DSDS 21
Permanent URI for this collection
This collection showcases outputs from the seventh Defence and Security Doctoral Symposium, hosted at the Defence Academy of the United Kingdom, Shrivenham by Cranfield University in association with AWE.
It is the only UK conference to provide research students and early career researchers in defence and security with an opportunity to present their work to a sector-wide audience. Covering both technology and social sciences research, contributions include technical papers, a 3MT (three minute thesis) competition, digital images and posters. In addition, there are plenary talks from thought-leaders, as well as opportunities to interact with industry, publishers and other employers of defence and security researchers.
News
Smith, Mandy (2021). 2021 Defence and Security Doctoral Symposium (DSDS21) hosted by Cranfield University in conjunction with DSTL and AWE: Symposium outputs. Cranfield Online Research Data (CORD). Collection. https://doi.org/10.17862/cranfield.rd.c.5600316.v9
Browse
Browsing DSDS 21 by Issue Date
Now showing 1 - 20 of 25
Results Per Page
Sort Options
Item Open Access Defence and Security Doctoral Symposium Flyer 2021("Cranfield University", 2021-09-02T13:31:36Z) Cranfield UniversityFlyer promoting the Defence and Security Doctoral Symposium 2021.Item Open Access Evaluation of Cyberbullying using Optimized Multi-Stage ML Framework and NLP(Cranfield University, 2021-12-11T19:02:52Z) Ketsbaia, Lida; Issac, Biju; Chen, XiaominDue to the evolution of technology, online hate is increasing, more specifically in areas of social media amongst the general population. Online hate has become a phenomenon that destructively impacts individuals, with victims suffering long-lasting mental and psychological issues. Since cyberhate is conveyed as an ever-growing social problem, researchers have tried to tackle the matter. One of the main methods researchers have focused on is through the means of Machine Learning to help classify whether a piece of textual data can be identified as cyberbullying or not. Therefore, the purpose of the research is to employ a multi-stage optimized Machine Learning Framework that will look at using a combination of two data balancing methods (RUS and SMOTE), the feature selection method PCA as well as the bio-inspired metaheuristic optimization techniques PSO and GA. The framework applied increases the performance of the Machine Learning Classifier Logistic Regression to help detect instances of cyberbullying. Furthermore, the paper will show the potential of using various NLP methods such as RoBERTa, XLNet and DistilBERT to find the most suitable model to use within the textual analysis of cyberhate.Item Open Access Towards realistic atomistic models of nitrocellulose degradation processes(Cranfield University, 2022-01-10T17:31:41Z) Gibbon, CatrionaComputational chemical modelling has many uses, including in situations involving chemicals that are unstable and mechanisms and reactions are hard to detect. A specific case is the degradation of nitrocellulose, a high energy material. It is known that nitrocellulose breaks down via both hydrolysis and thermolysis, but experimentally it is hard to investigate further, due to expense. Computational chemistry can be used to find the energies related to these reactions, and the effect of surrounding solvents, this information can be used to find optimum conditions for stability and long-term storage. Molecular dynamics can be used to create models of large scale of nitrocellulose, which can then be used as bulk structure for the study of these reactions. A stepping stone for building nitrocellulose systems is to build cellulose systems, a naturally occurring chemical that has been well studied. Cellulose naturally occurs in many different forms, from highly crystalline structures, such as cellulose I𝛼 and I𝛽, to amorphous structures, both of which are important for a full understanding of cellulose. Using cellulose I𝛽 as a starting point, structures of cellulose have been created, and the crystallinity broken by running simulations at high temperatures to form a paracrystalline structure. Periodic boundary conditions have been used to approximate an infinite system. Once accurate structures of cellulose have been created these are nitrated to varying levels of nitration. There are six possible locations of nitration in a cellulose dimer, and the nitrocellulose structures will be created with random combinations of 4-5 nitrated groups per dimer, giving approximately 12 % nitrogen content by mass, a level of nitration that is representative of experimental nitrocellulose.Item Open Access Synthesis and Characterisation of Lithium Silicides(Cranfield University, 2022-01-10T17:46:09Z) Azad, AtiaThermal batteries are primary (non-rechargeable) batteries. To activate the battery, a pyrotechnic heat source melts the solid electrolyte to a molten salt at high temperature (typically around 500°C). The battery activation starts by a pyrotechnic source such as Fe/KClO4. Thermal batteries are made from a positive electrode material such as FeS2, a molten salt electrolyte such as LiCl:KCl and a negative electrode material. Li13Si4 is the preferred anode material for thermal