This is a list of all currently active projects within ACCESS. Click on project title to view detailed information. You can also edit the information in a detailed view.
Online social networks like Facebook or MySpace, where people interact with their real-life and their Internet friends via dedicated websites, have become extremely popular. Yet they have some limitations to be overcome for a next generation of social networks: privacy concerns and requirements of Internet connectivity, both of which are due to the services being implemented as web-based applications on a central site whose owner has access to all the data. A central provider can dictate the terms of service, use data mining and targeted advertisement, and disclose user information to others.
To overcome these limitations, we envision a paradigm shift from client-server to a peer-to-peer infrastructure coupled with encryption so that users keep control of their data and can use the social network also locally, without Internet access. Extending the distributed approach by direct data exchange between user devices removes the strict connectivity requirements of web-based OSNs, while encryption and access control mechanisms ensure privacy protection.
To verify the feasibility of this approach, we model user and adversary behavior, develop mechanisms for distributed control, storage, opportunistic networking, and security and evaluate them by security analysis, simulation, emulation, and test-bed implementation. The insights gained can then be generalized and re-used, extending the field of computer networking to enable community-driven privacy-preserving communications.
This project aims at investigating important theories and design approaches of rateless codes for wireless networks. Rateless codes (also called fountain codes) are originally proposed for disseminating large bulk of data in computer networks, where channels are modeled as packet-erasure channels, and physical-level issues are generally not considered. Hence, there are many new design challenges for applying rateless codes to wireless networks. Among them, one of the main design issues for wireless networks is multipath fading, which are efficiently addressed by techniques such as multiple antenna (MIMO), efficient modulation, and cooperative communication. Optimal combination of these techniques and rateless codes is mostly unknown. Further, many wireless applications are very sensitive to delay performance, e.g., for real-time multimedia transmission. Yet, most current rateless codes are optimal only for very large information length. To design efficient rateless codes with finite source information length is quite valuable, especially jointly with the physical-level issues.
In this project, we shall study the rateless codes for fading and noisy channels. Thus, efficient approaches for combating fading, e.g., multiple antenna, and cooperative communication, differential modulation will be jointly considered with rateless codes. Both theoretical foundation and design principles for practical systems shall be developed. We shall also investigate the efficient design approaches for rateless codes with finite information length. Further, the combination of network coding and rateless codes will be investigated, because of their intriguing connection. Through these research activities, in a longer perspective, we are seeking to form a theoretical foundation and practical design principles for the development of robust and flexible infrastructure needed for a new generation of wireless applications.
Wireless Sensor Networks (WSNs) are ad hoc networks of tiny, autonomous devices equipped with wireless communication and sensing capabilities for communication, control
and sensing purposes. They are a topic of active research in different backgrounds, ranging from electronics, to communications, control theory and computer science.
WSNs are characterized by the scarcity of resources in terms of communication, computation and energy supply. They are now evolving from simple sensing networks to
functionally rich spatial-temporal distributed systems, where in addition to the already complex requirements of communication services, control and actuation applications are
being added. The need of achieving maximum efficiency from these systems motivates the careful integration of functionalities traditionally considered to be separated, as
communication, control, and computation. A close interaction of research from these different areas is required. Motivated by our successful preliminary results, with this
research project we hope to contribute to the theoretical advancements of the emerging theory of in-network computation in WSNs.
In network computation investigates the following issues. Consider a wireless network were some nodes are sensing physical phenomena and are transmitting corresponding
information. A central function must be computed in order to ensure energy efficient operations of the network, or some processing on information collected by nodes. Relevant
examples of this function is the sum of the transmit power of the nodes (which has to be minimized in order to compute power levels), a compression of data, or the computation of the optimal path from a node source to some destination. Traditional
communication paradigms are based on the idea that these operations can be efficiently performed centrally. By the contrary, this is not the optimal approach for WSNs, since it
is clearly inefficient for local nodes to send large amount of information to a central node. The scarce energy, memory and processing capabilities of the nodes may not be able to
meet delay and bandwidth requirements necessary for a satisfactorily data transmission. The alternative solution, proposed by Giridhar and Kumar in 2006, consists in distributing the computation in the
nodes of the network. Despite the apparent simplicity of the idea of in-network processing, it raises several fundamental questions: How distribute the computation? How the distribution of the
computation is affected by lossy channels? How to ensure that the distribution of the computation provides good performance? In this research project, we restrict ourselves
to three prominent aspects of in-network computation: distributed computation of radio transmit powers, distributed source coding, and, last but not least, their complex interaction in distributed routing algorithms. The focus of this research project is on efficient design of the distribution of the computation, and corresponding performance analysis.
