Electrical and Electronic Engineering - Research Publications

Permanent URI for this collection

Search Results

Now showing 1 - 10 of 15
  • Item
    No Preview Available
    A Group Formation Game for Local Anomaly Detection
    Ye, Z ; Alpcan, T ; Leckie, C (IEEE, 2023-01-01)
  • Item
    No Preview Available
    Online Trajectory Anomaly Detection Based on Intention Orientation
    Wang, C ; Erfani, S ; Alpcan, T ; Leckie, C (IEEE, 2023-01-01)
  • Item
    No Preview Available
    Robust Wireless Network Anomaly Detection with Collaborative Adversarial Autoencoders
    Katzef, M ; Cullen, AC ; Alpcan, T ; Leckie, C (IEEE, 2023-01-01)
  • Item
    No Preview Available
    Wireless Network Simulation to Create Machine Learning Benchmark Data
    Katzef, M ; Cullen, AC ; Alpcan, T ; Leckie, C ; Kopacz, J (IEEE, 2022-01-01)
  • Item
    No Preview Available
    Local Intrinsic Dimensionality Signals Adversarial Perturbations
    Weerasinghe, S ; Abraham, T ; Alpcan, T ; Erfani, SM ; Leckie, C ; Rubinstein, BIP (IEEE, 2022)
  • Item
    No Preview Available
  • Item
    Thumbnail Image
    Detection of Anomalous Communications with SDRs and Unsupervised Adversarial Learning
    Weerasinghe, S ; Erfani, SM ; Alpcan, T ; Leckie, C ; Riddle, J ; Cherkaoui, S ; Andersson, K ; AlTurjman, F (IEEE, 2019-02-08)
    Software-defined radios (SDRs) with substantial cognitive (computing) and networking capabilities provide an opportunity for observing radio communications in an area and potentially identifying malicious rogue agents. Assuming a prevalence of encryption methods, a cognitive network of such SDRs can be used as a low-cost and flexible scanner/sensor array for distributed detection of anomalous communications by focusing on their statistical characteristics. Identifying rogue agents based on their wireless communications patterns is not a trivial task, especially when they deliberately try to mask their activities. We address this problem using a novel framework that utilizes adversarial learning, non-linear data transformations to minimize the rogue agent's attempts at masking their activities, and game theory to predict the behavior of rogue agents and take the necessary countermeasures.
  • Item
    Thumbnail Image
    A game-theoretic analysis of the adversarial boyd-kuramoto model
    Demazy, A ; Kalloniatis, A ; Alpcan, T ; Bushnell, L ; Poovendran, R ; Basar, T (SpringerLink, 2018-01-01)
    The “Boyd” model, also known as the “OODA loop”, represents the cyclic decision processes of individuals and organisations in a variety of adversarial situations. Combined with the Kuramoto model, which provides a mathematical foundation for describing the behaviour of a set of coupled or networked oscillators, the Boyd-Kuramoto model captures strategic (cyclic) decision making in competitive environments. This paper presents a novel game-theoretic approach to the Boyd-Kuramoto dynamical model in complex and networked systems. A two-player, Red versus Blue, strategic (non-cooperative) game is defined to describe the competitive interactions and individual decision cycles of Red and Blue agent populations. We study the model analytically in the regime of near phase synchrony where linearisation approximations are possible. We find that we can solve for the Nash equilibrium of the game in closed form, and that it only depends on the parameters defining the fixed point of the dynamical system. A detailed numerical analysis of the finite version of the game investigates the behaviour of the underlying networked Kuramoto oscillators and yields a unique, dominant Nash equilibrium solution. The obtained Nash equilibrium is further studied analytically in a region where the underlying Boyd-Kuramoto dynamics are stable. The result suggests that only the fixed point of the dynamical system plays a role, consist with the analytical solution. Finally, the impact of other variations of the Boyd-Kuramoto parameters on the game outcomes are studied numerically, confirming the observations from fixed point approaches. It is observed that many parameters of the Kuramoto model affect the NE solution of the current game formulation much less than initially stipulated, arguably due to the time-scale separation between the underlying Kuramoto model and the static game formulation.
  • Item
    Thumbnail Image
    Information Constrained and Finite-Time Distributed Optimisation Algorithms
    Philip, BV ; Alpcan, T ; Jin, J ; Palaniswami, M (IEEE, 2017)
    This paper studies the delay-accuracy trade-off for an unconstrained quadratic Network Utility Maximization (NUM) problem, which is solved by a distributed, consensus based, constant step-size, gradient-descent algorithm. Information theoretic tools such as entropy power inequality are used to analyse the convergence rate of the algorithm under quantised inter-agent communication. A finite-time distributed algorithm is proposed to solve the problem under synchronised message passing. For a system with N agents, the algorithm reaches any desired accuracy within 2N iterations, by adjusting the step-size, α. However, if N is quite large or if the agents are constrained by their memory or computational capacities, asymptotic convergence algorithms are preferred to arrive within a permissible neighbourhood of the optimal solution. The analytical tools and algorithms developed shed light to delay-accuracy trade-off required for many real-time IoT applications, e.g., smart traffic control and smart grid. As an illustrative example, we use our algorithm to implement an intersection management application, where distributed computation and communication capabilities of smart vehicles and road side units increase the efficiency of an intersection.
  • Item
    Thumbnail Image
    Long-Term Stochastic Planning in Electricity Markets Under Carbon Cap Constraint: A Bayesian Game Approach
    Masoumzadeh, A ; Nekouei, E ; Alpcan, T (IEEE, 2016-01-01)
    Carbon price in an electricity market provides incentives for carbon emission abatement and renewable generation technologies. Policies constraining or penalizing carbon emissions can significantly impact the capacity planning decisions of both fossil-fueled and renewable generators. Uncertainties due to intermittency of various renewable generators can also affect the carbon emission policies. This paper proposes a Cournot-based long-term capacity expansion model taking into account carbon cap constraint for a partly concentrated electricity market dealing with stochastic renewables using a Bayesian game. The stochastic game is formulated as a centralized convex optimization problem and solved to obtain a Bayes-Nash Equilibrium (Bayes-NE) point. The stochastic nature of a generic electricity market is illustrated with a set of scenarios for wind availability, in which three generation firms (coal, gas, and wind) decide on their generation and long-term capacity investment strategies. Carbon price is derived as the dual variable of the carbon cap constraint. Embedding the carbon cap constraint in the game indicates more investment on renewable generators and less on fossil-fueled power plants. However, the higher level of intermittency from renewable generation leads to a higher carbon price to meet the cap constraint. This paves the way towards storage technologies and diversification of distributed generation as means to encounter intermittency in renewable generation.