welcome to ETRI
Insert title here

Software Research Laboratory

HOME > R&D > Software Research Laboratory > Embedded Software Research Department > areas for Research & Development

areas for Research & Development

  Developing an embedded software common platform for information and industrial
  devices

We are developing the Linux-based embedded software platform which covers

  • a domain-specific optimizing operating system to run information terminals,
  • development tools to help user develop applications conveniently and easily, and
  • testing and verification tools to enhance the platform safety.

The target application areas of our embedded software platform include information devices (such as PoSs, broadcasting set-top boxes, MIDs, PMPs and game devices) and industrial devices (such as industrial control systems, autonomous driving systems and precision instruments)

The domain-specific facilities of our embedded software common platform can be classified into the embedded operating system, UI and middleware and development environments tools.

The embedded operating system

We are developing a soft-realtime kernel that has both high responsiveness and low overhead for mobile devices and industrial control systems, and a fast boot system for information and industrial devices, and a power management technique that increases the battery life time by minimizing the power consumption of information and industrial devices, and a reliable system that provides the high availability.

The UI and middleware

We are developing a rich client technique that supports elegant user interfaces to users and improves the usability of users, and a communication middleware that supports reliable communication protocols such as WiFi, CDMA, HSDPA and WiBro.

The development environment tools

We are developing a platform reconfiguration tool that reconfigures system modules specialized for each hardware platform, and platform testing and verification tools that enhances the platform safety and checks whether the platform conforms to Linux APIs. These tools will remove difficulties developers can experience during developing embedded software.

The contribution of our embedded software common platform will be as follows.

  • Our embedded platform will help develop embedded systems more quickly and easily by using the providing facilities specialized on varied application ares,
  • Our embedded platform will help animate the related industries by developing technologies, standardizing and vitalizing communities.

  Mobile SW Platform Technology

ETRI is researching and developing a mobile SW platform which is leading mobile contents and the service industry. The SW platform that enables development and execution of mobile applications and services is a core technology leading digital convergence for the mobile environment.

We are researching and developing a service-oriented mobile middleware technology to propagate high-end mobile services in the mobile industry concentrated on contents and service. This research includes a mobile application and a service framework based on a mobile component technology.

We are also developing a mobile simulator based on hardware virtualization technology for satisfying the market needs of the integrated development environment. This study enables a test of various mobile applications without real mobile terminals. It includes an event-based test automation which supports testing of a full SW stack for mobile devices without any user interaction.

  Study on lightweight realtime embedded operating system

Supporting embedded applications such as environment, disaster prevention, digital home, national defense, and industrial equipment, we have been developing two lightweight operating systems which enables convenient development and efficient execution on hardware platforms, (1) NanoQplus for the use of wireless sensor nodes and (2) ROSEK for automobile ECU(Electronic Control Unit)s.

NanoQplus is a multi-threaded operating system which is designed so that users can write multi-thread applications very easily, provides the full-preemptive scheduler to obtain high system response, and overcomes the limitation of small memory size from possibly optimized kernel configuration by users. The mesh and tree based routing protocols minimized the energy consumption for data communication, thus it is expected to be a next-generation sensor network operating system.

ROSEK is a preemptive real-time multitasking operating system with high reliability, based on OSEK/VDX international standard. Since stability and reliability are the most important factors in the automotive industries, the operating system must support high reliability and fault tolerance to the level of software requested. In addition, it should provide communication protocols that can operate normally under any vibration or noise environments. ROSEK not only has the real-time control feature but also has the functions of multitask management, event management, alarm management, error tracing and debugging. Conclusively, it is possible that ROSEK can be used in high-pass terminal, automobile field, traffic light control, robot, and factory control in the future.

  Study on NanoQplus-based Sensor Network Simulator(RealSSim)

NanoQplus-based sensor network simulator, RealSSim(Realistic Sensor Network Simulator) is a sensor network simulator based on RF propagation analysis, and a simulator for large-scale sensor network systems. RealSSim helps users to establish an installation plan of a sensor network system, estimate performance of the system, and check possible fault occurrences before the system installation. Especially, RealSSim provides precise simulation results using RF propagation model and energy model based on measurement in real environment. RealSSim can be applied to various sensor network systems such as indoor location tracking system, U-city, national defense, shipbuilding, transport, etc.

RealSSim has the following features.

  • Visual Modelling of Sensor Network System
  • To describe space, sensor nodes, and scenario information required for simulation
  • Ray-tracing based RF Propagation Model
  • To compute the strengths of received signals based on the RF propagation model that considers the properties of spaces and reflection, penetration, and diffraction of RF signals
  • Simulation Execution and Result Analysis
  • Automatic deployment of sensor nodes based on RF propagation analysis and sensor network simulation and visualization based on VPSI(Virtual Protocol Stack Interface), which is a technology to be able to add and remove new protocol with ease and fastness
  • NanoQplus based Source Code Generation
  • After simulation is completed, to generate NanoQplus based source code for each sensor node

  Research on CPS technologies for autonomic SW with high dependability

CPS (Cyber Physical Systems) means autonomic embedded systems based on high dependability with realtime control of distributed physical systems through wired/wireless networks. While the legacy embedded software is focused on the system optimization of each embedded system, recently the novel embedded software technologies are required to reliably maintain the entire system with large number of mutually connected sub systems by means of realtime monitoring, concurrent control and automatic recovery. Therefore, we are developing system control softwares, namely CPS softwares, for safety critical services requiring reliability, security, fault tolerance and robustness.

The research covers the efficient development, verification and realtime operation of distributed intelligent autonomic systems.

  • Precise modeling and simulation for CPS reliability verification
  • Realtime middleware for control of distributed CPSs
  • Autonomic computing platform for highly reliable service support

The final results of this research will include

  • The realtime control software for intelligent defense systems, such as future combat systems and intelligent warship
  • The realtime control software for future intelligence buildings

The concept of CPS technologies for highly dependable autonomic control SW