無線通信系統中之通道編碼將相對的特別重要，尤其為了提高其通訊品質，相關 MIMO 也將為引進。視訊的傳輸又將是然而無線通信系統中的最重要的傳輸應用。結合此兩技術的相關研究將會是關鍵的研究。本計畫主要為實現在無線通信系統中，來傳輸視訊資料，其主要應用為實現在車用系統。相關技術可分為兩大部分，其中第一大部分乃因為傳輸大量資料於無線通信系統，所因應之相關錯誤更正編碼的技術之相關研究。近年來在無線通信系統的應用有極迅速的進步，在下一代無線通信系統的目標是更高的資料傳送速率和增加通信覆蓋區域面積。由於網路傳輸技術的進步以及多媒體視訊編碼的發展成熟，使得網路視訊傳輸成為現今數位生活化的主要訴求，行動無線通訊之要求將是一大考驗。在無線行動通訊之應用下，通道環境更為惡劣導致錯誤接收的發生更需被重視，前饋式錯誤更正技術提供了解決這個問題的另一途徑，資料經過前饋式錯誤更正編碼後可以在更低的信號強度下得到同樣的資料正確率，而編碼過程中所耗費的頻寬，則可以使用頻寬高效率調變技術以在固定的寬帶中傳送更多資料來解決。在現今的通信系統中,相連編碼(concatenated codes)被廣泛的使用，低密度奇偶校驗編碼之編碼增益可達到極接近向農界限之程度。本計畫之第二大部分，乃是植基於視訊系統中相關可伸縮性視訊編碼的技術。由於接收端的頻寬隨著通訊傳輸介質多樣化，而具有不同的接收能力，視訊編碼的功能通常只有在特定的位元率之下來進行，因此無法應用在多元化的接收頻寬環境中，所以可伸縮性視訊編碼因應而生，它主要可以分為時域性可伸縮性、空間可伸縮性及品質可伸縮性三種，藉由這三種可伸縮性的方式，使用者就可以依據本身的接收能力來選擇適合的品質觀看。一般訊息編碼無論是Image或Video傳輸的standard，許多都是針對無雜訊之通道 (noiseless channel) 所制訂的壓縮標準，而通道編碼則針對通道雜訊作最佳化的保護也制訂了相關的standard。但是在現實的環境裡必須考慮通道雜訊對訊息資料的影響，故本計畫將深入探討視訊/信道聯合編碼之最佳化傳輸架構，並適用於車用無線視訊會議系統。本計畫之相關係統以DVB-V通稱，其中對DVB-V中SVC之視訊技術制訂相關規格，在畫面解析度的部分可以從QCIF(176x144)支援到CIF(352x288)，在播放率的部分可以從 15 frame/s支援到30 frame/s，而相對應所需要的位元率是從128kbps到384kbps，即使在不同的接收頻寬條件之下，仍然可以維持最基本的視訊品質，並且在通道編碼中可以對可伸縮性編碼所產生的Base layer及Enhancement layer來做不同層次的保護動作。本計畫除了研究通道編碼與可伸縮性視訊編碼之技術外，更進而討論相關視訊/信道聯合編碼之概念，以求得整體系統應用之最佳表現。 In recent years, the wireless system is to transmit higher data rate and have a larger coverage area. However, radio transmission power needs to be kept to a minimum due to regulation and system power consumption reasons. Forward-error-correction (FEC) system can be employed to solve this problem. Traditionally, concatenated FEC based on convolutional codes are employed in wide variety of systems. It has been demonstrated that LDPC codes can achieve coding gain performance that is only several tenth of dB from the Shannon bound. For various transmission media and wireless transmission media, the bandwidth variation is obvious to various transmission media. However, the conventional video coding standard is constructed with many coding tools to improve coding efficiency for a given bit rate. Therefore, the conventional video coding standard is not suitable for applications of various bit rates. The newly established scalable video coding (SVC), which is capable of generating scalable bitstream for various bit rates, have been developed for universal access. The demand to improve quality and increase the capacity of such systems within limited bandwidth resources motivates the interest in efficient and error-resilient image coding methods. An important problem is how to efficiently encode images for transmission over limited-bandwidth channels while maintaining good decoded image quality and resilience to channel impairments. For SVC in DVB-V, three kinds of scalability are provided as QCIF (176x144) and CIF (352x288) of display size, 15 frame/s and 30 frame/s of temporal scalability and 128kbps to 384kbps of bit rate. Consequently, the basic visual quality can be achieved with various bite rates. Moreover, for improving flexibility of bit rate, the channel coding can apply various level of protection for Base layer and Enhancement layer, respectively. In addition, our project will target on the research of Joint Source-Channel coding (JSCC) to achieve the optimized performance in the system. 研究期間：9608 ~ 9707