batteries. The electrolyte is mixed with a binder material. MgO is a typical binder. The positive electrode material attracted attention with the aim of having a high capacity, a high voltage and good thermal stability. Previous work focused on new cathode materials and investigated battery discharge mechanisms. The negative electrode material is of interest because the high temperature structures and phase transitions have not been studied in the current literature. In this work, lithium-silicon phases were synthesised by a solid-state reaction between lithium metal and silicon powder inside evacuated quartz ampoules. The phases were characterised by powder neutron diffraction, carried out on the Polaris diffractometer at ISIS facility, Rutherford Appleton Laboratory, UK, differential scanning calorimetry (DSC) and magnetic measurements on the superconducting quantum interference device (SQUID). The lithium-silicon phases are remarkably stable at high temperature and remains crystalline, with phase transitions only occurring below room temperature.Item Open Access Doping with Tungsten - Bad for Humans, Great for Tanks(Cranfield University, 2022-01-11T15:44:28Z) Powell, DanielThe distribution of energy within an armour system is a fundamental principle. The joining of the hard disruptive outer layer and soft absorbing backing layer is often a source of weakness in armour systems, causing premature failure and thus defeat of the armour. Novel adhesive techniques to join these thin layers together could have a greater impact on the armour performance with minimal additional weight. Doping epoxy with tungsten powder is one such technique that is currently being considered.Item Open Access Visualising penetration in armour(Cranfield University, 2022-01-11T15:45:37Z) Powell, DanielThe human eye, CT scanning, X-ray radiography and Computer Aided Design (CAD) can all be used to visualise penetration in armour systems. This digital image demonstrates how these advanced techniques help us see a story that often seems non-existent to the human eye. X-ray radiography in the x and y axis paired with subsequent computer modelling allows the recreation of the crater and embedded material in 3D, enabling detailed analysis of the armour penetration. The volume of penetration can be measured with 95% accuracy from this CAD model when compared to the more expensive and complex, yet highly accurate, CT scanning. This demonstrates the reliability of this relatively simple and accessible methodology for quantifying damage in armour systems.Item Open Access Visualising penetration in armour(Cranfield University, 2022-01-11T15:50:02Z) Powell, DanielThe human eye, CT scanning, X-ray radiography and Computer Aided Design (CAD) can all be used to visualise penetration in armour systems. This digital image demonstrates how these advanced techniques help us see a story that often seems non-existent to the human eye. X-ray radiography in the x and y axis paired with subsequent computer modelling allows the recreation of the crater and embedded material in 3D, enabling detailed analysis of the armour penetration. The volume of penetration can be measured with 95% accuracy from this CAD model when compared to the more expensive and complex, yet highly accurate, CT scanning. This demonstrates the reliability of this relatively simple and accessible methodology for quantifying damage in armour systems.Item Open Access Assured Multi-Agent Reinforcement Learning for Safety-Critical Scenarios(Cranfield University, 2022-01-11T16:35:25Z) Riley, JoshuaMulti-agent reinforcement learning involves and facilitates a team of agents to solve complex decision-making problems in shared environments. This learning process is largely successful in many areas, but its inherently stochastic nature is problematic when applied to safety-critical domains.To solve this limitation, we propose our Multi-Agent Reinforcement Learning (AMARL), which uses a model checking technique called quantitative verification. Quantitative verification provides formal guarantees of agent compliance to safety, performance, and other non-functional requirements, while reinforcement learning occurs and after a policy has been learned.Our AMARL approach is demonstrated using three separate navigation domains, which contain patrolling problems. The multi-agent systems must learn to visit patrol points to satisfy mission objectives while limiting exposure to risky areas in these domains. Different reinforcement learning algorithms have been utilised within these domains: temporal difference learning, game theory, and direct policy search. The performance of these algorithms, while combined with our approach, are presented. Lastly, we demonstrate AMARL with differing system sizes in both homogeneous and heterogeneous multi-agent systems through our extensive experimentation. This experimentation shows that the use of AMARL leads to faster and more efficient performance than standard reinforcement learning and consistently meets safety requirements.Item Open Access Microstructure and Mechanical Properties of Inconel 718 and Inconel 625 Produced Through The Wire + Arc Additive Manufacturing Process(Cranfield University, 2022-01-12T12:26:44Z) James, WilliamIn developing