Army 8. Small Business Innovation Research (sbir) Proposal Submission Instructions
Yüklə 0,91 Mb.
|
| • The Domain Context is a model that defines the focus for the application of visualization approaches - i.e. where visualization approaches will be applied, who will be supported, and why those approaches are needed. 1
• Descriptive Aspects (DA) define what needs to be described for particular domain contexts. For example, DAs could be defined in terms of the various elements (or things) that are of importance, the relationships between those elements, and particular attributes that describe the elements and relationships. • The Visualization Approach dimension defines how the required information can be provided through computer-based visualization. Approaches are characterized in terms of the visual representations used (e.g., graphs, charts, maps) and related visual enhancements (e.g., use of overlays, distortion, animation). PHASE I: Using available literature resources, investigate and analyze key aspects of data visualization approaches. Assessments of task complexity, dynamic context elements, the cognitive cycle of perception / comprehension / projection, differences in individuals’ training and background, the transformation of raw data into information, and data streams that are typical and/or representative of the tactical environment are a partial list of the factors that should be considered. A methodology for combining those factors into a representation that could drive visualizations is desired. PHASE II: The methodology and framework should be refined during the first portion of phase 2. An assessment of how to validate specific approaches should be presented. The contractor will then develop a prototype that uses actual and/or representative data inputs, the various factors driving the visualization, and a Common Operational Picture that would be used by commander and staff to develop situational awareness and understanding in an optimal way. SME feedback into prototype visualizations is desired. Measurements of how well specific visualizations fit the objective are expected, as the usefulness and appropriateness of the visualization techniques are expected to be matured over the life of the SBIR. The contractor will work with the government to develop an initial plan for how the techniques could be integrated into systems of record. This would include proposed mechanisms for collecting user and situation data in real-time to update visualizations. PHASE III DUAL USE APPLICATIONS: During Phase 3 the prototype will be matured, and the contractor will work with the government to demonstrate the software to user juries in operational settings. Quick incorporation of feedback into the software baseline is expected. A real-world implementation/ deployment analysis should also be performed; the contractor will work with the government to establish a process for integrating the data collection and visualization approaches into one or more typical systems of record, such as the Command Post Computing Environment (CPCE). The contractor will work with the government to identify other potential adopters of the technology, such as the Department of Homeland Security and the Federal Emergency Management Agency. REFERENCES: 1. Gouin, Evdokiou, Vernik, A Showcase of Visualization Approaches For Military Decision Makers, 2004 2. Additional Q&A from TPOC (uploaded in SITIS on 1/5/18). KEYWORDS: Data, Visualization, Decisions, Informed, Cognitive, Aid, Technical, Evaluation.
TECHNOLOGY AREA(S): Information Systems The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), which controls the export and import of defense-related material and services. Offerors must disclose any proposed use of foreign nationals, their country of origin, and what tasks each would accomplish in the statement of work in accordance with section 5.4.c.(8) of the Announcement. OBJECTIVE: Utilize machine learning techniques to make inferences on the training set of another machine learning classifier, in order to manipulate inputs to generate desired outputs to harden network security applications. DESCRIPTION: Recent research has demonstrated an ability to utilize machine learning techniques in a manner to cause other models to leak information about the individual data sets they were trained on. It is proposed to extend this technique to Cyber defensive cases, in order to better understand and harden machine learning based network security solutions, such as Intrusion Prevention/Detection Systems (IPS/IDS). Utilizing a machine learning algorithm in an adversarial manner against a system already trained with a specific data set, it is possible to glean information on the original training set by manipulating inputs provided to the system and observing its reported outputs. It is the intention of this SBIR to evaluate the feasibility and commercial viability of techniques that could be easily adapted to test and evaluate the robustness of an already trained model, particularly one in which the internal classifier parameters are unknown. PHASE I: Evaluation of various machine learning networking security solutions and their implementations. An example is the open source project, Stratosphere. Evaluation of machine learning concepts, methods, and existing research applicable to this attack surface will aid in the eventual goal of an implementation of machine learning system concept(s) against a given IPS/IDS system to demonstrate manipulation of data inputs to generate specific responses from the classification system. PHASE II: Verification and validation of machine learning technique against additional IPS/IDS systems and surrogates. Enhancements to technique for real-time traffic manipulation to allow for dynamic triggers against