System identification concerns the construction and validation of mathematical models of dynamical systems from experimental data.
Our main objective is research on identification of structured systems. The theory for identification of linear systems is by now quite
advanced, while identification of more complex dynamical systems is a most challenging research task. We will in this project focus on the identification of block structured dynamical systems, where the components are rather simple but the overall system behavior can be more complex. We have recently studied Hammerstein systems and cascade systems.
It is very important to be able to analyze the estimated model with respect to its intended use. Here we will concentrate on control and fault
detection. The choice of input signal in the data experiment is very important and directly affect the quality of the estimated model. We will continue our long-term research effort in experiment design, but now concentrate on structured systems. We will also study model reduction and approximation of structured systems.
This project is a continuation of our ongoing VR project \\\\\\\"Experiment Design for System Identification of Complex Systems\\\\\\\". It is a truly joint project between the main applicant Bo Wahlberg and the co-applicant Håkan Hjalmarsson, and covers most parts of the basic research in system identification at Automatic Control, KTH.
Fundamental issues in radio frequency signals and systems
Principal investigator*
Peter Händel
Project leader*
Peter Händel
Project Start
2009-01-01
Project End
2011-12-31
Participants*
Peter Händel, Samer Medawar, Isaac Skog, J-O Nilsson
Abstract
Within this research proposal, we wish to explore fundamental problems within radio frequency
signals and systems. In this proposal, radio frequency typically means, but is not restricted to,
band-widths of the order of mega-Hertz, sampling rates in the order of hundreds of mega-
Hertz, and carrier frequencies in the giga-Hertz range. We are not restricted to a particular
application and the research within this proposal is expected to generate general results applicable
to a variety of applications; although the Ph.D.-student for which funding is applied
for will when appropriate illustrate the fundamental findings with experimental work employing
devices aimed at radio frequency mobile digital communication systems (3G and beyond),
such as analog-digital converters (ADCs), digital-analog-converters (DACs), power amplifiers
(PAs), terminal antennas, et. cetera. Accordingly, the proposal outlines research that is
important both theoretically and in practice.
Signalbehandling för olinjära dynamiska system är ett forskningsprojekt som bedrivs vid Kungliga Tekniska Högskolans institution för Signaler, sensorer och system på Campus Valhallavägen i Stockholm. Projektledare docent Peter Händel beskriver projektet översiktligt då han säger Vi behöver beskriva verkligheten runt om kring oss om vi ska ha möjlighet att förstå den och extrahera den information vi söker! Påståendet låter självklart, men efter lite eftertanke inser man att detta är lättare sagt än gjort. När vi försöker beskriva ett tekniskt system är de ofta både olinjära och dynamiska. Med system, i detta projekt, avses ett tekniskt system det kan vare något så påtagligt som en dieselmotor i en lastbil, en elektriskt integrerad krets, eller en utrustning för behandling av navigeringsdata från en GPS-mottagare med stöd från ytterligare sensorer. Dessa tre system har en sak gemensamt och det är att relationen mellan systemets in och utsignal har ett minne, det vill säga, nuvarande utsignal beror inte bara på den aktuella insignalen utan även dess tidigare värden. Vidare beskrivs förhållandet mellan in och utsignal genom ett olinjärt samband, det vill säga en dubbelt så kraftig insignal resulterar inte i en dubbelt så kraftig utsignal, utan i en utsignal som ges av ett mer komplicerat förhållande till insignalen. Som framgår av projekttiteln avser projektdeltagarna att behandla dessa in- och utsignaler med metoder som är lämpade för dagens datorhjälpmedel, vare sig det är med hjälp av skrivbordets persondator eller med dedicerad utrustning som är nödvändigt i vissa fall. I fallet med dieselmotorn är insignalen det tryck bränsle-luftblandningen har i cylindern, medan utsignalen är det akustiska buller som motorn genererar. I de två andra exemplen är signalerna elektriska. Signalbehandlingen syftar till att extrahera information för att förstå skeenden, eller att omforma de signalerna för att kompensera för olika oönskade effekter i de tekniska systemen. I dieselmotorns fall vill vi kunna styra trycket på bränsle-luftblandningen för bästa avvägningen mellan miljökrav på buller och bränsleförbrukning. I fallet med den integrerade kretsen vill vi, inom projektet, undersöka och utveckla signalbehandlingsmetoder för att förbättra linjäriteten hos, så kallade, analog-till-digitalomvandlande kretsar. I det tredje fallet vill vi modellera rörelseförändringar i navigeringssituationer. I en första anblick kan projektet upplevas som en spridd blandning av olika tekniska system. Peter Händel betonar dock att syftet med projektet är att ta fram och studera generella metoder applicerbara för flera applikationer. Projektet är grundforskning där de olika tekniska systemen beskrivs med matematiska modeller. Det handlar om att ta fram ett ramverk och generella verktyg för digital behandling av signaler härrörande från olinjära dynamiska system. Att studera olika tillämpningsområden parallellt med generella metoder är ofta fruktsamt då man, enligt Händel, får en positiv cross-over effekt mellan olika discipliner. Forskningsförslaget som beskrivs i denna ansökan är en del av den ovan beskrivna forskning inom signalbehandling för olinjära dynamiska system och syftar på att rekrytera en doktorand för vidare metodutveckling inom området, både när det gäller metodutveckling i allmänhet men också med inriktning mot metoder för navigering.
