Dariusz Mikulski

Distributed information-weighted Kalman consensus filter for sensor networks

Abstract Consensus-based algorithms for distributed Kalman filtering of the state of a dynamical target agent have attracted considerable research and attention during the past decade. In these filters, it is required for all agents to reach consensus about their estimates of the state of a target node. Distributed filtering techniques for sensor networks require less computation per sensor node and result in more robust estimation since they only use information from an agent’s neighbors in a network. However, poor local sensor node estimates caused by limited observability, network topologies that restrict allowable communications, and communication noises between sensors are challenging issues not yet fully resolved in the framework of distributed Kalman consensus filters. This paper confronts these issues by introducing a novel distributed information-weighted Kalman consensus filter (IKCF) algorithm for sensor networks in a continuous-time setting. It is formally proven using Lyapunov techniques that, using the new distributed IKCF, the estimates of all sensors reach converge to consensus values that give locally optimal estimates of the state of the target. A new measurement model is selected that only depends on local information available at each node based on the prescribed communication topology, wherein all the estimates of neighbor sensors are weighted by their inverse-covariance matrices. Locally optimal solutions are then derived for the proposed distributed IKCF considering channel noises in the consensus terms. Moreover, if the target has a nonzero control input, a method is giving of incorporating estimates of the target’s unknown input. Simulation case studies show that the proposed distributed IKCF outperforms other methods in the literature.

Trust method for multi-agent consensus

The consensus problem in multi-agent systems often assumes that all agents are equally trustworthy to seek agreement. But for multi-agent military applications - particularly those that deal with sensor fusion or multi-robot formation control - this assumption may create the potential for compromised network security or poor cooperative performance. As such, we present a trust-based solution for the discrete-time multi-agent consensus problem and prove its asymptotic convergence in strongly connected digraphs. The novelty of the paper is a new trust algorithm called RoboTrust, which is used to calculate trustworthiness in agents using observations and statistical inferences from various historical perspectives. The performance of RoboTrust is evaluated within the trust-based consensus protocol under different conditions of tolerance and confirmation.

Online adaptive learning for team strategies in multi-agent systems

During mission execution in military applications, the TRADOC Pamphlet 525-66 Battle Command and Battle Space Awareness capabilities prescribe expectations that networked teams will perform in a reliable manner under changing mission requirements and changing team and individual objectives. In this paper we first present an overall view for dynamical decision-making in teams, both cooperative and competitive. Strategies for team decision problems, including optimal control, N-player games (H∞ control, non-zero sum) and so on are normally solved offline by solving associated matrix equations such as the coupled Riccati equations or coupled Hamilton–Jacobi equations. However, using that approach, players cannot change their objectives online in real time without calling for a completely new offline solution for the new strategies. Therefore, in this paper we give a method for learning optimal team strategies online in real time as team dynamical play unfolds. In the linear quadratic regulator case, for instance, the method learns the coupled Riccati equations solution online without ever solving the coupled Riccati equations. This allows for truly dynamical team decisions where objective functions can change in real time and the system dynamics can be time-varying.

Trust dynamics in multi-agent coalition formation

We present a rigorous treatment of coalition formation based on trust interactions in multi-agent systems. Current literature on trust in multi-agent systems primarily deals with trust models and protocols of interaction in noncooperative scenarios. Here, we use cooperative game theory as the underlying mathematical framework to study the trust dynamics between agents as a result of their trust synergy and trust liability in cooperative coalitions. We rigorously justify the behaviors of agents for different classes of games, and discuss ways to exploit the formal properties of these games for specific applications, such as unmanned cooperative control.

Command and Control for Large-Scale Hybrid Warfare Systems

Emerging hybrid threats in large-scale warfare systems require networked teams to perform in a reliable manner under changing mission tactics and reconfiguration of mission tasks and force resources. In this paper, a formal Command and Control (C2) structure is presented that allows for computer-aided execution of the networked team decision-making process, real-time tactic selection, and reliable mission reconfiguration. A mathematically justified networked computing environment is provided called the Augmented Discrete Event Control (ADEC) framework. ADEC is portable and has the ability to provide logical connectivity among all team participants including mission commander, field commanders, war-fighters, and robotic platforms. The proposed C2 structure is developed and demonstrated on a simulation study involving Singapore Armed Forces team with three realistic symmetrical, asymmetrical, and hybrid attack missions. Extensive simulation results show that the tasks and resources of multiple missions are fairly sequenced, mission tactics are correctly selected, and missions and resources are reliably reconfigured in real time.

Trust-based coalition formation in multi-agent systems

In this paper, we provide a framework to study trust-based coalition formation in multi-agent systems using cooperative game theory as the underlying mathematical framework. We describe how to study trust dynamics between agents as a result of their trust synergy and trust liability in cooperative coalitions. We also rigorously justify the behaviors of agents for different classes of games and discuss how to exploit the formal properties of these games for cooperative control in an unmanned military vehicle convoy.

Stable structures of coalitions in competitive and altruistic military teams

In heterogeneous battlefield teams, the balance between team and individual objectives forms the basis for the internal topological structure of teams. The stability of team structure is studied by presenting a graphical coalitional game (GCG) with Positional Advantage (PA). PA is Shapley value strengthened by the Axioms of value. The notion of team and individual objectives is studied by defining altruistic and competitive contribution made by an individual; altruistic and competitive contributions made by an agent are components of its total or marginal contribution. Moreover, the paper examines dynamic team effects by defining three online sequential decision games based on marginal, competitive and altruistic contributions of the individuals towards team. The stable graphs under these sequential decision games are studied and found to be structurally connected, complete, or tree respectively.

