行動合作網通之無線訊號與訊息處理技術研究-子計畫三：合作式通訊網路之協同綠能編碼技術研究與設計

Abstract

合作式通訊為未來無線通訊技術之趨勢，透由無線通訊裝置彼此間之合作進行資訊的互相協助傳 遞與處理可達成更高效能之傳輸功能、容量及品質。由於通訊自由度的增加，合作式通訊引發諸多研 究議題，目前已開始被廣泛的研究與探討，其中最重要亟需探討議題之一為耗能問題，特別全球環保 意識抬頭之下，綠能電子是必要的技術發展趨勢。針對此議題，本研究進行實體層低耗能之合作式通 訊號處理演算法研究與設計。有別於傳統針對個別無線通訊訊號處理功能進行低耗能演算法設計，本 計畫除了進行無線裝置彼此(Inter)之低耗能合作演算法之設計外，並將合作之概念延伸至個別裝置內 (Intra)的方塊間之協力(Synergy)低耗能訊號處理處理演算法設計，我們並將不同型式的訊號均視為 “碼”，而其操作與處理均可視為 “編碼”，因此本計畫的目標在於達成優化合作式、具有綠能概念之無 線通訊訊號處理編解碼設計。 基於上述理念，我們除了將探討合作式無線網路之整合綠能編碼技術，以及無線裝置間之低耗能 網路編碼設計、優化之中繼台及天線功率分配外，亦將探討裝置內之低耗能協同訊號處理技術。本計 畫原來為去年申請三年計畫但僅獲核准一年，今年度進行約6 個月已經達成諸多初步成果，本申請計 畫為延續今年度執行中之計畫，因此規劃兩年，此兩年計畫為原始計畫之第二及第三年計畫。本計畫 第一年(明年，101 年)主要研究議題有: (i) 進階取樣理論與技術之研究。近十年來有一些突破性新取樣 理論與技術之進展，其可以較低於傳統理論之Nyquist 取樣率之取樣率取得完整訊號資訊，因此或可 能降低系統實現複雜度及耗能。目前我們正進行基礎理論之研究，明年第一年我們將設計出低耗能之 取樣技術及相對之適應性類比數位訊號轉換取樣演算法與相對的傳送端資料編碼設計以達成降低功 耗的效果; (ii) 高效能之通道編解碼法及合作式通道解碼演算法設計，明年(第一年)將設計出高效能且 低耗能之基因輔助LDPC 解碼演算法及合作式LDPC 與Turbo; (iii) 進行較新、非2 多次方之完整點數 及不規則、非完整點數的FFT/IFFT 之低耗能演算法設計。FFT/IFFT 為傳輸編碼“transmission coding” 運算中之一主要運算，這些較新的計算出現於一些通訊系統，如LTE, LTE-A, IEEE 802.16m, DVB-T2， 並涵蓋OFDM/OFDMA與 SC-FDMA傳輸方式均需用到; 目前正研究各標準之FFT 規格包含不同FFT 點數及不完整點數FFT(適用於FDMA 或OFDMA 系統)，明年第一年將完成一個高效能、低耗能、高 彈性的FFT 處理器架構；(iv) 低耗能之合作通訊站台辨識演算法設計; 目前正研究低耗能之辨識演算 法，明年第一年將進一步研究新的LTE 系統PSS 及SSS 碼；其可較標準系統具更低之耗能，但效果 至少不輸於現有之辨識碼; (v)低耗能之多資料源與中繼站之網路編碼設計，目標在於提升整體系統傳 輸效能且盡可能降低耗能; 目前已有一初步多sources 與destinations 的網路編碼方法，雖然效能不 差，但仍有諸多問題要克服；(vi) 優化天線及中繼台耗能之分配; 目前我們針對等距多中繼台架構已 有初步成果，明年第一年將針對一般非等距之中繼hop 網路作優化研究(vii) 研究LTE-A 及802.16m 實體層之規格及相關之通訊訊號處理。 第二年(102 年)主要工作有: (i) 將延續第一年之初步研究成果進行更高效能之技術開發設計; (ii)依 據新的取樣技巧，進行第一年編碼運算法之協同、修改、優化設計，例如通道編碼設計、FFT 運算; (iii) 以LTE/LTE-A，或802.16m 通訊標準為應用載具進行演算法之C 語言或、及DSP 實現設計，並與其 他子計劃進行整合設計模擬及作最終展演系統之設計。

Cooperative communication has been becoming the future mainstream communication technology since recent years. Via cooperation, wireless mobile devices can collaborate such that performance of overall network can be greatly improved. Due to collaboration, various research and design issues are incurred. Among those issues, energy consumption is particularly important, as the global awareness on environmental protection receives significant attention worldwide. As a result, green electronics is a necessary foundation for future environment-aware telecommunication technologies. In response to this issue, this project is aimed to investigate and develop low energy-consuming PHY signal processing algorithms for cooperative network. Different from conventional approaches that only consider low-power design of a single function module in the entire baseband transceiver, this project will consider the synergetic low-energy designs of “inter” cooperative wireless stations and “intra” composing function modules within a station. Conceptually, we also treat various signal forms simply as “codes”, and all sorts of operations and manipulations on the codes are simply “coding operations”. The concept and vision, in an abstract form, is suggested in the project title which also summarizes the main theme of the project. With the vision in mind, in addition to investigating “green” coding algorithms for cooperative inter- station operations, this project will also investigate synergetic green coding algorithms for intra-station baseband operations. The project spans for two years. In the first year, the main research issues are: (i) Develop new low-power sampling techniques and their associated signal processing techniques. There were some breakthroughs in sampling techniques in the recent ten years, which can sample signals with sub-Nyquist rate. That means system complexity (and therefore the energy consumption) can possibly be reducel. (ii) Design of high-performance cooperative channel coding methods which can achieve higher performances than the state-of-art coding methods (such as LDPC and turbo codes), but with lower energy consumption. Particularly, we will combine the best of both LDPC and turbo codes, and also apply genetic algorithm to LDPC decoding. (iii) Design new low energy-consuming FFT algorithms for various new DFT operations with non-power-of-2 DFT sizes and/or partial (and random) inputs. Those DFT operations are commonly seen in many advanced systems such as LTE, LTE-A, IEEE 802.16m, DVB-T2 system, and in both OFDM/OFDMA and SC-FDMA transmission modes. We will then also design low-power, flexible FFT architecture for all these heterogeneous FFT operations. (iv) Development of green station identification algorithm and codes. We will design new LTE PSS and SSS codes which have much lower power consumption, but with comparable performances to the standard codes. (v) Design new network coding algorithms for multi cooperative mobile and relay station with low energy consumption and high network capacity. (vi) Power optimization of cooperative antenna and relay stations. We will optimize power-capacity tradeoff for general unequally-spaced multi-hop relay network. (vii) Study and design of LTE-A and 802.16m PHY functions and associated signal processing algorithms. In the 2nd year, the main tasks include: (i) continuing and improving the first year researches; (ii) based on the first investigation on new advanced sampling techniques, new energy-aware sampling techniques, adaptive low-power ADC sampling method and the associated transmitter data recoding algorithms will be proposed; (iii) considering the new sampling methods, conduct cooperative and synergetic design of the first year research results for low-power consideration; (iv) based on LTE/LTE-A, or 802.16m system, realizing the developed techniques with C language and/or DSP platform, and conduct C/DSP integration design with other subprojects for final demonstration.