COoperative DEcision making based on Slepian-Wolf/multiple Access wireless Networks

在本项目中，我们成功地估计和利用了分布式协作无线网络中的信源相关性，以达到接近极限的Slepian-Wolf多址接入信道(Slepian-Wolf Multiple Access Channel，SWMAC)速率域的理论性能。我们发现，使用Memory-1Rate-1掺杂累加器可以很容易地到达外在信息传递(退出)图中非常接近(1，1)互信息的点，从而使误差平台看不到(10^-7以下)。对于宽带传输，我们的多输入多输出(MIMO)Turbo均衡器的工作效率非常高，其性能与理论上的MIMO/SW-MAC限制仅相差约1分贝。即使使用非常简单的存储器卷积码，我们的系统也能很好地工作。潜在的应用是功率受限的无线多跳和自组织无线网络、传感器/监测网络，其中传感器收集相关数据并以更高的频谱效率将其传输到公共目的地(融合中心)。该项目的主要出版物是关于IEEE信号处理的会刊、IEEE通信通讯和一些重要的IEEE国际会议。

In this project, we have successfully estimate and exploit the source correlation in distributed cooperative wireless network to achieve close-limit Slepian-Wolf multiple access channel (SW-MAC) rate region theoretical performance. We found that point very close to (1,1) mutual information of the extrinsic information transfer (EXIT) chart can be reached easily by using memory-1 rate-1 doped accumulator, which make the error-floor invisible (below 10^-7). For broadband transmission, our multiple input multiple output (MIMO) turbo equalizer work very efficiently to approach a performance only about 1 dB away from the theoretical MIMO/SW-MAC limit. Our system also work well even when very simple memory-1 convolutional codes are used. The potential applications are for power limited wireless multi-hops and ad-hoc wireless network, sensor/monitoring networks, where sensor collect the correlated data and transmit it to a common destination (fusion center) with higher spectrum efficiency. The main publications of the project are on IEEE Transaction on Signal Processing, IEEE Communication Letters and some major IEEE international conferences.