| Security has always been a stepchild to the developers of embedded systems. In the past they were able to rely on physical protection. Due to the connectivity and ubiquity of today's embedded systems this is no longer possible. The aim of SecFutur is to develop and establish a security engineering process for embedded systems. In order to achieve this overall goal SecFutur will provide a set of implemented resource-efficient security building blocks for embedded systems, each addressing a specific complex non-functional requirement, and a security engineering framework that supports the developer in integrating these building blocks into the overall engineering process. SecFutur targets the developer of embedded systems who by using the project results will be able to follow an application driven security engineering approach and increase the overall security of the system. Practical scenarios from several security-relevant application areas will be used in SecFutur to evaluate and demonstrate the advances towards secure resource-efficient embedded systems. |
This project addresses reliable delivery in partition-tolerant and
resource-constrained networks in three directions:
|
| The goal of this work is to study the issues associated with the deployment of Hastily formed networks (HFN) in a disaster management scenario with heterogeneous actor groups and diversified communication and leadership/cultural traditions. The vehicle for the study will be a technical platform in the form of a simulation environment in which algorithms for reliable and secure communication will be developed and tested. The organisation and interaction aspects will be studied with the aim of enhancing interoperability among the diverse end users, and establishing a conversation space. |
| This project is based on a long term industrial research problem for managing networks of cooperating UAVs connected via adhoc networks. An example application for such networks is time-constrained reconnaisance missions. One of the first issues that distinguish such networked UAVs from other MANETs in the literature are their specific mobility patterns. The project will start by studying suitable mobility models for UAV networks and goes on to model services delivered to mobile ground stations in presence of constraints. The considered constraints will range over safety-related, security-induced, and resource constraints leading to QoS requirements (bandwith optimisation, latency, etc). The work is carried out by an industrial PhD student that is employed at Saab Aerospace. |
| This project is part of the current national aerospace research program (NFFP4-S4207), with a focus on distributed modular avionics (DMA) architectures and their incremental verification. The project is carried out in cooperation with Chalmers university (Dependable Real-time Systems group), Saab Aerospace (Linköping) and Saab avitronics (Jönköping). Our part of the project concerns assurance of system dependability by focusing on faults within components and the propagation of their effects to other components. We aim to adapt and apply techniques for compositional formal verification of component-based systems to DMA architectures. The project also has a demonstration phase in which developed processes for certification of COTS increments to an existing avionic subsystem will be illustrated. |
The goal of this project is to establish an engineering
discipline for systematic development of component-based software for safety-
critical embedded systems. The main innovation of SAVE is the interdisciplinary
combination of architectural and component based design with analysis and
verification, in the specific context of safety and real-time. The main
challenges in component-based development of safety-critical applications
are to handle the multitude of conflicting requirements, including safety
vs. cost and time-to-market. SAVE will address the above by developing
a general framework for component-based development of safety-critical
vehicular systems, including
|
|
The goal of this project is to develop platforms for real-time data services that involve techniques for managing unpredictability of the environment, handling imprecise or incomplete knowledge of the workload, reacting to overloads and unexpected failures (i.e., those not expressed by design-time failure assumptions), to achieve the performance requirements and temporal behavior necessary for accomplishing the specified tasks. To address this problem, we aim at developing a management framework for real-time data services that provides guarantees on QoS and QoD in terms of several fundamental performance metrics for real-time applications. In order to provide guarantees, the system must be able to adapt to changes to the external environment, e.g. applied workload and/or execution time estimation errors. Initial research shows that feedback control scheduling is able to deal with dynamic systems that are both resource insufficient and that exhibit unpredictable workloads. Feedback control real-time scheduling defines error in terms for system behavior, i.e. the difference between desired and actual system behavior. The error is continuously monitored, and the system is adjusted to maintain desired performance. Our research focuses on using feedback control scheduling as initial research has been promising in providing satisfactory QoS and QoD guarantees. |
| This project aims to develop distributed algorithms for adaptive anomaly detection in resource-constrainted adhoc networks. In particular, we study how detection of attacks and intrusions can be performed with little overhead in terms of computational resources, but utilising the knowledge about the normal behaviour of user applications in terms of network resource consumption, QoS parameters, amd their adaptation to dynamic changes in the network. This work complements emerging trust models for adhoc and P2P networks. Its applications are envisaged where the adhoc solutions are used as an ingredient to bridge fixed infrastructure networks during failures or overloads, or simply to extend the reach of existing critical infrastructures. |
