CAREER: Key Problems in Wireless Communications

这项研究可以分为两个主要主题，其动机是无线系统的以下关键瓶颈：无线电带宽的稀缺性，以及将固有损耗的无线链路与高速有线网络连接的问题，高速有线网络的损耗主要是由于拥塞造成的。第一个研究主题是如何提高直接序列CDMA的容量和性能，它有望成为未来无线系统的强大主力技术。由于多址干扰对CDMA性能造成了主要限制，因此考虑了两种广泛的方法来减轻其影响：分散自适应接收和集中多用户接收。我们的目标是促进这两种方法的理论发展，以期开发实用的接收器结构。所解决的问题包括设计使自适应干扰抑制在无线多址信道的严重时间变化上工作的方法，使用直接序列CDMA支持多种通信量（因此是多种速率），发明接近最优的自适应非线性接收器，将天线分集纳入具有合理复杂性的干扰抑制方案，对于不适用自适应方法的系统，推导出基于多用户检测和估计原理的接收机。第二个研究主题是开发一种设计接口的理论，以确保无线和有线网络之间的兼容性，同时最大限度地减少宝贵的无线带宽的使用。这些问题的表述虽然具有更广泛的意义，但其动机是由于无线网络必须与两种有线技术兼容，即新兴的ATM网络和TCP/IP。所考虑的问题包括在有损无线链路上的服务保证的制定和实现它们的设计原则，通过使用基于每个接收器看到的信道状态的传输控制来优化广播或多址无线信道的无线带宽，通过无线接口操作TCP/IP流量控制和ATM速率控制所使用的反馈机制，在具有无线和有线链路的异构网络上提供良好的端到端性能，并设计一种不受TCP/IP缺点影响的流量控制机制。

This research can be classified into two major themes, motivated by the following key bottlenecks for wireless systems: the scarcity of radio bandwidth, and the problem of interfacing inherently lossy wireless links with high-speed wireline networks where loss occurs mainlv due to congestion. The first research theme is concerned with enhancing the capacity and performance of direct-sequence CDMA, which holds great promise as a robust workhorse technology for future wireless systems. Since multiple-access interference imposes a major limitation on CDMA performance two broad methodologies are considered for alleviating its effects: decentralized adaptive reception and centralized multi-user reception. The goal is to further the theoretical development of both these methodologies with a view to developing practical receiver structures. The issues addressed include devising methods for making adaptive interference suppression work over the severe time variations of a wireless multiple-access channel, supporting multiple classes of traffic (and hence multiple rates) using direct-sequence CDMA, inventing near optimal adaptive nonlinear receivers, incorporating antenna diversity into interference suppression schemes with reasonable complexity, and deriving receivers based on multi-user detection and estimation principles for systems to which adaptive methods do not apply. The second research theme is the development of a theory for designing interfaces that ensure compatibility between wireless and wireline networks while minimizing the use of precious wireless bandwidth. The problem formulations, while of broader interest, are motivated by the fact that two wireline technologies with which a wireless network must be compatible are the emerging ATM networks and TCP/IP. The problems considered include the formulation of service guarantees over a lossy wireless link and the design principles for implementing them, optimization of wireless bandwidth for a broadcast or multiple-access wireless channel by using transmission control based on the channel state seen by each receiver, manipulation of the feedback mechanisms used by TCP/IP flow control and ATM rate control by the wireless interface to provide good end-to-end performance over a heterogeneous network with both wireless and wireline links, and the design of a flow control mechanism that does not suffer from the shortcomings of TCP/IP.