New class of linear-complexity overhead-optimized random linear network codes

对数据广播、大数据下载、在线视频、互联网电视、电话会议等的需求增加，已经导致通信网络资源的使用不断增加。思科预计，到2017年，互联网流量将达到1.4万亿千兆字节，互联网连接设备的数量将达到人均3台设备。现有的网络通信协议将无法处理这种巨大的需求，除非以增加许多新的基础设施为代价，增加连接带宽或增加服务器的数量。显然，增加新的基础设施是一个非常昂贵的解决方案，而且由于资源稀缺，并不总是可行的。此外，大多数现代通信网络（诸如无线网络）的动态性质使得传统协议甚至更加低效。因此，我们需要新的解决方案，否则现有网络基础设施的容量很快就会不堪重负。

Increased demand for data broadcasting, large data downloads, on-line video, internet TV, teleconferencing, etc., has led to an ever increasing usage of communication network resources. Cisco has projected the Internet traffic to reach 1.4 trillion gigabytes and the number of Internet connected devices to reach 3 devices per capita by 2017. Existing network communication protocols will not be able to handle this enormous demand unless at the cost of adding much new infrastructure, e.g., increasing the connection bandwidth or increasing the number of servers. Obviously, adding new infrastructure is a very costly solution and is not always feasible due to scarcity of resources. Furthermore, the dynamic nature of most modern communication networks (such as wireless networks) makes the conventional protocols even more inefficient. As a result, new solutions are needed or the capacity of our existing network infrastructure will soon be overwhelmed.