Senior researcher position in Network Interconnected Sensing and Control
Principal investigator*
Karl H Johansson
Project leader*
Karl H Johansson
Project Start
2006-01-01
Project End
2008-12-31
Abstract
Information technology is being embedded into a growing range of physical systems linked together through communication networks. In order to manage complexity and flexibility of these systems, new theories and tools are needed that support the convergence of control with computing and communication. The objective of this proposal is to develop new design methodologies and explore fundamental properties of systems with networked sensor and control nodes.
Three particularly important and interesting problems will be considered: (1) Distributed control over packet-switched and wireless networks. Design methods for network-aware controllers, which should tolerate not only variations in traffic load and transportation delay bud also flexible and unreliable sensors and actuators, will be developed. (2) Resource allocation algorithms to support control-aware network protocols and communication links. Propose new structures with network states that can be negotiated in a middleware between the transport layer and the control application. (3) Coordination strategies for cooperating and mobile sensor and control nodes. Research on fundamental performance limitations caused by time-varying communication topology, symbolic sensor and actuator data, and layered control and information structures. The main part of the work will be basic research, but also experimental evaluations will take place through small-scale demonstrators and collaborations in European projects.
Hard limits are at the core of many branches of engineering and science.
These limits provide bounds on what we can and cannot achieve. Examples can be found in estimation theory, information theory, and control theory. Technology from all of these fields are used together in modern networked control systems. In the physical sciences there are also well-known hard limits. We call these physical hard limits. We argue that physical hard limits also give hard limits on the performance of entire networked system. The focus here is on physical hard limits arising from thermal noise and from limits on available energy. The specific goal is to quantify hard limits of physical implementations of actuators, sensors, controllers, and to interconnections of them. Our preliminary results indicate that a good approach to quantify these limitations is to match lossless dynamical systems to device models. As an application for the results we plan to use a control-theoretic heat engine that transfers uncertain energy (heat) into useful energy (work) using feedback control. That is an interesting problem in its own right, with close connections to thermodynamics. This project is an attempt to bring in some of the physical limitations into the picture of hard limits for networked control systems. Since current research has mostly been focusing on limits arising from information theory and control theory, this should help broaden our understanding of networked control systems.
The main objective in the proposed research project is to develop modeling methods for systems that act in information-rich environments. Systems in information-rich environments are systems where we have access to large amounts of measurement data. Often these systems are networked systems, and they can be both man-made and natural. Examples are found in, for instance, the car and truck industries, in the mobile phone industry, and in systems biology.
The work plan includes three related projects. The first project deals with the problem of fitting simple structured models to networked information-rich systems. This will require the methods to be able to take large amounts of measurement data into account. The methods should also give guidance on how to simplify the interconnection structure and the dynamics of the model.
The second project deals with
the problem of simplifying models with respect to simulation time. This is important to be able to use the models in time-critical applications, such as in control engineering. The third project is more theoretical and deals with the problem of finding fundamental limitations for control and estimation in large networked systems. This is important to help us understand what we can and cannot achieve with our systems. These results are expected to be useful both in industry and in academic fields where various aspects of modeling of these systems are considered.
The project aims at dramatically increasing the reliability of wireless sensor networks by viewing reliability as a first-class systems property instead of a detail in a few layers of the system. Our holistic approach combines the modularity benefits of layering with the performance benefits of vertically integrated solutions. To future-proof the project we identify and explore two emerging technologies, on-chip integrated communication peripherals and software-defined radio.
Funding
VINNOVA - Swedish Governmental Agency for Innovation Systems
WIDE will develop a novel rigorous and integrated framework for advanced control and real-time optimization of large-scale spatially distributed processes that exploits wireless sensor networks for pervasive and reconfigurable information collection, and at validating the approach on a real city water distribution system.