Deep-learning trust estimation in multi-agent systems

In this paper we propose a trust algorithm, dubbed NeuroTrust, based on a multi-layered neural network. Previous work introduced trust as a performance estimation algorithm between team members in multi-agent systems, to allow for behavior optimization of the team. The trust model was developed based on an Acceptance Observation History (AOH) and confirmation and tolerance parameters to control trust growth and decay. Further work proposed certain improvements, in an autonomous vehicles convoy scenario, by considering agent diversity and a non-linear relationship between trust and vehicle control. In this work we show a further optimization using a deep recurrent neural network. This multi-layered neural network delivers trust as a probability function estimation with AOH as a sliding window batch input. The neural network is pre-trained using supervised learning, to emulate the previous trust model, as baseline. This pre-trained model is then exposed to future optimization using on-line reinforcement learning. The proposed trust model could be adaptable to a variety of systems, external conditions, and agent diversity. One application example where such a biologically-inspired trust model is suitable would be for soldier-machine teaming. Furthermore, particularly in the autonomous convoy scenario, we can account for the trust-control relationship nonlinearity in the trust domain, thus simplifying the control algorithm.

Incremental learning in trust-based vehicle control

In many multi-agent teams, entities fully trust their teammates and the information that they provide. But we know that this can be a false assumption in many cases, which can lead to sub-optimal performance of the team. In this paper, we build off of prior work in developing a simple model of estimating and responding to different levels of trust between team members. We have chosen to use a vehicle convoy application to generate data and test the operation of the trust estimation algorithm and its evolution. We build on prior work, where a cruise control algorithm to maintain following distance was implemented, as were algorithms to adjust follow distance based on trust in the leader and the capability for a lead vehicle to “look back” and adjust its speed based on the follow distance of the vehicle behind. In this paper we introduce a mechanism, based on trust, which negotiates between two follow behaviors, either follow the vehicle ahead or drive towards a set of fixed waypoints. We also add a nonlinear relationship between trust and follow distance to provide a knob to adjust convoy performance and the paper shows that it does adjust performance, somewhat as expected.

Special Issue: Modeling & Simulation for Cyber Security of Autonomous Vehicle Systems

The transportation and information gathering needs of our world are changing. Transportation systems are becoming more vital for our society and also more congested. Networking holds out the hope for more efficient use of multiple autonomous systems but also opens the door for increased risk exposure to failures and malicious attacks. Developments in internet technologies, online auctions and games, social networks, high-technology engineered systems with multiple dynamic components, and elsewhere motivate new advancements for decision, interaction, and control in cyberphysical systems (CPS) with autonomous dynamical subsystems. Examples of vehicle CPS can be found in areas as diverse as automobiles, air transportation, civil infrastructure, trains, and surface and subsurface water vehicles. Information Assurance must provide authentic, accurate, secure, reliable, and timely information to vehicle decision and control systems in order to achieve information dominance, mission performance, and survivability in risky changing environments. Computing and information processes must be carried out over distributed and heterogeneous networked systems for resilience to minimize single point of failure to malicious attacks. Recent advances in autonomous cyberphysical systems show promise for improved reconnaissance, surveillance, and security in foreign and domestic conflicts. However, CPS also present new and unevaluated sets of attack surfaces in communications and controls. As such, the modeling and simulation of cyber security in autonomous systems is a critical research area that will ultimately dictate the ways these systems should interact with people and each other. Thus, the goal of this Special Issue is collect a comprehensive body of descriptive and prescriptive approaches to modeling and simulating cyber security threats, defense techniques, and overall frameworks in highly dynamic and hostile environments. This special issue is composed of five papers that show the current state of modeling, simulation, decision, and control in some important areas of cyberphysical systems. The paper by Yagdereli, Gemci, and Aktas x presents a study on cyber security of unmanned vehicles, including cars, civilian unmanned aircraft, trains, and boats. Types of cyber-attacks are characterized, including passive and active attacks. Potential vulnerabilities of autonomous vehicles are detailed. Metrics for cybersecurity are proposed. A framework for mitigation strategies based on modeling and simulation is presented. The paper by Bergin presents a cyber-attack and simulation framework based on a live-virtual-constructive (LVC) environment. The focus is on wireless mobile network interfaces for military autonomous vehicles. LVC offers opportunities for training and assessment of humanin-the loop performance. Security threats and vulnerabilities are discussed, and attack techniques are outlined including denial of service, passive attack, and malicious agents. A cyber-attack and simulation framework is detailed including methods of data modeling, storage, and transfer. The paper by Cavagnaro and Tiller presents a framework for risk assessment of force allocation and countermeasures in autonomous vehicle air operations. A probability theory approach is used to define tipping points to quantify game outcomes that represent mission success. Probability distributions are presented for risk assessment, and random walks are employed to yield a distribution of outcome results. Based on these tools, the Tiller software wargame simulation Modern Air Power has been developed and is applied in this paper to force allocation and countermeasures based on game-changing capabilities.