| This project is conducted under the Sixth Framework Programme of the European Community and is a Specific Targeted Research and Innovation Project (STREP). The goal of the project is to provide a concept for optimizing dependability in distributed component based systems. It comprises an architecture, well defined technology integration rules, metrics for evaluation, as well as prototype implementations. The aim is not to obtain a new middleware, but integrate and reuse existing off-the-shelf products (as COM, CORBA, EJB). The idea is to use replication for transparent fault tolerance and persistence, the focus being on the trade-off between consistency and availability. For this a combination of synchronous and asynchronous replication is used. The studies are oriented towards measuring the trade-off and configuring it to allow an application-specific optimum of availability. The project also aims at the design and implementation of a fault-tolerant naming service adapted to the context of availability-consistency trading. |
|
This research project focuses on the data management in software of real-time and embedded systems. This is an important class of systems since they are used in different areas. Distinguishing features of such systems are that the amount of data items in such applications can be high with freshness requirements on the data items and at the same time have deadlines on calculations. By using a central repository for data management, one can avoid unnecessary storing of data at different processes, which enhances software maintainability and fosters better software evolution due to the simpler structure and the removal of data subscription models. This also simplifies the programmers' tasks since large parts of data synchronization can be performed by the database, and that time constraints, such as data validity, can be enforced by the database. Current focus of this research project is on maintaining data freshness such that transactions in a database system use fresh data and the required load imposed by making data fresh are state dependent. This means that resources can be better utilized instead of designing the data management for the worst-case scenario. Freed CPU resources can for instance be used for enhanced diagnosis of the system. |
| This project aims to combine architectures and algorithms for adaptive management of resource allocation in networked applications. In particular, we study how user requirements in terms of Quality of Service (QoS) can be met by system level resource allocation, when there are several types of uncertainties present. We study algorithms for resource allocation both locally (in one node) and at a global level. The project started by studying the load control problem for radio network controllers in 3rd Generation mobile telecom, and will continue by studying the general problem in multi-resource multi-criteria settings. We adopt a combination of techniques from the areas of real-time systems, artificial intelligence, and control theory. The research is conducted in cooperation with researchers from UNCC who are supported by a three year travel grant from NSF. |
| This project studies the application of fault-tolerant techniques in distributed systems. We study formal models of fault-tolerance for achieving safety (in safety-critical systems) and availability (in telecommunication applications). The work focuses on replication techniques in general, and group services in the case of software intensive monitoring systems, in particular. In this project we will study the impact of other typical demands on a system, e.g. real-time demands and dynamic resource allocation in the achieved levels of fault-tolerance. To this end, trade-off studies are performed to study various approaches for implementing fault-tolerance in the middleware in presence of real-time and resource utilisation requirements. |
| This project is a multi-disciplinary effort to promote improvements in the system engineering process, in particular systems with heterogenenous components ranging from mechanical, hydarulic and electromechanical elements to software and electronics. The project is organised in a number of tracks including the following three at RTSLAB: |
The goal of this research is to bridge the gap between embedded systems, real-time systems and database systems, with a particular focus on the software development tools. Significant amount of research has focused on how to incorporate database functionality into real-time systems without jeopardizing timeliness and how to incorporate real-time behavior into embedded systems. However, research for embedded databases used in embedded real-time systems, that explicitly address (i) the development and design process, and (ii) the limited amount of resources in embedded systems is sparse. This type of research inherits the challenges from component-based software engineering, embedded systems and real-time systems. Further, this research explicitly addresses system resource demand for the system in the design of the embedded database in order to minimize system resource usage. At a high level, the goal is to build an experimental research platform for building
embedded databases for embedded real-time systems. At a high-level, the platform consists of two parts. First, we intend to develop a component library, which holds a set of methods, that can be used when building an embedded database. Initially, we will develop a set components that deal with concurrency control, scheduling, main-memory techniques. At the next step, we
develop tools that, based on the application requirements, will support the designer when building an embedded database using these components. More importantly, we want to develop application tools and techniques that:
|
| Research on monolithic real-time systems has resulted in many effective technologies for scheduling and resource management. Multi-processor realisation of real-time systems extend the same technologies, where the end-to-end performance requirements of a system are studied as a central problem via analysis based on parameters for each processor. In this project we propose to study how real-time techniques can be incorporated into open distributed system architectures at design stage, in particular in development platforms such as CORBA. The goal is to facilitate trade-off studies at early design stage, and support component based software development. Examples of interesting trade-offs such as real-time/security and real-time/fault-tolerance are considered. |
| This project is part of the current national aerospace research program (NFFP 3), with a focus on system safety and reliability in integrated hardware/software real-time systems. The project is organised in cooperation with the department of electrical engineering (Vehicular systems) and Saab AB. Our part of the project concerns improving system dependability by studies of how faults that lead to system failures can be avoided. At the early design stage we consider the use of langauges with formal sematics for specification of system components (in particular reconfigurable components, FPGAs), and the potential for formal verification and automatic code generation techniques. We further study how these techniques can be combined with existing methods for safety analysis (FTA, FMEA) and how consideration of random failures (e.g. radiation) affect the design and verification process. |
