Papers by Georgios Theodoropoulos
Future Generation Computer Systems, 2014
h i g h l i g h t s • We present an optical/electrical interconnect for large-scale Clusters/Data... more h i g h l i g h t s • We present an optical/electrical interconnect for large-scale Clusters/Datacenters. • The system uses reconfigurable optical planes to optimize network communications. • Low-cost electronic planes bypass traffic during the optical plane reconfiguration. • A DeBruijn-based topology ensures good tradeoff between radix and latency at scale. • Performance approaches a fully connected network for stable traffic at lower cost.
Concurrency and Computation: Practice and Experience, Jul 1, 2008
Large, experimental Multi-Agent System (MAS) simulations are highly demanding tasks, both computa... more Large, experimental Multi-Agent System (MAS) simulations are highly demanding tasks, both computationally and developmentally. Agent-toolkits provide reliable templates for the design of even the largest MAS simulations, without offering a solution to computational limitations. Conversely, Distributed Simulation architectures offer performance benefits but the introduction of parallel logic can complicate the design process significantly. The motivations of distribution are not limited to this question of processing power. True interoperation of sequential agentsimulation platforms would allow agents designed using different toolkits to transparently interact in common abstract domains. This paper discusses the design and implementation of a system capable of harnessing the computational power of a distributed simulation infrastructure with the design efficiency of an agent-toolkit. The system permits integration, through an HLA federation, of multiple instances of the Java-based lightweight-agent simulation toolkit RePast. Comparative results of the performance of the new system vs. sequential RePast are presented and conclusions are drawn on the challenges faced in creating such mappings in general.
arXiv (Cornell University), Jun 19, 2018
Graph embeddings have become a key and widely used technique within the field of graph mining, pr... more Graph embeddings have become a key and widely used technique within the field of graph mining, proving to be successful across a broad range of domains including social, citation, transportation and biological. Graph embedding techniques aim to automatically create a low-dimensional representation of a given graph, which captures key structural elements in the resulting embedding space. However, to date, there has been little work exploring exactly which topological structures are being learned in the embeddings process. In this paper, we investigate if graph embeddings are approximating something analogous with traditional vertex level graph features. If such a relationship can be found, it could be used to provide a theoretical insight into how graph embedding approaches function. We perform this investigation by predicting known topological features, using supervised and unsupervised methods, directly from the embedding space. If a mapping between the embeddings and topological features can be found, then we argue that the structural information encapsulated by the features is represented in the embedding space. To explore this, we present extensive experimental evaluation from five stateof-the-art unsupervised graph embedding techniques, across a range of empirical graph datasets, measuring a selection of topological features. We demonstrate that several topological features are indeed being approximated by the embedding space, allowing key insight into how graph embeddings create good representations.
Domain names can be traded via online market places and auctions. Speculation can be lucrative, w... more Domain names can be traded via online market places and auctions. Speculation can be lucrative, with high value transactions reaching into tens of millions of dollars. This paper proposes a framework for automated domain name appraisal and evaluates several formulations of the problem with real world data. A dynamic nonlinear valuation modelling process is defined using machine learning techniques. Attributes or value factors are derived from short domain name text strings as well as a variety of other contextual data able to be obtained online from open sources. A data set of 9.975 million domains is used for evaluation and results show that search engine query data is a primary driver of value but using extracted text features can facilitate higher performance in distinguishing high value domains particularly when used with ensemble learning.
J. Wirel. Mob. Networks Ubiquitous Comput. Dependable Appl., 2015
The insider threat problem is a significant and ever present issue faced by any organisation. Whi... more The insider threat problem is a significant and ever present issue faced by any organisation. While security mechanisms can be put in place to reduce the chances of external agents gaining access to a system, either to steal assets or alter records, the issue is more complex in tackling insider threat. If an employee already has legitimate access rights to a system, it is much more difficult to prevent them from carrying out inappropriate acts, as it is hard to determine whether the acts are part of their official work or indeed malicious. We present in this paper the concept of "Ben-ware": a beneficial software system that uses low-level data collection from employees' computers, along with Artificial Intelligence, to identify anomalous behaviour of an employee. By comparing each employee's activities against their own 'normal' profile, as well as against the organisational's norm, we can detect those that are significantly divergent, which might indicate malicious activities. Dealing with false positives is one of the main challenges here. Anomalous behaviour could indicate malicious activities (such as an employee trying to steal confidential information), but they could also be benign (for example, an employee is carrying out a workaround or taking a shortcut to complete their job). Therefore it is important to minimise the risk of false positives, and we do this by combining techniques from human factors, artificial intelligence, and risk analysis in our approach. Developed as a distributed system, Ben-ware has a three-tier architecture composed of (i) probes for data collection, (ii) intermediate nodes for data routing, and (iii) high level nodes for data analysis. The distributed nature of Ben-ware allows for near-real-time analysis of employees without the need for dedicated hardware or a significant impact on the existing infrastructure. This will enable Ben-ware to be deployed in situations where there are restrictions due to legacy and low-power resources, or in cases where the network connection may be intermittent or has a low bandwidth. We demonstrate the appropriateness of Ben-ware, both in its ability to detect potentially malicious acts and its lowimpact on the resources of the organisation, through a proof-of-concept system and a scenario based on synthetically generated user data.