an IPS/IDS in a specific manner. Demonstration of technique effectiveness from both inside and outside of a protected network. PHASE III DUAL USE APPLICATIONS: Extension of technique beyond network security. Potential commercialization options include, but are not limited to: - Use technique to validate robustness of machine learning algorithms to inference attacks - Technique applicability to keyword manipulation to guard against advanced tracking mechanisms to enhance security and privacy - Masking “honeypot” networks by manipulating traffic to appear already compromised - Utilizing technique to validate effectiveness of other classifiers’ ability to handle malicious or targeted junk data Military transition paths for network security applications of this concept include Product Manager elements and product lines within PEO IEW&S, PM EW&C. Elements of this SBIR would directly feed into established, planned, and already transitioned I2WD mission funded efforts relating to Cyber security, awareness, and understanding. Aspects of Phase III deliverables will support situational understanding and modeling of Cyber assets and defensive techniques. It is expected that, if successful, this SBIR will transition directly to elements within PM EW&C, as part of long-term and ongoing product line support. Commercially, a successful implementation of this SBIR in Phase III would aid in heightened Cyber defensive and penetration testing techniques, providing Internet Service Providers (ISPs), cloud-based architecture providers, and other Cyber security research organizations a robust validation method. Specific transition partners, operational use cases, and military applications are classified. Generic descriptions and high-level transition paths are provided to provide unclassified clarification as much as possible. REFERENCES: 1. R. Shokri, et al. "Membership Inference Attacks Against Machine Learning Models". 38th IEEE Symposium on Security and Privacy. 2017. 2. N. Carlini, D. Wagner. "Towards Evaluating the Robustness of Neural Networks". 38th IEEE Symposium on Security and Privacy. 2017. 3. Stratosphere IPS Project. Accessed June 7, 2017. [Online] https://stratosphereips.org/ 4. H. Yang, et al. “How to Learn Klingon Without Dictionary: Detection and Measurement of Black Keywords Used by Underground Economy”. 38th IEEE Symposium on Security and Privacy. 2017. 5. R. Sommer, V. Paxson. “Outside the Closed World: On Using Machine Learning for Network Intrusion Detection”. 2010 IEEE Symposium on Security and Privacy. 6. S. Mukkamala, et al. “Intrusion detection using neural networks and support vector machines”. Proceedings of the 2002 International Joint Conference on Neural Networks. 2002. 7. W. Lee, S. Stolfo. “Data Mining Approaches for Intrusion Detection”. 7th USENIX Security Symposium. 1998. 8. C. Tsai, et al. “Intrusion detection by machine learning: A review”. Expert Systems with Applications. Vol. 36, Iss. 10. pp. 11994-12000. December 2009. 9. D. Tsai, et al. “A hybrid intelligent intrusion detection system to recognize novel attacks”. IEEE 37th Annual 2003 International Carnahan Conference on Security Technology. 2003. KEYWORDS: Machine Learning, Cyber, Network Security, Intrusion Prevention System, IPS, Intrusion Detection System, IDS, Neural Networks, Behavioral Modeling
TECHNOLOGY AREA(S): Electronics OBJECTIVE: The objective of this topic is to develop network resource-aware methodologies for coordination and cooperation that will optimize the information flow for Army ad-hoc tactical networks operating in congested and contested environments. DESCRIPTION: The Army has a definite need to understand the degree to which ad-hoc wireless communications can still occur in congested and contested environments where power sources are short-lived, the spectrum is limited and where jamming (both friendly and adversarial) can occur spontaneously and without warning. The achieving of wireless communications may entail that the agents/nodes on the network cooperate and coordinate to an extent that will enable them to pass needed information to each other despite being embedded in a congested and contested environments. The degree of cooperation and coordination between these agents can be obtained by viewing the interactions between agents as an optimization problem, the solution of which will indicate what kinds of activities and just how much data exchange between the agents are necessary for passing needed information. The methodology to be developed will address the problem as one that will utilize coordination and cooperation in obtaining the solution. This is to be contrasted with one where the agents alone and in isolation decide the optimal course of action for operating effectively in a congested and contested environment. The methodology should make use of any of the techniques coming from the fields of machine learning and artificial intelligence but should not make use of techniques from game theory. This analysis should be done assuming a base frequency ranging from 400 MHz to 2.4 GHz. Free space communications is assumed. It will be assumed that there are N (friendly) network nodes enclosed in bounded region R of space in the environment and distributed uniformly in this region. Two other (adversarial) nodes will be assumed to be outside of R but transmitting a periodic signal of period T and power P at the same frequency as the friendly nodes. These two nodes will represent adversarial jamming nodes. Two of the friendly nodes are assumed to transmit at ten (10) different frequencies (to be chosen by the investigator) between 400 MHz and 2.4 GHz but at twice the power of the other friendly nodes. The duration and starting times of these signals will be chosen according to a deterministic pattern. These