the wire + arc additive manufacturing (WAAM) process for heat resistant alloys used in high-speed flight applications, structures were built from nickel-based superalloys Inconel 718 (IN718) and Inconel 625 (IN625). In this paper, wall structures were deposited in both superalloys, using a plasma transferred arc process. The microstructure was analysed optically and under SEM; both alloys were seen to be of typical dendritic structure with long columnar grains, with little variation between the alloys. The findings suggest that the structures included significant segregation of heavy metals, with potential Laves phases and δ-phases also found across the alloys, which showed significantly more segregation of Nb and Mo at the grain boundaries and inter-dendritic regions. The alloys also underwent room temperature mechanical testing, in addition to this IN625 specimens were tested after a solutionising and ageing treatment. Hardness measurements indicated that in general the WAAM process has the effect of increasing material hardness by approximately 10%, when compared to wrought alloy in a solutionised state. In IN625 the heat-treated specimens showed an increase in hardness of around 6%, when compared with its as-deposited condition. Elongation in IN625 showed much greater values. Overall, IN718 showed a greater strength with less elongation than IN625. A comparison between both alloys and their stated maximum UTS and YS values from literature revealed that WAAM built IN718 and IN625 in its as-deposited condition can achieve just over half the maximum achievable UTS, with no post-process treatment. The heat-treatment process tested in IN625 marginally reduced the gap in UTS performance by 3.5%.Item Open Access Home Made Explosives (HME), how to understand the threats and model the risk?(Cranfield University, 2022-01-12T12:46:55Z) Halleux, FrancisWithin the scope of countering the threat of Improvised Explosive Devices, HME constitute a major public security concern, requiring specific risk assessments for first responders such as Explosive Ordnance Disposal (EOD) teams. An initial literature review helped identify current HME threats worldwide, i.e. the synthesis of peroxide explosives and/or the mixing of highly energetic pyrotechnics. Following a first study on Urea Hydrogen Peroxide as a potential emerging threat, Flash Powder (FP) is now investigated, considering the ease of manufacture combined with the availability of its precursors. FP can also be directly extracted from well-known fireworks, which are still widely available despite strong regulation efforts. These materials are much more than “just pyro’s”. Because of their explosive output, flash compositions have become attractive and largely misused by experimenters, ranging from backyard hobbyists to insurgents. Flash powder has been studied in the literature, but mainly in the context of regular applications. Research data are still lacking when it comes to performance as an explosive. After initial chemical, physical, and thermal characterization, sensitivity and performance assessments will be executed, including free-field and underwater firings and heat flux measurements.Item Open Access CED3: Comparative Evaluation of DDoS Defences(Cranfield University, 2022-01-12T15:05:18Z) Otung, AndikanDDoS defence validation provides a way to capture the usefulness of defensive solutions to one of the most notorious Internet attacks of our generation. A uniform method of defence evaluation, enables not only the individual assessment of defences but, if well formed, offers a valuable mechanism by which different DDoS defences can be objectively analysed and commensurably compared. Success in this area would not only enable individual organizations to make better individual decisions on which defences to implement, but would facilitate inter-organizational collaboration for the improvement of national infrastructure. This paper presents CED3 (pronounced “Seed”), a DDoS defence evaluation methodology that enables objective comparison of DDoS defences. Instead of gauging a defence’s effectiveness by testing it under benchmark scenarios that are formed independently of the defence in question, CED3 starts with theoretical analysis that considers “metrics of distinction” to identify relevant tests. This approach enables CED3 to more completely evaluate defences by capturing both strengths and limitations. CED3 introduces the notion of “true effectiveness” in a defence’s evaluation, which encapsulates not just the performance of a defence under test but also the cost to an attacker necessary to overcome that defence. Lastly, CED3 provides a taxonomy-based defence-map, using which a defence’s scope and evaluation performance can be more clearly visualised. The CED3 methodology was applied to three notable defence schemes: capacity enlargement, Passport and TrustGuard and the formulated. Tests were performed via simulation, using the NS-3 software executing on a high performance computing cluster, and comprised of hundreds of thousands of CPU hours. The resulting comparative evaluation, discussion and conclusions are presented in this paper.Item Open Access Detecting LFM Parameters in Joint Communications and Radar Frequency Bands(Cranfield