The recent trend for Software as a Service and other types of cloud services is driving demand fo... more The recent trend for Software as a Service and other types of cloud services is driving demand for data centers of ever increasing scale. This will require the scaling up and scaling out of existing data center architectures and will ultimately need new architectures featuring new topologies for the data center networks that interconnect its constituent nodes. The preferred tool for the detailed evaluation of such designs is simulation, as it gives finer detail than analysis at lower cost than testbed evaluation. Realistic simulation times require that the simulation itself be capable of being scaled out, i.e., it should be amenable to parallelization. We describe a simulation framework, together with a methodology for partitioning the relevant simulation models to allow their parallel implementation, and demonstrate its validity by applying it to the simulation of a hybrid optical/electrical network architecture using a cluster of high-end servers. We report on results capturing the performance of our simulator and discuss how these are associated with the underlying hardware hosting the simulation.
Simulation Practice and Theory, Mar 1, 2000
Synchronous VLSI design is approaching a critical point, with clock distribution becoming an incr... more Synchronous VLSI design is approaching a critical point, with clock distribution becoming an increasingly costly and complicated issue and power consumption rapidly emerging as a major concern. Hence, recently, there has been a resurgence of interest in asynchronous digital design techniques as they promise to liberate VLSI systems from clock skew problems, oer the potential for low power and high performance and encourage a modular design philosophy which makes incremental technological migration a much easier task. This activity has revealed a need for modelling and simulation techniques suitable for the asynchronous design style. Contributing to the quest for modelling and simulation techniques suitable for asynchronous design, and motivated by the increasing debate regarding the potential of CSP for this purpose, this paper investigates the suitability of occam, a CSP-based programming language, for the modelling and simulation of complex asynchronous systems. A generic modelling framework is introduced and issues arising from the parallel semantics of CSP/occam when the latter is employed to perform simulation are addressed.
Page 1. A Continuous Matching Algorithm for Interest Management in Distributed Virtual Environmen... more Page 1. A Continuous Matching Algorithm for Interest Management in Distributed Virtual Environments Elvis S. Liu and Georgios K. Theodoropoulos School of Computer Science University of Birmingham {ESLiu, GKTheodoropoulos}@cs.bham.ac.uk ...
Abstract In many real time interactive simulations the problem of interest management is that of ... more Abstract In many real time interactive simulations the problem of interest management is that of filtering updates through application-specified range queries. This" associative'form of memory access is often implemented more simply as a reference memory access pattern ...
Traditionally Interest Management (IM) in distributed environments has been performed through a '... more Traditionally Interest Management (IM) in distributed environments has been performed through a 'top-down' expression of interest patterns in the model. For many models the interest patterns are complicated or highly dynamic and are not suitable for expression in this way. Top-down expression also forces the model to contain data about the network communication which should strictly be encapsulated by the infrastructure. We propose a system, based on a Distributed Shared Variable abstraction, capable of automatically deriving interest expressions at the infrastructure level in a 'bottom-up' manner. We demonstrate through a simulated testbed that for certain access patterns this dramatically improves the precision of the IM system without sacrificing transparency from a modeller perspective.
Page 1. An Approach for Parallel Interest Matching in Distributed Virtual Environments Elvis S. L... more Page 1. An Approach for Parallel Interest Matching in Distributed Virtual Environments Elvis S. Liu and Georgios K. Theodoropoulos School of Computer Science University of Birmingham {ESLiu, GKTheodoropoulos}@cs.bham.ac.uk ...
