Performance of Networked Passive Radar Systems with Multiple Transmitters and Receivers

该项目的目标是展示具有多个发射机和接收机的无源雷达系统的突出但未开发的潜力。无源雷达是一种强大的方法，它使用现有的环境通信信号，如无线电和电视广播，或卫星，蜂窝和WiFi信号，来检测，成像或分类物体，并估计它们的位置和运动。由于无源雷达使用现有的通信信号，它可以大大降低成本，复杂性和能源使用，同时在需要快速部署雷达的紧急情况下尤为重要。通过理论分析、算法评估，该项目将展示通过适度增加实际雷达系统模型的发射和接收天线数量所获得的巨大性能增益。这些贡献将对信号处理、传感器网络、机器学习和雷达系统研究产生重大影响。由于新的统计问题将被考虑，新的理论发展应该提供数学和统计学的贡献，同时导致实用的算法，并最终改善雷达系统的空中交通管制，国土安全，执法（穿墙成像），监视，海洋监测，天气监测和环境监测。这些研究应有助于主动雷达、声纳、超声波、声学和其他类似的主动和非主动传感器技术中多目标情况的性能分析。它应该鼓励智能家居、企业和汽车的新应用。 该项目还将提供充足的机会，为教育研究生，最好是来自代表性不足的群体，在信号处理和能源的重要跨学科领域，通过本研究项目，课程和利哈伊的综合电力网络（INE）倡议之间的协调，PI正在领导。 该项目中的传感研究与INE倡议中的几项活动很好地结合在一起。研究结果也将被纳入当前和未来的Lehigh课程，希望课堂笔记将演变成一本书和短期课程的网络无源雷达，以提供广泛的教育影响。最佳可能的性能，实际估计的位置和速度矢量的对象使用的无源雷达与M发送和N接收站将首次得出。基于简化系统模型的初步结果，该项目预计将展示通过适度增加MN为现实系统模型所获得的巨大性能增益。这些成果迄今尚未被观察到，应该鼓励MN 1无源雷达技术研究活动的巨大增长。这些贡献将对信号处理、传感器网络、机器学习和雷达系统研究产生重大影响。 所提出的方法将采用局部/非局部/贝叶斯/非贝叶斯有限MN性能的界限;最近的收敛结果的依赖随机变量的总和，以指导启发渐近分析;仔细选择相关反射系数，相关噪声和其他重要退化模型;增强模型的电磁理论基础上;最近开发的目标和杂波模型;以及最有希望的机会信号，包括显示出重大前景的MIMO通信信号。将采用多用户/迭代检测和干扰信道的成熟主题，以包括估计机会发射信号时所引起的劣化，并考虑可能泄漏到被认为仅是反射信号的直接路径信号的任何分量。同时采用几种不同类型的信号的机会和不同的电台布局的影响将被揭露。

The goal of this project is to demonstrate the outstanding but untapped potential of passive radar systems with multiple transmitters and receivers. Passive radar is a powerful approach that uses existing ambient communication signals such as radio and television broadcasts, or satellite, cellular and WiFi signals, to detect, image or classify objects and estimate their position and motion. Since passive radar uses existing communication signals it can drastically reduce cost, complexity and energy usage while being especially important in emergency settings where one needs to quickly deploy a radar. Through theoretical analysis, algorithm assessment, the project will demonstrate the tremendous performance gains obtained through moderate increases in the numbers of transmit and receive antennas for realistic radar system models. These contributions should have significant impact to signal processing, sensor networking, machine learning and radar systems research. As new statistical problems will be considered, new theory developed should provide contributions in mathematics and statistics while leading to practical algorithms and ultimately improved radar systems for air traffic control, homeland security, law enforcement (through-wall imaging), surveillance, ocean monitoring, weather monitoring, and environmental monitoring. These investigations should provide contributions relating to the performance analysis of multiple target cases in active radar, sonar, ultrasound, acoustics and other similar active and nonactive sensor technologies. It should encourage new applications for smart homes, businesses and cars. This project will also offer ample opportunities for educating graduate students, preferably from under-represented groups, in the important cross-disciplinary areas of signal processing and energy via coordination between this research project, classes and Lehigh's Integrated Networks for Electricity (INE) initiative, which the PI is leading. The sensing research in this project couples well with several activities within the INE initiative. Research results will also be incorporated into current and future Lehigh classes with the hope that class notes will evolve into a book and short course on networked passive radar to provide broad educational impact.The optimum possible performance for realistically estimating the position and velocity vectors of objects using a passive radar with M transmit and N receive stations will be derived for the first time. Based on presented preliminary results for a simplified system model, the project is expected to demonstrate the tremendous performance gains obtained through moderate increases in MN for realistic system models. These gains have not been observed to date and should encourage a tremendous increase in research activity on passive radar technology with MN 1. These contributions should have significant impact to signal processing, sensor networking, machine learning and radar systems research. The proposed approach will employ local/nonlocal/Bayesian/nonBayesian bounds for finite MN performance; recent convergence results for sums of dependent random variables to guide enlightening asymptotic analysis; carefully chosen models for correlated reflection coefficients, correlated noise and other important degradations; enhanced models based on electromagnetic theory; recently developed target and clutter models; and the most promising signals of opportunities, including MIMO communication signals which show significant promise. The well-developed topics of multiuser/iterative detection and interference channels will be employed to include the degradation incurred when estimating the transmitted signals of opportunity and to account for the any components of the direct path signals that may leak into what is thought to be only the reflected signals. The impact of simultaneously employing several different types of signals of opportunity and different station placements will be uncovered.