WISA develops wireless sensor and actuator networking technologies for measurement and control. The project takes a cross-layer perspective, and aims at developing networking protocols, sensor fusion techniques and control algorithms that work in harmony, enabling a wide deployment of wireless technologies in automation and monitoring.
Funding
VINNOVA - Swedish Governmental Agency for Innovation Systems
SERAN stands for simple and efficient radio access networks, and is a joint project between Ericsson Research and KTH looking at advanced radio resource management and network coding techniques for ITM advanced and beyond.
KTH, GET, INRIA, France Telecom, Alactel Lucent, CERTH
Abstract
Euro-NF is a Network of Excellence on the Network of the Future, formed by 35 institutions (from the academia and industry) from 16 countries. Its main target is to integrate the research effort of the partners to be a source of innovation and a think tank on possible scientific, technological and socio-economic trajectories towards the network of the future.
Vital Infrastructure Networks, Information and Control Systems Management
Principal investigator*
Karl H Johansson
Project leader*
Karl H Johansson
Project Start
2008-10-01
Project End
2010-12-31
Participants*
Gunnar Karlsson, Henrik Sandberg, Carlo Fischione and György Dán
Abstract
Society is increasingly dependent on the proper functioning of the electric power system, which in turn supports most critical infrastructures: water and sewage systems; telecommunications, Internet and xomputing services; air traffic and other transportation. Many of these other infrastructures are able to operate without power fort shorter periods of larger power outarges may be difficult and time xonsuming to restore. Such outargets might thus lead to situations of the non-functioning societies with devastating economical and humanitarian consequences. For this reason, this consol decided to concentrate its research to the systems for transmission and distribution of electric power. We anticipate of the results will be applicable to the protection of other critical infrastructures. The operation and management of the electric power system depend on computerized industrial control systems. Keeping these systems secure and resilient to external attacks as well as to internal operational errors is thus vital for uninterrupted service. However, this is challenging since the control systems are extremely complex. Yet, the systems are operating under stringent requirements on availability and performance. If control and supervision are not done in real-time, the power network may come to collapse.
The objective of the project is to develop, test and evalutate methodologies for the analysis, design and operation of resilient and secure industrial control systems for critical infrasturcures. Methodologies will be developed with a particular focus on increased robustiness of the control system. As mentioned, the focus is on power transmission and distribution networks. The project conbines a holistic management prespective- in order to counteract sub-optimization in the design- with in-depth analysis and development of security solutions adapted to the specific requirements of networked control systems.
Mikael Skoglund, Henrik Sandberg, Carlo Fischione, Ulf Jönsson and Ragnar Thobaben
Abstract
The revolutionary developments in microelectronics over the past decades have led to the production of cheap yet powerful devices that can communicate with one another, can sense and act on their environment and can be deployed in large numbers to deliver an abundance of data. Such devices and the networks they form( broadly grouped under the term wireless sensor networks) bring together communication, computation, sensing and control and have enabled monitoring and automation at an unprecedented scale. Especially challenging in this context are networked control systems, where feedback control loops are closed over networked , distributed communication platforms. To take full advantage of this technology novel design methods are necessary that transcend the traditional borders between disciplines, to apply the principles of feedback to complex, interconnected systems. The objective of the FEEDNETBACK project is to generate precisely such a xo-design framework, to integrate architectural constraints and performance trade- offs from control, communication, computation, complexity and energy management. This will allow the development of more efficient, robust and affordable networked control technologies that scale and adapt with changing application demands. By focusing on wirelessly connected networks and leveraging on recent advances in sensor networks, we will study networked control from a fundamental point of view. We will extend the current scientific state-of-the-art in networked control and will develop a software tool set to support our co-design framework. To demonstrate and evaluate this framework, we will apply it to two industrial case studies: a smart camera network for surveillance and motion capture and an underwater inspection system that comprises autonomous surface and underwater vehicles. The case studies are chosen to test the applicability and limitations of the developed approach over a variety of resource constraints, such as network size, cont.
CoopNet develops fundamental theory for control and optimization of networked systems. Particular attention is given to robust and distributed optimization, and its applications in data networking, wireless communications and industrial automation.
The AURES project (autonomous UGV systems for reconnaissance and surveillance) aims at developing and demonstrating autonomous functions in a UGV system used for reconnaissance and surveillance. A UGV is an unmanned ground vehicle, and the focus is on developing three basic capabilites for a cooperating group of such vehicles: First, the UGVs should be capable of patrolling areas in an efficient manner. Second, they should be able to autonomously position themselves to provide continuous camera coverage of specified areas and buildings. Third, the UGVs should be able to search and secure an area or building, so that an intruder cannot escape undetected. In addition to these fully autonomous capabilities, the project will also evaluate different concepts for control with a flexible level of autonomy.
