for a physical channel of an IP packet in a conventional HSDPA. 5 Fig. 2 shows an example of the number of bits of a header or the like in the physical channel mapping of a conventional IP packet. Figures 3(a) and (b) show examples of the distribution of IP packet sizes on the cable system internet. 4(a) to (c) are flowcharts showing the IP packet mapping method of the present invention. 10 Figure 5 shows the classification of the pattern of the IP packet mapping method of the present invention. Fig. 6 shows an example of the relationship between the IP packet and the MAC_PDU when the variable length is divided. Fig. 7 shows an example of the relationship between the IP packet and the MAC_PDU when the partition is divided into variable lengths and divided into a certain length or less. 15 Fig. 8 shows an example of the relationship between IP packets and MAC-PDUs when divided and variable lengths are divided by a fixed length or the like (only one variable length). Fig. 9 shows an example of the relationship between the IP packet and the MAC-PDU when there is a divided and fixed length. Fig. 10 is a view showing an example of the configuration of a communication system 20 to which the IP packet mapping method of the present invention is applied. Fig. 11 (a) and (b) show an example in which a channel coding block is generated from an IP packet via a MAC_PDU. Figure 12 is a conceptual diagram of adaptive modulation demodulation channel coding (A M C). Fig. 13 shows an example of a combination of a data modulation method and a channel coding rate. 25 1307229 Figure 14 (8), (b) shows an example of mapping a radio frame from a channel coding block. Figure 15 shows the processing of hybrid ARQ. The first (8) and (8) are diagrams showing the re-sending (four) 35-difference detection type hybrid arq according to the 5 CRC code used in the embodiment of the present invention. Fig. 17 (8) and (9) show an example of a conventional hybrid QQ generation request for retransmission of a signal error. Figure I8 shows an example of compensation for the retransmission signal error 10 formed by the conventional RLC layer. Fig. 19 is a view showing another configuration example of the communication system to which the IP packet mapping method of the present invention is applied.苐20 Figure (a) 7 (c) shows an example of a NACK signal. Fig. 21 shows an example of combining the ACK/NACK signal I5 indicating the overall reliability. Figure 22 shows an example of a block. Fig. 23 (8) to (c) show the types of processing of the hybrid ARQ. Fig. 24 shows an example of the configuration of the decoder. Fig. 25 shows an example in which the soft decision output is used and the reliability is detected. 20 Fig. 26 shows an example of detecting the reliability using the number of code inversions in the repeated decoding process. Fig. 27 is a diagram showing an example of calculation processing when the reliability is detected using the cross entropy in the repeated decoding. Fig. 28 shows an example of detecting reliability using the received SINR. 26 1307229 [Description of main component symbols] ίο... Transmitter 11... Transmit buffer 12... Split connection unit 13... Packet coding unit 14... Channel coding unit 15... Mapping unit 16... Modulation unit 17... Retransmission control unit 18... Control unit 20...receiver 21...demodulation unit 22...reception quality measurement unit 23...retransmission control unit 24...IP packet reproduction unit 25...reliability detection unit 100...base station 201...terminal 202...terminals S1 to S5...step S11 to S14...Steps S21 to S24...Steps S31 to S34...Steps P1 to P25...Packet 27