University, 2022-01-12T15:22:33Z) Zhang, KaiyuAs the traditional radar waveform, linear frequency modulation (LFM) is widely used in military applications to detect targets. Recently, civilian applications such as internet of vehicle and unmanned aerial vehicle also apply LFM waveform to sense the nearby surroundings information. However, this complicated environment usually contain other waveforms, which may adversely influence LFM signal. Thus, there has been increasing interest in using the same radio spectrum to enable the radar and communication signals to coexist. In this poster, we select the orthogonal frequency division multiplexing (OFDM) signal as the communication waveform and discuss how to detect LFM parameters under communication and radar spectrum sharing scenarios. Firstly, the traditional method, the discrete chirp Fourier transform (DCFT), is applied in this scenario to estimated LFM parameters. Secondly, the alternative approach, the Hough transform, is proposed by considering the intrinsic feature of OFDM receivers. Through simulations, we demonstrate the DCFT method and the use of the Hough transform to confirm that these can be identified to a high degree of accuracy.Item Open Access Boron Carbide-Silicon Carbide nanocomposites for next generation armour(Cranfield University, 2022-01-12T15:34:31Z) Payne, Henry (Harry)Modern warfare is developing rapidly; technology is becoming more advanced and the modern soldier requires more of it. With the increase in equipment there is an increase in weight which reduces mobility. Threats have also developed, in such environments greater protection can be required. Combining these two results shows a need to produce lighter armour with a higher level of protection. Boron Carbide (B4C) has shown great potential for use in armours. It has a high hardness and low density, but unfortunately it sometimes fails prematurely. This has been explained by a phase transformation involving polytype collapse. This research aims to mitigate structural breakdown by microstructural design. It is hypothesised that a composite containing nano grains of B4C and Silicon Carbide (SiC) can mitigate the polytype breakdown, whilst combining the low density of B4C with the ballistic reliability of SiC. Current work has focused on creating robust nanopowder production techniques. Silica particles, the precursors to SiC, have been successfully produced at 100 nm average size. This has been achieved by investigating various reaction ratios in order to determine the repeatability of the reaction and the suitability for industry. Further to this, successful analysis of B4C-SiC powders, produced using SiO2 nanoparticles has been completed. This work pushes forward the project into the final composite production phase with the aim of using nanoindentation to determine their suitability as ceramic armours.Item Open Access Identifying prototypical trust signals in open-source software libraries: A think aloud study(Cranfield University, 2022-01-13T09:51:30Z) Peace, RobertA huge (and increasing) amount of veracity neutral information is available to online information seekers in today’s digital world. However, knowing which information to trust is difficult for users because of the existence of disinformation. Consequently, understanding which information to trust can have serious security implications for users. As an example, Open-source software (OSS) libraries are a useful resource for both experienced and inexperienced coders. However, the open nature of the OSS libraries allows malicious actors to hide numerous types of harmful code within scripts. This has ramifications for users because malicious code can be difficult to detect. For instance, inexperienced users may not have the know how to detect harmful code. Whereas experienced users may not correctly evaluate the trustworthiness of the code due to time constraints. Consequently, it is important to understand the digital trust signals that are being utilised by users to make credibility judgements about code within OSS libraries. This poster presents research on prototypical digital trust signals using a think aloud methodology. We recruited computer science students from the two UK Universities. Participants were asked to think aloud their credibility judgements, whilst interacting with a prototypical OSS library. The collected data will be analysed to identify prototypical trust signals for OSS users. The potential implications of helping to secure OSS libraries will be discussed.Item Open Access Virtual Reality interactive training for maritime crews in response to terrorist attacks on cruise ships(Cranfield University, 2022-01-13T11:20:48Z) Robinson, Selina; Meenaghan, AmyDespite increasing concern from policymakers and maritime security experts regarding the risk of terrorist attacks on maritime vessels, research within this domain remains somewhat neglected. To date, such attacks have been mercifully rare. Nonetheless, the potential for significant casualties and loss of life, particularly when considering ‘soft’ targets such as cruise liners and passenger ferries, indicate the need for robust training schedules for ship masters and crew members.The current research describes an ongoing