The capture and rendering of 3D video is rapidly evolving. Initial applications are likely within entertainment and informational fields, but a multitude of further application areas will follow. This means that the demand for the transmission of 3D video is likely to start growing rapidly in the not too distant future. Coding algorithms will play an important role as 3D video is an extremely demanding use of telecommunication networks. 3D video rendering (also referred to as Free Viewpoint TV (FTV)) allows arbitrary viewer positions, which, within certain limits, are different from the camera positions during the capture. This in turn allows a range of video renderings which spans from rendering on 2D displays with user selectable viewpoints to spatial video renderings via stereo or 3D displays.
Service- oriented cross-layer infrastructure for distributed smart embedded devices
Principal investigator*
Karl H Johansson
Project leader*
Karl H Johansson/ Mikael Johansson
Project Start
2007-01-01
Project End
2009-12-31
Abstract
The goal of the SOCRADES project is to create new methodologies, technologies and tools for the modelling, design, implementation and operation of networked hardware/ software systems embedded in smart physical objects. The smart embedded system is to be applied in perception and control systems in intelligent environments, in which enhanced system intelligence is achieved by co-operation of smart embedded devices pursuing common goals. These devices with embedded intelligence and sensing/actuating capabilities are expected to be heterogeneous yet they need to interact seamlessly and intensively over a network (wired/wireless).
The middleware technologies to be developed in this project will be based on the Sevice-Oriented Architecture approach, will be generic to any networking technology or transmission medium, and will provide open interfaces that enable interoperability at the semantic level to any third party. A SOCRADES service will be implemented as a software component that encapsulates device- specific functionality. This functionality is advertised to the outside world, in order to be located and invoked by other networked devices and/or applications, without the latter being aware in any way of how the functionality is implemented.
The SOCRADES approach is to create system intelligence by a large population of small and smart networked embedded devices at a high level of granularity, as opposed to the traditional approach of focusing intelligence in a few large and monolithic applications. This increased granularity of intelligence distributed among loosely coupled intelligent physical object facilitates the adaptability and reconfigureability of the system, allowing it to meet business demands not foreseen at the time of design. Focus from a functional perspective will be in managing the vastly increased number of intelligent devices and the associated complexity. Focus from a run-time infrastructure view will be on a new breed of very flexible real-time embedded devices (wired/wireless) that are fault-tolerant, reconfigurable, safe and secure.
Mikko Alava, Erkki Oja, Samuel Kaski, Esko Ukkonen, Antti Kupiainen
Abstract
The Academy of Finland - Finland Distinguished Professor - project - Statistical Physics,
Distributed Systems, and Computational Biology -
aims to create, within the FiDiPro framework, an internationally leading research team in the fertile and highly active cross-disciplinary area
spanning computer science, physics, and computational aspects of biology. The team
will build on, and bring together, the expertise of Prof. Erik Aurells research groups
at Stockholm (KTH), the KTH Linnaeus Center ACCESS, and the traditions of a number of Finnish groups,
involving two national Centres of Excellence active at TKK. The team will be based
at TKK partly at the Department of Information and Computer Science and partly
at the Department of Engineering Physics.
Communication systems form a defining aspect of modern society. Particularly important in this respect is the transmission of audio-visual information, since it is a fundamental means of human interaction. The rapid growth in network capacity and the increasing diversity of services is associated with a rapidly increasing heterogeneity of network attributes and quality of services (QoS). Current state-of-the-art coding systems for audio and video are designed for a particular scenario, and will fail if faced with a communication channel with different properties. In contrast, we propose to introduce flexibility into source and channel coding algorithms. To this purpose we will develop a new paradigm that replaces off-line coder design by real-time coder design, thus ensuring the optimality of the coder over a large range of environmental conditions (e.g., rate, quality, and robustness requirements imposed by the channel). The new algorithms are expected to retain or improve on the efficiency of current state-of-the-art algorithms. The new paradigm is based on the usage of universal statistical models of the sources and channels, and generic models of distortion. The flexibility is obtained through the use of models that allow analytically derived source and channel coding methods that do not require training.
