Signal Processing Solutions for the Networked Battlespace

现代战场的性质正在发生巨大变化。电子通信使平台之间前所未有的数据和信息交换成为可能。电子技术的进步使低成本联网无人值守传感器成为可能。因此，对从各种网络通信和武器平台获取的大量多传感器数据进行智能和稳健的处理，对于保持军事优势和减轻在无政府状态和扩展的作战区域（战场空间）内呈现多重威胁的智能对手至关重要。因此，我们组成了一个独特的学术专家联盟，来自拉夫堡、萨里、斯特拉斯克德和卡迪夫大学，以及六个工业项目合作伙伴QinetiQ、Selex-Galileo、泰雷兹、德州仪器、PrismTech和steep Ascent，以开发变革性的新信号处理解决方案，为Dstl、国防部和整个英国带来好处。为了实现这一目标，我们提出了一个由以下五个相互关联的工作包组成的五年综合工作计划：(1)网络化战场空间的高维自动统计异常检测和分类，我们的目标不仅是检测异常，而且是识别其性质和细微差别，当在高维复杂网络环境中获得时。将使用数据质量和模糊度量来确保常态性模型不被不可靠和模糊的数据破坏；(2)处理不确定性并结合领域知识，我们的目标是利用网络化战斗空间的世界模型来提高性能和信心，并将不确定性降低到前所未有的水平。此类信息的例子有关于地形和油田布局的数字地图、平台之间的几何关系和天气等操作条件；(3)信号分离和宽带分布式波束形成，其中我们的目标是设计低复杂度的鲁棒算法，用于欠确定和卷积源分离，以及基于低秩和稀疏表示的宽带分布式波束形成，以及它们的快速实现；(4)多输入多输出（MIMO）和分布式传感，其中我们打算为分布式MIMO雷达系统在混乱的网络战场空间中运行创建新的范例；(5)低复杂度算法和高效实现，其中我们与德州仪器，PrismTech和steep Ascent合作，旨在制定和实现网络环境中一系列复杂信号处理算法的新颖实现策略。这些相互关联的工作包经过精心设计，与Dstl和EPSRC确定的研究主题和挑战相结合，我们有明确的策略来获取数据集，执行评估和交流发现。我们设计了一个结构严谨的财团管理团队，包括一个由知名外部独立专家组成的总体指导小组，这些专家的专业知识涵盖了工作计划的范围。联合体的运营将由联合体董事和联合体管理团队负责。我们联盟管理的一个关键组成部分是鼓励被聘用的研究人员和学生定期借调到联盟内其他合作者的实验室，以便从合作大学和行业的互补知识和技能中受益；获得访问特权数据集和/或设备的权限；或共享资源，并在解决特定的Dstl挑战时提供临界质量。管理结构和协调措施的设计是为了使联合体有能力承担牵头联合体的角色，如果需要，与Dstl和EPSRC合作，建立一个信号和数据处理的实践社区，并确保英国在该领域拥有世界领先的能力。

The nature of the modern battlefield is changing dramatically. Electronic communication is allowing unprecedented interchange of data and information between platforms. Advances in electronics are allowing the possibility of low cost networked unattended sensors. Intelligent and robust processing of the very large amount of multi-sensor data acquired from various networked communications and weapons platforms is, therefore, crucial to retain military advantage and mitigate smart adversaries who present multiple threats within an anarchic and extended operating area (battlespace). Hence we have composed a unique consortium of academic experts from Loughborough, Surrey, Strathclyde and Cardiff universities together with six industrial project partners QinetiQ, Selex-Galileo, Thales, Texas Instruments, PrismTech & Steepest Ascent, to develop transformational new signal processing solutions to the benefit of Dstl, the MoD, and the UK in general. To achieve this goal we are proposing a five-year integrated programme of work composed of the following five interlinked work packages: (1) Automated statistical anomaly detection and classification in high dimensions for the networked battlespace, in which we aim not only to detect anomaly, but also to identify its nature and nuance, when acquired in a high dimensional complex network environment. Data quality and ambiguity measures will be used to ensure the models of normality are not corrupted by unreliable and ambiguous data; (2) Handling uncertainty and incorporating domain knowledge, within which we aim to exploit the world model of the networked battlespace to improve performance and confidence, and to reduce uncertainty to an unprecedented level. Examples for such information are digital maps about terrain and layout of the field, geometric relations between platforms and operational conditions such as weather; (3) Signal separation and broadband distributed beamforming, in which we target at designing low-complexity robust algorithms for underdetermined and convolutive source separation, and broadband distributed beamforming, facilitated by low-rank and sparse representations, and their fast implementations; (4) Multi-input and multi-output (MIMO) and distributed sensing, within which we intend to create novel paradigms for distributed MIMO radar systems operating in the cluttered networked battlespace; and (5) Low complexity algorithms and efficient implementation, in which with Texas Instruments, PrismTech & Steepest Ascent we aim to formulate and realize novel implementation strategies for a range of complex signal processing algorithms in a networked environment. These interlinked workpackages have been very carefully designed to marry up with the research themes and challenges identified by Dstl & the EPSRC and we have clear strategies for attaining datasets, performing evaluation, and communicating findings.We have designed a carefully structured consortium management team including an overarching steering group with renowned external independent experts with expertise covering the scope of the work programme. The operation of the consortium will be the responsibility of the Consortium Director and the Consortium Management Team. A key component of our consortium management is to encourage research staff and students employed to be periodically seconded to the labs of other collaborators within the consortium to benefit from complementary knowledge and skills at partner universities and industry; gain access to privileged datasets and/or equipment; or share resources & provide critical mass when addressing a particular Dstl challenge.The management structure and coordination measures have been designed for the consortium to have the capacity to assume the role of lead consortium, if required, working with Dstl & EPSRC to establish a community of practice in signal and data processing, and to ensure the UK has world leading capability in the area.