Adaptive Bayesian Receivers in Fading Channels: A Sequential Monte Carlo Filtering Design Paradigm

摘要即将到来的无线通信技术将使信息可及性有一个巨大的飞跃。所谓的第四代及以后的无线系统的先进特性，如与多媒体应用程序兼容的数据速率，将使当前无线系统无法实现的许多新兴应用成为可能。然而，目前还不清楚如何优化设计无线接收器，以满足未来无线系统中固有的更宽带宽和更高数据速率所带来的技术挑战。人们普遍认为，未来无线接收器的真正利基在于自适应系统的发展，以执行复杂的信号处理功能。但是，目前还缺乏可以用来设计这些未来接收器的具体原则。在这个阶段，重要的是获得可能有助于在这一领域引发革命性突破的见解和理论工具。研究了单用户和多用户衰落信道中自适应贝叶斯接收机的设计方法。该方法将未知时变信道中的信号接收问题表述为多元贝叶斯推理问题。时序蒙特卡罗滤波方法是统计领域最近发展起来的一种相对简单但功能极其强大的数值技术，将用于开发用于计算通道和数据的贝叶斯估计的自适应系统。在时序蒙特卡罗贝叶斯估计的统一框架下，将处理无线通信中发现的一系列接收器设计问题，如缓解各种射频干扰（包括多址干扰、窄带干扰、脉冲噪声）、衰落信道跟踪、解决多径信道色散、多天线时空处理、利用编码信号结构等。本计画的理论工作可望最终形成适用于未来无线系统的新型接收器设计概念。

AbstractThe coming generation of tetherless communication technology promises a giant leap forward in informationaccessibility. Advanced features of the so-called fourth-generation wireless systems and beyond, such as data rates compatible with multimedia applications, will enable many emerging applications not possible with current wireless systems. However, it is not at all clear how wireless receivers should be optimally designed to meet the technical challenges introduced by the wider bandwidths and higher data rates inherent in the future wireless systems. It is generally believed that the real niche for future wireless receivers lies in the development of adaptive systems to perform sophisticated signal processing functions. But, at this time there is a lack of concrete principles that can be used to design these futuristic receivers. It is important at this stage to acquire the insights and theoretical tools that may help spark revolutionary breakthroughs in this field.Investigation of design methodologies of adaptive Bayesian receivers in single-user and multiuser fading channels is proposed. The approach is to formulate the problems of signal reception in unknown time-varying channels as multivariate Bayesian inference problems. Sequential Monte Carlo filtering methods, the relatively simple but extremely powerful numerical techniques recently developed in the field of statistics, will be employed to develop adaptive systems for computing the Bayesian estimates of the channels and data. An array of receiver design problems found in wireless communications, such as mitigation of various types of radio-frequency interference (including multiple-access interference, narrowband interference, impulsive noise), tracking of fading channels, resolving multipath channel dispersion, space-time processing by multiple antennas, exploiting coded signal structures, etc., will be treated under the unified framework of sequential Monte Carlo Bayesian estimation. The theoretical effort in this project is expected to culminate in the formulation of novel receiver design concepts applicable for future wireless systems.