The aim of the proposed project is to develop specific new techniques that abate the effects of packet loss at either no increase in rate or a very small increase in rate and that can be used with existing source coders. The project is motivated by the desire to minimize the need for increased network capacity, which is particularly relevant for the increasingly common wireless network, by the observation that packet loss occurring in the commonly used packet networks is relatively low. For such networks, the solutions proposed herewithin make high levels of redundancy unnessecary. To facilitate the introduction of the new methods in practical applications, they are designed to function with existing source coders, and, thus, will not contribute to the ongoing proliferation of new coding standards.
Mobile communications is an important economic driver generating growth. Significantly improved transmission capabilities are increasingly required to support content-rich data oriented services in order to connect people as well as machines to the information society.
The support of broadband services for mobile and wireless applications towards IMT-Advanced, with excellent user experiences, are key trends for future radio access technologies (RAT), providing deployment scenarios with reduced operators CAPEX and OPEX. The WINNER+ project addresses these challenges from a technical, standardisation and regulatory perspective.
Based on the basic system concept, which was developed in the FP6 WINNER and WINNER II projects, this project will develop, optimise and evaluate a competitive IMT-Advanced candidate proposal by integrating innovative and cost-effective additional concepts and functions and providing an evolution path towards further improved performance of IMT-Advanced. This development is ongoing in a globally competitive environment. Technology candidate proposals are expected from Europe, IEEE, China, Japan and Korea. In order to reduce fragmentation and to ensure a competitive European position in the global context, this project is mobilising manufacturers and operators in Europe and the research community for a collaborative research effort. The consortium is based on the leading role of European companies in the global market on mobile and wireless communications.
Innovations are expected in the areas: radio-resource management, including active management of interference, heterogeneous network deployment including relaying as integral part, in combination with antenna concepts, spectrum sharing and its flexible usage, multi- and broadcast capabilities, exploitation of peer-to-peer links between user terminals and incorporation of network coding. These activities will take into account the results of WRC 2007 in November 2007.Frequency spectrum, as a scarce resource, requires efficient reuse and sharing techniques between different RATs. The WINNER+ system focuses on tight and low latency cooperation across RATs but also across WINNER+ networks for efficient spectrum use.
Key technologies and selected concepts will be evaluated and optimised through prototypes and emulators in order to show the feasibility of the developed radio interface. The WINNER+ project will contribute essential technical information to the forthcoming standardisation and regulatory process after WRC 2007 via well-established channels, significantly increasing the opportunities to exploit the WINNER+ system concept.
Magnus Jansson, Petter Wirfält, (Karl Werner alumni)
Abstract
The main goal of this project is to analyze existing and develop
new efficient estimation algorithms for applications in signal
processing. The approaches that will be studied combine statistical
modeling with robust computational tools from numerical
linear algebra and optimization.
In the project proposal several, partly overlapping,
open problems are identified in:
subspace system identification of linear dynamical systems; spectral
analysis; array signal processing; multidimensional wireless
communication channel estimation; reduced rank linear regression; and
structured matrix estimation. We intend to approach these specific
problems from a statistical estimation theory
viewpoint and study fundamental algorithmic limitations as well
as algorithm performance optimization.
Information flow security concerns the problem of determining and controlling the nature and amount of information flowing to and from some a component of computer system. Information flow security underpins such basic security concepts as confidentiality and integrity, and is of fundamental importance in computer security. Since the late 90´es substantial progress has been made in identifying basic models for information flow security, mostly based on a multi-level security model, where the goal is to completely prevent all flows from higher to lower levels. Unfortunately, the practical impact of this research has been minimal. There are a number of reasons for this, including an overly restrictive security model, and an overly restrictive scope that largely ignores, for instance, timing attacks. The goal of the project is to develop models that can more precisly constrain applications to allow those information channels that are intended, i.e. inherent, for the application at hand, while disallowing others, including timing channels. The security model will be phrased in terms of a new concept of \\\\\\\"secure sub-computation\\\\\\\", capturing the situation where parts of a computation must be strictly constrained when applied to secret, or high-integrity data, whereas other parts can act freely on public data. The concept will be examined from both a theoretical and a practical point of view, in terms of efficient algorithms, implementations on Java bytecode, and experiments.