two nodes represent friendly jammers. It will be assumed also that all nodes have finite power sources that could be replenished either with batteries or with RF energy harvesting. The goal will be to develop a practical computable methodology that will: (i) Show explicitly how cooperation and coordination among agents can optimize network resources (spectrum, bandwidth, energy), and the processing power of the agents using algorithms that will scale with the number of agents. The exact notion of cooperation and coordination should be innovative and must not utilize those from the game theory literature. (ii) Using computational geometry or similar techniques, segment the environment to indicate which geographical regions can optimally support wireless communications. Each of these regions is to be ranked as to its effectiveness in supporting wireless communications using a quantitative risk measure. For each agent this risk measure must reflect the many reasons why the agent won’t be able to obtain useful information from other agents, such as congestion, unavailability of slots or routes, or signal interference. (iii) Indicate explicitly which network factors/covariates and their measurements are needed for cooperation and coordination and to what degree these factors contribute to optimal wireless communications in contested and congested environments. (iv) Estimate the amount of needed information that can be exchanged in a tactical network using coordination and cooperation. PHASE I: Explore and define a mathematical framework to capture the interactions between agents in a tactical network embedded in a contested and congested environment. Use this framework to show how cooperation and coordination between agents is to occur and show explicitly what network data is needed to accomplish this. Identify the geographical segmentation algorithm to be used and the risk measure(s) to be assigned to each geographical segment. PHASE II: Create simulation and/or analytical models to illustrate the optimality of the cooperation and coordination framework. Give examples of the models over real Army tactical networks. Develop software that can be implemented in a tactical network that will realize what was shown in these models. This software is to be written in a language that is implementable on an Army tactical communications platform. PHASE III DUAL USE APPLICATIONS: Demonstrate a radio system that is field-ready utilizing the software developed in Phase II, and demonstrate interoperability with other transceivers in a tactical network environment that would be used in all Army echelons. REFERENCES: 1. Additive consistency of risk measures and its application to risk-averse routing in networks, R. Cominetti and A. Turrico, arXiv: 1312.4193v1 [math.OC] 15 Dec 2013. 2. Cooperative learning in multi-agent systems from intermittent measurements, N. Leonard, A. Olshevsky, arXiv: 1209.2194v2 Sept 2013. 3. Learning of coordination: exploiting sparse interactions in multiagent systems, F. S. Melo and M. Veloso, Procs of 8th Int. Conf on Autonomous Agents and Multiagent Systems, 2009. 4. Collaborative multiagent reinforcement learning by payoff propagation, J. R. Kok and N. Vlassis, Journal of Machine Learning Research 7, 2006. 5. Collective decision-making in ideal networks: the speed-accuracy tradeoff, V. Srivastave, N. E Leonard, arXiv 1402.3634v1 Feb 2014. 6. The topology of wireless communication, E. Kantor, Z. Lotker, M. Parter, D. Peleg, arXiv 1103.4566v2 Mar 2011. 7. A review of properties and variations of Voronoi diagrams, A. Dorbin. 8. Risk Measures for the 21st Century, Giorgio Szego (Editor), Wiley; 1 edition 2004.
TECHNOLOGY AREA(S): Electronics The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), which controls the export and import of defense-related material and services. Offerors must disclose any proposed use of foreign nationals, their country of origin, and what tasks each would accomplish in the statement of work in accordance with section 5.4.c.(8) of the Announcement. OBJECTIVE: To develop and demonstrate technologies that are capable of detecting the lasers associated with Laser Beam Rider (LBR) Anti-Tank Guided Missiles (ATGMs) with high angular accuracy. DESCRIPTION: The ATGMs pose a significant threat to Army combat vehicles. With advanced sensors and countermeasures, ATGMs can be detected from significant distances, and countered prior to impacting their intended targets. The LBR ATGM launchers project an infrared laser field by which the missile is guided to the target. This laser energy can be detected to help warn vehicle crews, or aid other systems in locating the inbound missile. Unlike laser rangefinders and designators which use relatively higher powered lasers, LBR laser energy is much lower since it only requires a one-way guidance link from the launcher to the back of the missile in flight. The laser energy that impacts the ground vehicle is very low, making it both difficult to detect and find its origin. Current state-of-the-art, commercially available laser warning systems are capable of detecting LBR ATGMs and locating them within a quadrant, but do not provide the angular accuracy required to enable improved countermeasure solutions, evasive maneuvers, or line-of-bearing for counter fire. Fielded laser warning systems also typically have beam-rider sensitivity on the order of 5-10uW/cm^2, or are not optimally designed to detect beam-riders. Instead, they are designed optimally for rangefinder or laser designator detection, with a reduced capability to detect beam-riders. The goal of this Small Business Innovation Research is to develop and demonstrate sensor technologies capable of accurately detecting this laser energy at the maximum effective ranges of LBR ATGMs.