| Interactive simulation environments are considered one of the current promising emerging technologies. We are creating autonomous agents as football players for RoboCup and rescue workers for the RoboCup Rescue simulation environments. This research focuses on the system for specifying the behavior of the agents, the mechanisms and abstractions required for controlling their dynamic behavior, end-user programming of the actors, and the interaction between the agent simulator and the dynamic environment simulator. The work is being conducted in a realistic setting by integrating a prototype of the actors with the RoboCup soccer server and the RoboCup Rescue simulation environment. See also: |
SAFEGUARD aims to enhance the dependability and survivability of Large Complex Critical Infrastructures (LCCIs), such as distributed electric and telecommunication networks. Modern automation systems underlying LCCIs include different levels of automation, regulation, and control, but "intelligent" functions relating to critical issues such as safety and system survivability are usually monitored or executed by human operators. We are primarily interested in improving dependability and survivability of large infrastructures by:
|
| The project is directed towards the development of an interface standard which will allow the complete set of tools in design of airplanes and spacecrafts and their avionics systems (hardware, software, mechanical design, and implementation, project management, etc.) to communicate in such a way that an integrated project support environment is available from commercial tools. The interface standard will be possibly based on the STEP interface used in CAD applications, but will be extended to add the semantics appropriate to the aircraft industry. |
| The project is directed towards the development of an interface standard which will allow the complete set of tools in design of airplanes and spacecrafts and their avionics systems (hardware, software, mechanical design, and implementation, project management, etc.) to communicate in such a way that an integrated project support environment is available from commercial tools. The interface standard will be possibly based on the STEP interface used in CAD applications, but will be extended to add the semantics appropriate to the aircraft industry. |
|
This project is directed at the development of a data exchange standard which will allow the complete set of tools used in analysis and design of complex heterogeneous systems such as air- and spacecrafts and their avionics systems (hardware, software, mechanical design, and implementation, project management, etc.) to communicate in such a way that an integrated project support environment can be implemented from commercial tools. The proposed interface standard is developed within the STEP (ISO 10303) framework and is assigned the identifier AP-233. The aim of the project is to drive the standardisation within STEP and to validate the proposed standard through tool interface development and real data exchanges. The role of RTSLAB in this project is to co-ordinate and harmonize the development of the standard. The SEDRES-2 project is funded by the CEC in the IST programme and is a successor to the highly successful SEDRES project (see SEDRES homepage). |
| The project is concerned with improved programming environment for the family of synchronous languages (Lustre, Signal, Esterel, and a version of Statecharts). The work in the project is organised in several work packages: combination of imperative and declarative synchronous paradigms, program verification, code distribution and multi-tasking, integrating synchrony and asynchrony, connection with hardware/software codesign, and integration of analog/discrete synchronous design. Our contributions are mainly in the last work package, dealing wit hybrid systems incorporating synchronous controllers. |
| Interactive simulation environments are considered one of the current promising emerging technologies. We are creating autonomous agents as pilots for aircraft simulation environments. This research focuses on the system for specifying the behavior of the agents, the mechanisms and abstractions required for controlling their dynamic behavior, end-user programming of the simulated pilots, and the interaction between the agent simulator and the dynamic environment simulator. The work is being conducted in a realistic setting by integrating a prototype of the simulated pilots with TACSI - the tactical aircraft simulator produced by Saab AB. See also the EASE agent development environment homepage and CSIRO RoboCup Applications World-wide. |
| The co-design procedure for embedded systems starts from a high level description and applies partitioning algorithms in order to realise the final design as a range of hardware and software components. Common languages for design at the system level includes VHDL and C. In this project we propose enhancing the co-design procedure by adding a formal verification capability to capture early design errors prior to partitioning. For this purpose, formal languages which capture both data-flow and control-flow are explored and the integration of formal verification tools with a co-design environment is proposed. |
| This project is part of the European Agents and Middleware (AMI) initiative within the IST program. The goal of the project is to provide insights on the cost of incorporating generic fault-tolerance mechanisms in a middleware. The idea is to relieve the application writer in a distributed (multi-tier client server system) from writing the code that deals with fault tolerance when a server crashes. Instead, provide support in the middleware so that at the conception of the application, the appropriate replication mechanism (warm/cold passive, active), and the required parameter settings can be supported by automatic code generation within the middleware. We have chosen to study CORBA as a generic middleware and to study which performance, code size, state size, and other trade-offs exist in extending the CORBA infrastructure in compliance with the recently proposed FT-CORBA specification. Part of the trade-off study is the overhead (time) when there are no failures and the time taken for fail-over. |
Ericsson Radio Systems