project that aims to design, test and validate a comprehensive training programme for cruise ship masters, security personnel and crew. Detailed reviews of existing emergency response measures, and focus groups with maritime security experts, practitioners, and commercial training providers have been used to inform the development of a training package using a simulated cruise ship bridge. In this paper, the pilot phases of the research will be described. The context for the proposed training package will be discussed, and the rationale for the use of virtual reality as a training tool will be evaluated. Initial trials designed to test the effectiveness of the training package in enabling trainees to achieve specific learning objectives will be described. Subsequent stages for the development of project will then be outlined. It is proposed that the successful validation of this virtual training tool will demonstrate the potential for adaptation to other modes of transport at risk of terrorist attack. Additionally, the package can be modified for use in related areas of risk (such as piracy) for a wide range of maritime vessels.Item Open Access Highly efficient conversion of laser energy to hard X-rays in high intensity laser-solid simulations(Cranfield University, 2022-01-13T11:21:10Z) Morris, StuartWe present simulations which predict significantly higher laser to X-ray efficiencies than those previously found in high intensity (1020-1022 Wcm-2) laser-solid simulations. The bremsstrahlung emission is shown to last for 10-100 ps, which is difficult to model with conventional particle-in-cell (PIC) codes. The importance of collective effects is also demonstrated, showing the limitations of Monte Carlo modelling in these systems. A new, open-source hybrid-PIC code with bremsstrahlung routines has been developed to model this X-ray production. Special boundary conditions are used to emulate complex electron refluxing behaviour, which has been characterised in 2D-PIC simulations. The peak X-ray efficiency was recorded in thick gold targets, with 7.4% conversion of laser energy into X-rays of energy 1 MeV or higher. The target size is shown to play a role in the conversion efficiency and angular distribution of emitted X-rays, and a simple analytic model is presented for estimating these efficiencies.Item Open Access Vibration Analysis of Non-Metallic Materials(Cranfield University, 2022-01-13T17:44:01Z) Sriraman, ShabarishA 3MT talk on developing a hybrid method, which combines experimental testing and simulation tuning, to characterise the dynamic vibration properties of materials. Topics discussed include: - Why we need a new characterisation method - How the method works - Capability and key innovation, including a new approach to damping and reducing estimations in simulations - Validation and resultsItem Open Access De-sovereignisation as an instrument of Russia’s coercive diplomacy in the context of post-Soviet de facto states. A comparative study of Moldova and Ukraine (2014–2021).(Cranfield University, 2022-01-13T17:44:17Z) Barbieri, JaroslavaThis project examines the de-sovereignisation processes occurring in Moldova and Ukraine as a result of Russian-instigated activities in the breakaway territories of the Pridnestrovian Moldavian Republic (PMR) and the Donetsk and Luhansk Peoples’ Republics (DNR/LNR). De-sovereignisation is defined as the process(es) whereby a state selectively facilitates state- and nation-building activities that interfere with another state’s ability to exercise control and authority over portions of its territory and population. By strengthening various attributes of sovereignty and statehood in de facto states (i.e. territories having de facto yet lacking de jure sovereignty), Russia automatically weakens those of their respective parent states, thereby making them more vulnerable to its influence. With this in mind, this thesis addresses the following research questions: How does Russia’s role in supporting the viability of post-Soviet de facto states affect reintegration prospects into their parent states? How does Russia’s involvement in the DNR/LNR compare to that in the PMR? To guide empirical research, this thesis relies on a refined version of coercive diplomacy theory, which implies the use of (verbal and non-verbal) threats and incentives to instigate behavioural change in a target state to the coercer’s advantage. Thus, this project looks at how Russian state and non-state actors have been instrumentally promoting state- and nation-building activities in the PMR and the DNR/LNR with the aim to pressurise their respective parent states into complying with Russia’s shifting demands. Critically, this project provides the first comparative study of such activities, operationalised into five domains: 1) political-diplomatic; 2) economic-financial; 3) military-security; 4) educational-cultural; 5) socio-legal. At the same time, this study distances itself from the conventional way of treating political and military leaders in de facto states as mere Russian proxies deprived of independent agency. Rather, it interrogates the extent to which local state- and nation-building initiatives are coordinated with Russia’s ruling elite and examines whether