PHASE I: The goal of Phase 1 is to produce a conceptual design and breadboard suitable to demonstrate the component technology in a laboratory environment. Surrogate lasers may be used to emulate the laser of the actual missile systems in order to demonstrate the concept. PHASE II: The goal of Phase 2 is to produce a prototype component that could potentially be integrated with other military sensors for Army ground vehicles. This prototype will be demonstrated both in the laboratory and controlled field environments in order to show the laser detection capability at long ranges. The prototype may be tested against actual ATGM guidance lasers depending on their availability and security considerations. Required Phase 2 deliverables will include two major design reviews, technical documentation of the prototype, operator’s manual, the prototype hardware, and a final demonstration. PHASE III DUAL USE APPLICATIONS: The Phase III developed component will be a Technical Readiness Level 6, and is intended to transition to the Army Vehicle Protection Suite Program of Record. The product could be a stand-alone component integrated on to Army ground vehicles, but will more likely be integrated and packaged with other laser warning and threat warning sensors. Potential commercial applications for this sensor technology include laser surveying, ranging, and optical communications. REFERENCES: 1. Title : Laser Warning Receiver, Corporate Author : NATIONAL AIR INTELLIGENCE CENTER WRIGHT-PATTERSON AFB OH Personal Author(s) : Mei, Jin Full Text : http://www.dtic.mil/get-tr-doc/pdf?AD=ADA315183 2. Patents/Research Patent: Laser beam rider guidance system Publication Number: US 4111385 A 3. Patent: Panoramic laser warning receiver for determining angle of arrival of laser light based on intensity Publication number: US 9448107 B2 4. Market Potential: http://www.satprnews.com/2017/07/05/laser-warning-system-market-to-witness-comprehensive-growth-by-2025/ KEYWORDS: Ground combat vehicle, laser, Anti-Tank Guided Missile, ATGM, Laser Beam Rider, electronic warfare, infrared, sensor, detector, laser warning receiver
TECHNOLOGY AREA(S): Information Systems OBJECTIVE: The objective of this topic is to design and develop a joint MAC and routing protocol that supports tactical radios operating in full-duplex (FD) UHF/VHF band, and simultaneous electronic warfare (EW) and communications capability. The development of full-duplex radios is emerging for the unlicensed band and the current research has focused on development of the physical layer to provide point to point links and increase capacity. Taking advantage of the physical layer breakthroughs in full-duplex, which cancels self-interference, will entail the re-design of some of the upper layer protocols, such as scheduling and routing. DESCRIPTION: The development of full-duplex radios has a benefit for the Army in that it allows an increase in MANET network capacity and also simultaneous electronic warfare (EW) and communications. However, to get the maximum gain out of the unique characteristics of a full-duplex capability for future MANET wireless communication, it is important to design intelligent full-duplex scheduling and routing protocols. Current tactical radios cannot simultaneously transmit and receive on the same channel since the self- interference generated when transmitting is orders of magnitude stronger than the received signal. In addition, the scheduling and routing protocols are designed for half-duplex radios. Current breakthroughs at the physical layer such as circulators, analog circuitry, digital signal processing (DSP) techniques and the antenna technologies promise to provide 40-80db in self-interference cancellation. Its feasibility has been shown with off-the-shelf components [1, 2, 3]. However current scheduling designs for full duplex have been so far centralized in nature and geared towards hub and spoke network configurations. Some research has been done on cellular networks with the goal of implementing a full-duplex MAC protocol that builds on IEEE 802.11 [4] and in wireless networks where a novel MAC algorithm is developed that exploits self-interference cancellation and increases spatial re-use [5]. As this research indicates, some of the key challenges for full-duplex MAC development are to coordinate multiple simultaneous transmissions that respects the selection of FD transmission modes and nodes, the fairness among nodes, the hidden node problem, and the contention in asynchronous FD mode. Also, this research indicates that the full-duplex transmission in wireless MANET networks needs a direct coupling between the routing layer and the MAC layer in order to alleviate the cross-interference relationship between the links in the network. This cross-interference can make it very difficult to fully exploit the potential of efficient FD transmission. To this end, innovative research is required to develop a cross-layer framework that encapsulates both routing and a distributed MAC with power control that is implementable on FD wireless ad-hoc networks. The routing coupled with the MAC should have the capability that all links in any routing path can be activated simultaneously, effectively forming a cut-through route [8], wherein each node along the route can simultaneously receive a new packet from the upstream node and forward a previously received packet to its downstream node. In addition to this requirement, the approach should allow a collection of nodes from the network to have the capability of engaging in jamming and communications simultaneously. The resulting framework should take the form of a decision engine that allows this collection of nodes to be intelligently selected in order to optimize jamming and communications capability. All of these features should be scalable and show performance gains in throughput, latency, and packet loss. Yüklə 0,91 Mb. Dostları ilə paylaş:
|