合作式通訊之媒介存取協定設計

Abstract

傳輸的可靠度是無線通訊中最重要的議題之一，為了改善傳輸的可靠度，針 對包括實體層 (PHY)及媒介存取層 (MAC)的許多不同技術紛紛被提出。藉著從多 輸入多輸出系統(MIMO system)得到的靈感所提出的合作式通訊，能夠在得到分 集增益的同時避開使用多天線技術的複雜度，而得到的增益在適當的使用下即可 改善傳輸可靠度及減少功率的損耗。 合作式通訊中，在傳送者 (source device)鄰近的裝置我們稱為中繼者 (relay)，他可以將從傳送者收到的訊框 (frame)再次送到接收者 (destination)。要達成這樣的合作式傳輸，包含傳送者、中繼者及接收者在內 的裝置必須知道 (1)什麼時候應該合作 (2)應該跟誰合作。為了滿足這些要求， 在媒介存取層中這三方互相的信令傳遞與協調是必須的。 在這篇論文中，我們設計了一種支援合作式傳輸的新的媒介存取協定，這個 新的協定允許裝置在動態的無線網路環境中使用不同的傳輸方式 (是否與他人 合作)。遵循我們的協定，這些裝置可以互相交換關於通道品質的資訊，並藉由一 個完全分散式的方法，透過競爭，來進行中繼者的選擇。為了選到最好的中繼者， 我們也制定了一套｢忙線音｣中繼者分級系統。除此之外，我們也將一項名為｢自 適應調變｣的跨層最佳化技術整合進我們的協定之中以獲得更佳的成果。 我們除了數學的分析也藉由網路模擬軟體 Opnet Modeler，進行多樣的模擬 來展現我們的成果。在低信噪比的環境下，整體的吞吐改善量約在 10%到 20%之 間變動，這個結果也證明了我們的提出的協定是可行的，並且在考慮到所有額外 負擔的情形下也能有相當的吞吐改善量。

Transmission reliability is one of the most important issues in wireless communi- cations. Different techniques have been proposed to improve transmission reliability in both physical (PHY) and medium access control (MAC) layers. Inspired by the multi-input multi-output (MIMO) system, cooperative transmission was introduced to obtain diversity gain without the complexity of using multi-antenna radios. This additional gain, when being used properly, can improve transmission reliability or reduce power consumption. In cooperative transmission, neighbors of the source device, referred to as relays, forward the data frame received from the source device to the destination device. To enable such cooperation, devices including the source, the relays, and the destination must know (1) when to cooperate and (2) whom to cooperate with. To address these issues, MAC-layer signaling and coordination among the three parties are needed. In this thesis, we design a new medium access control (MAC) protocol that sup- ports cooperative transmission. The new protocol enables devices to adapt trans- mission schemes (direct or cooperative) to dynamic wireless environments. With our protocol, the devices exchange the link information and perform relay selection — via contention — in a fully distributed manner. In order to select the best relay, we also introduce a ”busy-tone” relay raking system. To further improve the performance, a cross-layer optimization technique, namely adaptive modulation, is integrated into our protocol. We performed mathematical analysis and conducted extensive simulation in Opnet Modeler to show the performance. The overall throughput improvement varies from 10% to 20% in low-SNR environments. The results show that our protocol is a feasible iisolution and delivers considerable improvement even when taking into account all signaling overheads.