these may generate policy dilemmas for Russia. There has been a tendency to emphasise the causes of Russia’s military incursions into neighbouring countries, with little empirical analysis of the long-term consequences of such incursions on the ground. By contrast, this project acknowledges the complex and ‘messy’ realities of Russia-backed de facto states and uses apparent similarities as a starting point to examine in comparative terms idiosyncratic local developments and the ensuing challenges for reintegration prospects. This study relies on case-study analysis and draws on sources such as state documents and public statements issued by the de facto authorities and the Russian leadership, local newspaper releases, videos from state-controlled media outlets as well as semi-structured interviews with officials and experts from the countries under study.Item Open Access Converged Security: Building an Evidence-Based Road Map(Cranfield University, 2022-01-13T17:44:51Z) Boakes, EmmaIncidents involving cyber-physical systems are increasing , , and are predicted to escalate further in the coming years . Convergence involves the integration of security resources within an organisation; it has been advocated since the early 2000s as a way of helping understand and mitigate vulnerabilities in cyber-physical systems. There is, however, little empirical research exploring converged security, and no clear roadmap for organisations who want to adopt the approach. My research examines the experiences of organisations who have taken a converged approach to security and uses this to develop an evidence-based roadmap for others to follow. I have carried out three studies to understand the implementation of converged security from different perspectives. Firstly, a literature review and a series of pilot interviews with senior security staff helped to identify the scope of converged security and key factors that facilitate effective convergence. These factors were then used to construct a three-round Delphi study with 23 security professionals working within converged security functions across the world. This study validated 22 critical success factors for implementing converged security. The third study comprised fifteen email interviews with senior staff involved in the decision to converge across different organisations and industry sectors. The interviews examined the move to convergence, and the decisions around its implementation. Taken together these studies provide an evidence-base of the activities that organisations need to adopt when deciding how to implement converged security, such as: effective communication to sell the idea, achieve buy-in and support; having a common goal that aligns converged security with business strategy; and the importance of culture, relationships and respect in ensuring collaboration within and between security resources. Grounded in the real-world experiences of a range of security professionals across different geographies and industries it is apparent that while there is no standard approach to convergence there are commonalities across different implementations. The methodological contribution of this research lies in mixed qualitative methods used remotely. The substantive contribution is an evidence-based road map for the delivery of converged security.Item Open Access Getting a Whiff of Crime: The Chemistry Behind Detection of Illicit Substances by a New Sniffer Dog(Cranfield University, 2022-01-13T17:45:13Z) Hardy, IonaTo combat the constantly shifting threat related to the trafficking of illicit drugs and terrorism and meet Law Enforcement Agencies (LEAs) operational requirements, the exploitation of the latest scientific and technical advances is key. A valid alternative to the detection technologies currently used/explored by UK and International markets, is the CRIM-TRACK sniffer device, which allows detection of multiple illicit substances in a non-destructive, rapid, and accurate manner. The CRIM-TRACK sniffer includes a colourimetric sensor system (Microchip, Figure 1) consisting of chromic dyes reacting to vapour traces of illicit substances (Analytes).The aim of this PhD project is to further develop the CRIM-TRACK sniffer device’s colourimetric sensor through chemical analysis of the interaction between analyte and dye molecules. Responsive (Colour changing) and non-responsive dyes from the current colourimetric microchip for the analyte of interest are determined through detection experiments using the CRIM-TRACK sniffer. The detection mechanisms of four analytes – a selection of chemical precursors and controlled drugs – with two families of dyes, diazobenzenes and sulfonephthaleins, have been ascertained. This is achieved by applying analytical techniques such as Nuclear Magnetic Resonance and UV-Vis Spectrophotometry and comparing the data produced by analyte:dye mixtures and their single components. Evaluation of these mixtures is aided by comparison of experimental data against modelling data produced by colleagues at Sheffield Hallam University. Understanding the detection mechanisms allows the identification and selection of other responsive dyes for improvement of the current microchip and the design of scenario specific microchips.