Agile MIMO Systems for Communications, Biomedicine, and Defense
Principal investigator*
Björn Ottersten
Project leader*
Björn Ottersten
Project Start
2009-01-01
Project End
2013-12-31
Participants*
Peter Händel, Mikael Skoglund, Magnus Jansson, Mats Bengtsson, Per Zetterberg
Partners*
Petre Stoica, UU
Abstract
This proposal targets the emerging frontier research field of multiple-input multiple-output (MIMO) systems along with several innovative
and somewhat unconventional applications of such systems. The use of arrays of transmitters and receivers will have a profound impact on
future medical imaging/therapy systems, radar systems, and radio communication networks. Multiple transmitters provide a tremendous versatility and allow waveforms to be adapted temporally and spatially to environmental conditions. This is useful for individually tailored illumination of human tissue in biomedical imaging or ultrasound therapy. In radar systems, multiple transmit beams can be formed simultaneously via separate waveform designs allowing accurate target classification. In a wireless communication system, multiple communication signals can be directed to one or more users at the same time on the same frequency carrier. In addition, multiple receivers
can be used in the above applications to provide increased detection performance, interference rejection, and improved estimation accuracy. The joint modelling, analysis, and design of these multidimensional transmit and receive schemes form the core of this research
proposal. Ultimately, our research aims at developing the fundamental tools that will allow the design of wireless communication systems
with an order-of-magnitude higher capacity at a lower cost than today; of ultrasound therapy systems maximizing delivered power while
reducing treatment duration and unwanted illumination; and of distributed aperture multi-beam radars allowing more effective target location, identification, and classification. Europe has several successful industries that are active in biomedical imaging/therapy, radar systems, and wireless communications. The future success of these sectors critically depends on the ability to innovate and integrate new technology.
Multi-antenna Transmission and Scheduling in IMT-Advanced
Principal investigator*
Mats Bengtsson
Project leader*
Mats Bengtsson
Project Start
2008-01-01
Project End
2010-12-31
Participants*
Mats Bengtsson, Björn Ottersten, Mikael Skoglund, Per Zetterberg, Ragnar Thobaben
Partners*
Ericsson Research, Uppsala University
Abstract
The project will study technical solutions for next generation mobile access technologies. The project intends to develop and validate techniques for scheduling and controlling transmission to/from multiple users over multiple antennas. This includes multi-antenna base stations and transmission over multiple relay nodes (cooperative relaying).
The expected results are within the areas of multi-antenna transmission techniques, multiple access, Spatial division multiple access, adaptive transmission, opportunistic and cooperative transmission, scheduling and MAC layer design. These will utilize and strengthen the results from earlier and ongoing EU projects, such as WINNER, WINNER II, WINNER+, COOPCOM and SENDORA, and is also expected to provide input to standardization work. Closer cooperation between KTH, Uppsala University and Ericsson Research, which forms a good basis for future EU programs.
The work will be based upon radio channel measurements, to be used to evaluate and refine different radio transmission algorithms. Similar work based on common channel data and with regular technical project meetings, has lead to fruitful results in earlier projects.
Handling Radio Resources in Space, Time and Frequency
Principal investigator*
Mats Bengtsson
Project leader*
Mats Bengtsson
Project Start
2007-01-01
Project End
2009-12-30
Participants*
Mats Bengtsson
Abstract
The request for higher data rates and better connectivity in future wireless systems present a never ending challenge for the system designers, since the radio
spectrum is a limited resource. Also, current systems carry mostly speech and SMS whereas the future traffic will be a mixture of file transfers, web access, streaming video and voice, with very differing demands on delays and error rates.
One approach to handle these problems is to use multiple antennas, possibly both at the user terminals and access points. However, an efficient solution needs to determine the spatial processing jointly with the scheduling in time and frequency. The goal of this project is to develop theories and algorithms that use the available radio resources as efficiently as possible.
Cooperative and Opportunistic Communications in Wireless Networks
Principal investigator*
Björn Ottersten
Project leader*
Per Zetterberg
Project Start
2006-10-01
Project End
2010-09-30
Partners*
Ericsson, TSI, UPC, EURECOM, UCL, RATCI
Abstract
Cooperation and opportunism are two concepts that have recently revolutionized the way engineers think about wireless system design. Both have their origin at information theoretic studies and target the maximization of the spectral and power efficiency at the system level. Successful incorporation of these concepts in wireless system standards requires the solution of difficult theoretical and practical problems. Consequently, these concepts are expected to be long-term research topics of fundamental nature in the communications area. We plan to:
1. explore performance limits of cooperative and opportunistic schemes (e.g., information theoretic analysis of combined cooperative-opportunistic schemes, large-system analysis for gain quantification of cooperative schemes) and multiuser diversity with limited feedback;
2. develop efficient strategies and coding schemes for cooperative and opportunistic communications, maximizing diversity and coding gain under various operational scenarios (homogeneous versus inhomogeneous channel statistics, full versus partial channel knowledge, codes for a combined cooperative-opportunistic scenario);
3. develop efficient resource allocation with limited feedback, including decentralized resource allocation, by targeting the optimal exploitation of various grades of channel information at the transmitters (full, partial, statistical);
4. implement selected cooperative and opportunistic schemes on a test-bed, in order to fully understand practical implementation constraints, assess the benefits of cooperation and opportunism in realistic environments, and support proof-of-concept.