European Simulation Multiconference on Simulation, Jun 16, 1998
Despite the recent increase in the popularity of blockchain, the technology suffers from the tril... more Despite the recent increase in the popularity of blockchain, the technology suffers from the trilemma trade-off between security decentralisation and scalability prohibiting adoption, and limiting the efficiency and effectiveness of the induced system. Addressing the trilemma trade-off calls for dynamic management and configuration of the blockchain system. In particular, choosing an effective and efficient consensus protocol for balancing the trilemma trade-off when inducing the blockchain-based system is acknowledged to be a challenging problem given the dynamic and complex nature of the blockchain environment. DDDAS approaches are particularly suitable for this challenge, and in previous work, the authors presented a novel DDDAS-based blockchain architecture and demonstrated that it offers a promising approach for dynamically adjusting the parameters of a system and optimising for the trade-off. This paper presents a novel simulation tool that can support and satisfy the DDDAS requirements for a dynamically re-configurable blockchain system. The tool supports the simulation and the dynamic switching of consensus protocols, analysing their trilemma trade-off. The simulator design is modular and allows the implementation and analysis of a wide range of consensus protocols and their implementation scenarios, along with the ability for parallelization. The paper also presents a quantitative evaluation of the tool. CCS CONCEPTS • Computing methodologies → Simulation tools; • Computer systems organization → Peer-to-peer architectures.
arXiv (Cornell University), Aug 21, 2019
Graphs have become a crucial way to represent large, complex and often temporal datasets across a... more Graphs have become a crucial way to represent large, complex and often temporal datasets across a wide range of scientific disciplines. However, when graphs are used as input to machine learning models, this rich temporal information is frequently disregarded during the learning process, resulting in suboptimal performance on certain temporal inference tasks. To combat this, we introduce Temporal Neighbourhood Aggregation (TNA), a novel vertex representation model architecture designed to capture both topological and temporal information to directly predict future graph states. Our model exploits hierarchical recurrence at different depths within the graph to enable exploration of changes in temporal neighbourhoods, whilst requiring no additional features or labels to be present. The final vertex representations are created using variational sampling and are optimised to directly predict the next graph in the sequence. Our claims are supported by experimental evaluation on both real and synthetic benchmark datasets, where our approach demonstrates superior performance compared to competing methods, outperforming them at predicting new temporal edges by as much as 23% on real-world datasets, whilst also requiring fewer overall model parameters.
ACM Transactions on Modeling and Computer Simulation, Jun 18, 2019
Using computer simulation to analyze large-scale discrete event systems requires repeated executi... more Using computer simulation to analyze large-scale discrete event systems requires repeated executions with various scenarios or parameters. Such repeated executions can induce significant redundancy in event processing when the modification from a prior scenario to a new scenario is relatively minor, and when the altered scenario influences only a small part of the simulation. For example, in a city-scale traffic simulation, an altered scenario of blocking one junction may only affect a small part of the city for considerable length of time. However, traditional simulation approaches would still repeat the simulation for the whole city even when the changes are minor. In this article, we propose a new redundancy reduction technique for large-scale discrete event simulations, called exact-differential simulation, which simulates only the altered portions of scenarios and their influences in repeated executions while still achieving the same results as the re-execution of entire simulations. This article presents the main concepts of the exact-differential simulation, the design of its algorithm, and an approach to build an exact-differential simulation middleware that supports multiple applications of discrete event simulation. We also evaluate our approach by using two case studies, PHOLD benchmark and a traffic simulation of Tokyo.
Uploads
Papers by Georgios Theodoropoulos
In AD 1071, the Byzantine Emperor, Romanos IV Diogenes, set out from Constantinople for the eastern borders of his Empire with an army described as “more numerous than the sands of the sea”. His military campaign culminated in defeat by the Seljuk Sultan Alp Arslan at the Battle of Mantzikert. This defeat was to have profound consequences for both Byzantine and Turkish history and is still commemorated in the modern state of Turkey. Yet the historical sources for this campaign contain significant gaps and we know more about the political intrigues surrounding the emperor than we do about how the army moved and fed itself.
The ‘Medieval Warfare on the Grid’ project (2007-2011) was funded by an AHRC-EPSRC-Jisc e-Science grant and set out to use computer simulation to shed new light on the Mantzikert campaign. In this book we present the results of the project and demonstrate that computer simulation has an important role to play in the analysis of pre-modern military logistics. It can give new context to historical sources, present new options for the interpretation of past events and enable questions of greater complexity to be asked of historical military campaigns. It can also highlight the similarities that exist across time and space when armies need to be mobilised, moved and fed.