基於分子擴散之數位通訊

Abstract

作為奈米裝置間最有前途的通訊機制，分子通訊在通訊理論界與研究社群間都獲得極大關注。基於分子擴散之通訊指的是分子單純地依靠粒子擴散的理論來到達目的地。 此論文中，我們定義了兩種不同的數位分子通訊：同調及非同調分子通訊。對這兩種分子通訊，基於分子擴散的數位通訊系統之基礎架構設計均被提出。通道之記憶效應在分子通訊中為一決定性因素，而目前出版文獻卻都疏於考慮。本論文中提出之架構首先以分析之方式將任意量級的通道記憶納入考量。基於消息理論，可將互消息最大化之接收端偵測方案被提出。一低複雜度且不需前置機率訊息之接收端偵測方案亦被提出。數值結果顯示，本論文中所提出之低複雜度接收端偵測方案可同時確保兩種數位分子通訊系統在無限大的通道記憶環境下，系統亦可穩定運作。藉由延長位元區間長度，每一次通道使用達到一個位元傳輸之最大通道容量最終可被達成。 在所提出的數位分子通訊架構下，為了對抗多使用者干擾之效應，數種適用於多輸入多輸出通訊系統之多樣性技巧均被提出。傳送端多樣性、選擇結合、最大比例結合、決策融合、空間多工等適用於多輸入多輸出分子通訊系統之通訊技巧均被提出並分析以求得錯誤機率。透過虛擬多輸入多輸出通訊技巧來達到合作多樣性之概念亦被提出並探討其於分子通訊上之應用。數值結果顯示所提出之多樣性技巧可在多使用者干擾環境下顯著降低錯誤機率。若傳送端可獲得通道狀態訊息並適當分配分子數目，效能可獲得更進一步增進。

Molecular communication, being one of the most promising communication mechanisms among nanoscale devices, has attracted considerable interests from the communication theorists and research communities. Diffusion-based molecular communication refers to the situation where molecules reach the destination relying solely on the laws of molecular diffusion. In this thesis, two different families of digital molecular communication are identified, namely coherent and non-coherent digital molecular communication. For each family of molecular communication, the fundamental framework for designing a digital communication system based on molecular diffusion is proposed. Such a framework is the first to analytically take into account the effect of channel memory with arbitrary order, which is a key property of the diffusion channel that most of the existing literature fails to capture. Based on an information-theoretic approach, an information-optimal receiver detection scheme that can achieve maximum mutual information is proposed. A low-complexity receiver detection scheme without the knowledge of a priori information is also proposed. Results show that the low-complexity receiver detection scheme guarantees both families of molecular communication to be operated without failure in the case of infinite channel memory. A channel capacity of 1 bit per channel utilization can be ultimately achieved by extending the duration of the signaling interval. Under the proposed framework of digital molecular communication systems, various diversity techniques for Multi-Input Multi-Output (MIMO) transmissions are proposed to combat the effect of Multi-User Interference (MUI). The transmit diversity, selection combining, Maximum Ratio Combining (MRC), decision fusion, and Spatial Multiplexing (SM) for MIMO molecular communication are formulated and analyzed to obtain the error probability. The notion of virtual MIMO to achieve cooperative diversity is also investigated and proposed to be applied to molecular communication. Numerical results show the proposed diversity techniques can significantly lower the error probability in the presence of MUI. Further performance improvement can be obtained by properly allocating the molecules among the transmission nodes if the Channel State Information (CSI) is available at the transmitting end.