The goal is to attain an order of magnitude higher aggregate data rates than what is currently envisioned in 4G research, while supporting much tighter QoS guarantees.
Access technologies: multiple access, multiple users, multiple distributed antenna systems
Principal investigator*
Björn Ottersten
Project leader*
Lars Jonsson
Project Start
2008-07-01
Project End
2011-06-30
Participants*
Mats Bengtsson
Partners*
Lund University, Beihang University, Tsinghua University, Huazhong University of Science and Technology
Abstract
Multiple-Input Multiple-Output (MIMO) technology has a great potential for future wireless communications in achieving high throughput and reliable transmission. Existing MIMO techniques can be directly applied for cellular systems with intra-cell and inter-cell interference suppressed by frequency partition and orthogonal multiple access, which are not optimal in general cases. In fact, a system of multiple antennas has the capability of suppressing both the inter-cell and intra-cell interference. An important bottleneck for implementing MIMO techniques in a mobile communication system exists at the user terminal with multiple antennas, which is required to be simple and of small size. Consequently, the strong electromagnetic coupling and signal correlation that result from the close proximity of the multiple antennas in a small confined volume inevitably reduce the information carrying capability of the MIMO channel. Another bottleneck to achieve the promised benefits provided by MIMO techniques is the interference which will lead to considerable performance loss. In future cellular systems, the channel quality may be hard to predict due to the bursty nature of the interference caused by packet switched data, link adaptation, varying channels, and opportunistic scheduling. This will lead to a new challenge for the resource allocation and interference suppression.
Funding
VINNOVA - Swedish Governmental Agency for Innovation Systems
This research proposal targets the research field of multi-user multiple-input multiple-output (MIMO) wireless communications. The use of arrays of antennas at the transmitter and receiver in radio communication links allows spatially selective transmission and reception providing increased link capacities as well as improved spectral efficiency. In future packet based wireless communication systems, MIMO technology in combination with multi-user diversity employing adaptive modulation and coding will be key components. Within this proposal, we will identify and formulate a number of critical research topics regarding the modeling, analysis, and design of multi-user MIMO systems. The subjects will be fundamental in nature and the goal is to provide understanding that is independent of a specific wireless system or standard. However, our participation in targeted research efforts in the European framework programs, FP6 and FP7, allows the dissemination of our research results and ultimately this will influence the design of future wireless systems.
Networked systems pervade many of todays technologies
such as the Internet, power distribution grids and process
manufacturing. Comparatively little has been done from a modeling and,
especially, an identification perspective for such systems.
Firstly, we aim to develop an accurate model for window/acknowledgment
based congestion control and to analyze its implications on the
possibilities and limitations of this control paradigm. Within the
on-going project we have developed a strong candidate model showing very
good resemblance to packet level simulations, in single
bottleneck scenarios. This study is motivated by that this is the
control paradigm used in Internet today and a key issue is to
understand if it is able to meet future demands in terms of capacity
and increased heterogeneity in traffic and links.
We also intend to study model errors in system
identification of networked and decentralized systems.
Our focus will be on the case where models are locally identified at
nodes and how local information is to be combined as to give a relevant
representation of the total model error. Here the challenge lies in
that disturbances that are correlated across the nodes imply that the
model errors are correlated across the nodes. The starting
point of our analysis will be a geometrical approach to assess model
errors. The problem is important in control of networked and
decentralized systems.
The objective of this project is to contribute to the basic understanding and design of wireless networked embedded systems. These systems, which consist of digital devices that can sense, communicate and act on their environment, are present in every aspect of life. Today their use is mostly in monitoring, but their potential capability goes far beyond that since they can have control functionality: they can react to signals and act on the environment in an autonomous and intelligent way. It is an open research problem how to design wireless networked embedded systems in a rigorous and systematic way. On this project we will contribute with fundamental research in the area by studying three important problems: (1) control and estimation over wireless networks, (2) network protocols and communication for real-time control, and (3) coordination of mobile network nodes. In the first problem, we will investigate control architectures and algorithms that can deal with communication imperfections and network variations. The second problem is on how to adjust the network to improve its ability to serve control applications. In the third problem, we will study how mobility of network nodes can improve communication, and thereby enable better coordination of networked systems. The main part of the research will be fundamental theoretical work, but also experimental evaluations will take place through small- scale demonstrators and collaborations in European projects.
