Differentiated and adaptive robustness in wireless access networks

由于两个因素，传统电信基础设施的健壮性正在减弱：无线技术的使用增加，其本质上比有线更容易受到攻击；以及向IP的融合，这是在设计时考虑到“尽最大努力”的哲学。这种方法有明显的管理优势，但几乎可以肯定的是，由于无线和互联网技术的不足，这一综合系统的可靠性将受到影响。对于普通家庭用户来说，这可能不是一个特别紧迫的问题，但对于愿意付费的用户来说，期望可靠性和服务质量可能非常重要。这在紧急情况下也是极其重要的，即使故障不会直接影响系统，也会有大量流量激增。在这种情况下，网络可能过载，并且可能发生不能中继通信的情况，因为它发生在最近的事件中。这些都是非常严重的问题，必须在无线/互联网融合完成之前尽快解决。在本项目中，我们将研究当前提出的一些技术如何在有线互联网中提供可用于提高健壮性的某些形式的服务保证。另一个困难是，这些技术必须与无线技术一起工作，这是一个到目前为止还没有研究过的问题。这一点非常重要，因为几年后，我们就可以拥有一个完全基于互联网的网络，拥有大量的无线用户。很可能在紧急情况下，这些用户，包括救援和安全人员，都无法访问网络，因为紧急情况造成了普遍的过载。这显然不是一种选择，这就是为什么我们想现在就开始研究这些问题。

The traditional telecommunications infrastructure robustness is diminishing because of two elements: the increased use of wireless technologies that, by its nature, are more vulnerable than wireline, and the convergence towards IP, which was designed with a "best-effort" philosophy in mind. There are obvious management advantages to this approach but it is almost certain that the reliability of this integrated system will suffer from the inadequacies of both wireless and Internet technology. This may not be a particularly pressing problem for the casual home user, but it could be very important for users that would pay to expect reliability and Quality of Service. It would be also extremely important in emergencies situations, where even if a failure did not affect directly the system, there would be a large surge of traffic. In this case, the network may be overloaded and it may happen that no communications can be relayed as it happened in recent events. These are very serious problems that have to be addressed as soon as possible before the wireless/Internet convergence is complete. In this project, we examine how some techniques currently proposed provide some forms of service guarantees in the wired Internet that can be used to improve robustness. An additional difficulty is that these techniques must be made to work with wireles technology, a problem that has not been examined so far. This is very important since in a few years, we could have a completely Internet-based network with a large population of wireless users. It could very well happen that in an emergency situation, none of these users, including relief and security workers, could access the network, because of the general overload caused by the emergency. This is clearly not an option and that is why we want to start right now to look at these issues.