具備多封包接收能力的無線網路中的機會式媒體接取控制

Abstract

電信產業即將成為我國的第三個兆元產業。而無線通訊與網路是電信國家型 計畫的三個組成部份中最受看好的領域。由於可利用的無線電頻譜先天上相對 於光纖頻寬是很稀少的，在可預見的未來我們仍然需要更先進的無線通訊與無 線網路技術。 我們將藉由這個研究計畫設計一個全新的無線網路媒體接取控制方案。這個 媒體接取控制方案將整合無線網路中的多封包接收(Multiple Packet Reception)技 術和機會式媒體接取控制(Opportunistic Medium Access Control)技術，並取得技 術領先地位。 機會式媒體接取控制主要是利用無線通訊管道的時變性(如fading)，只使用 瞬間情況良好的管道傳送封包，以增加無線網路的總體產出(throughput)。傳統的 通訊理論把通訊管道的時變性視為必須克服的缺點；而機會式媒體接取控制則將 無線通訊管道的時變性視為一種可以利用的資源。目前，大多數關於機會式媒體 接取控制的文獻都是假設基地台不能同時接收多個封包。 多封包接收能力是指一個基地台可以在同一時間(time slot)、同一頻帶 (frequency band)同時成功地接收並解碼來自不同網路節點的多個封包。在傳統的 (0,1,e)通道模型中，同時傳送兩個以上的封包只會造成碰撞(collisions)，對網路 Throughput 並沒有任何貢獻。然而，網路消息理論(Network Information Theory) 與先進的通訊技術(如CDMA+Multi-user Detection 及Smart Antennas+SIC等)允許 基地台可以同時成功地接收多個封包。以先進的通訊技術為基礎的medium access control scheme 可以達到傳統的medium access control schemes 所無法達到 的無線網路效能。 我們將使用機率模型與排隊理論來做網路效能分析。我們預期這個研究計劃 將導致至少6 篇IEEE 期刊論文的發表。此外，我們也希望能有一專利權的發表。 我們希望藉由相關的論文發表進一步地提升我國在相關學術界的國際位階。

The telecommunications industry will soon become the third thousand-billion-dollar industry of Taiwan. Among the three parts of the national telecommunications project, wireless communications and networking is the most promising one. Since the wireless spectrum is inherently scarce, it is necessary to develop novel wireless communications and networking technology that utilizes the wireless bandwidth efficiently. In this project, we will design a novel opportunistic medium access control scheme for wireless networks with multiple packet reception capability. We expect to become one of the leading teams in this promising research field. An opportunistic medium access control scheme takes advantages of the time-varying nature of wireless communications channels. In particular, in order to maximize the network throughput, an opportunistic medium access control scheme uses only channels with good instantaneous qualities to transmit packets. In conventional communications theory, channel fading is seen as a detrimental property. In contrast, from the viewpoint of opportunistic medium access control, channel fading is beneficial to network throughput. Currently, most works on opportunistic medium access control use the conventional (0, 1, e) collision model. In a wireless network with multiple packet reception, an access point could simultaneously receive/decode a number of packets from different sources, which use the same frequency band to transmit packets. In the conventional (0, 1, e) channel model, simultaneous packet transmissions to an access point results in packet collisions and leads to a decrease in network throughput. However, network information theory implies that it is possible for the access point to simultaneously receive/decode a number of packets from different sources. The multiple packet reception capability could be realized by advanced communications technologies such as CDMA-based multi-user detection and smart antenna with successive interference cancellation. More important, the capacity region of an information-theoretic medium access control scheme is larger than that of an arbitrary pure TDMA-based medium access control scheme. We will use probability theory and queueing theory to analyze the network performance, when the proposed opportunistic medium access control scheme is used in wireless networks with multiple packet reception. We plan to publish at least six IEEE journal papers within three years. The publication of our research work should help us get recognized worldwide.