Reinhardt Karnapke

Challenges for 100 Gbit/s end to end communication: Increasing throughput through parallel processing

Todays applications and services become more dependent on fast wireless communication, for the upcoming years data-rate demands of 100Gbit/s can be easily expected. However, fulfilling that demand is a task which cannot simply be solved by upscaling existing technologies. While most of the research tackles the challenges regarding the transmission technology from the physical layer up to base-band processing, we focus on the challenges concerning the handling of that vast amount of data. The overall goal is to bring together the transmission technology with the operating system to create a suitable end-to-end communication solution. In this paper we argue that communication can be understood as a soft-realtime problem and how that helps introducing parallelism into protocol-processing.

CoBaS: Introducing a Component Based Scheduling Framework

Many-Core systems and heterogeneous systems are getting more and more common and may soon enter the mainstream market. To harvest their capabilities to their full potential, the runtime systems scheduling policies have to be adapted and, in many cases, tailored to the specific system. The runtime system can be both an operating system or management infrastructure of an infrastructure as a service (IaaS) platform. Developing, implementing, and testing those scheduling policies is a challenging task in general. In this work we present CoBaS, a component based scheduling framework for multi and many-core runtime systems. The main purpose of CoBaS is the simplification of the scheduling policy implementation and an increased code reuse to save time during development. CoBaS uses a novel approach to reach that goal. It allows the breakdown of the policy implementation into several components that can be reused. Through composition, a fast prototyping, testing and evaluation of new scheduling policies is possible without implementing every functional part again. CoBaS uses an event based approach to distribute information about system states and state changes between the runtime system and components as well as between components themselves. Furthermore, it has a facility to hand over ordered task sets between components. We have adapted both the Linux and Free BSD kernel to use CoBaS by completely removing the native scheduler. The integration of CoBaS into those kernels shows the feasibility of our approach.

On the Need of Systemic Support for Spatio-Temporal Programming of Mobile Robot Swarms

In this paper, we present SwarmOS -- a distributed operating system for mobile robot swarms. SwarmOS features transaction-based spatio-temporal programming of mobile robot swarms on a systemic level. We show the programming model and resource management in SwarmOS. Swarm applications consist of concurrent, distributed and context-aware actions. We provide distributed transactions in order to guarantee atomic execution of a set of dependent actions. We distinguish between schedulability and executability of a set of actions: the first one is checked by the space-time scheduler which is a core service of the execution environment. The scheduler plans actions in space and time and computes spatio-temporal trajectories if robot movement is necessary. In order to guarantee executability of a distributed transaction of spatio-temporal actions, we present the concept of path alternatives and a time-based two-phase commit protocol in order to assure consistency. We show the feasibility of our approach by performing experiments on our testbed.

Systemic support for transaction-based spatial-temporal programming of mobile robot swarms

In this paper, we present an approach to support transaction-based spatial-temporal programming of mobile robot swarms on a systemic level. We introduce a programming model for swarms of mobile robots. Swarm applications consist of concurrent, distributed and context-aware actions. We provide distributed transactions in order to guarantee atomic execution of a set of dependent actions. We distinguish between schedulability and executability of a set of actions. In order to guarantee executability of a distributed transaction of spatial-temporal actions, we present the concept of path alternatives and a time-based two-phase commit protocol in order to assure consistency. We show the feasibility of our approach by a proof-of-concept.

Unidirectional link counter - a routing protocol for wireless sensor networks with many unidirectional links

Experiments with wireless sensor networks have shown that asymmetric and unidirectional links occur often. Protocols that detect these links on the link layer have been proposed, however, unidirectional links represent a challenge as well as a chance across all layers. Still, most routing protocols prefer to use only bidirectional links. In this paper we present Unidirectional Link Counter, a routing protocol that collects information about the unidirectional links on a path to make routing decisions. The collected information can furthermore be used for more effective neighborhood management and better quality estimation.

Simulation-based tracing and profiling for system software development

Tracing and profiling low-level kernel functions (e.g. as found in the process scheduler) is a challenging task, though, necessary in both research and production in order to acquire detailed insights and achieve peak performance. Several kernel functions are known to be not traceable because of architectural limitations, whereas tracking other functions causes side effects and skews profiling results. In this paper, we present a novel, simulation-based approach to analyze the behavior and performance of kernel functions. Kernel code is executed on a simulated hardware platform avoiding the bias caused by collecting the tracing data within the system under observation. From the flat call trace generated by the simulator, we reconstruct the entire call graph and enrich it with detailed profiling statistics. Specifying regions of interest enables developers to systematically explore the system behavior and identify performance bottlenecks. As case study, we analyze the process scheduler of the Linux kernel. We are interested in quantifying the synchronization overhead caused by a growing number of CPU cores in a custom, semi-partitioned scheduler design. Conventional tracing methods were not able to obtain measurements with the required accuracy and granularity.

Spatio-Temporal Coordination of Mobile Robot Swarms

Context-aware applications that require access to physical space and time are a necessity in cyber-physical systems. We focus on the design of a cyber-physical operating system in which a space-time scheduler is the core-component responsible for resource management. Given a set of space-time tasks and a set of mobile robots that move through physical space, a main objective remains in finding a mapping of tasks to robots. In this paper, we address the problem of scheduling a set of tasks with spatio-temporal constraints in space and time. We present an online-scheduler that computes collision-free spatio-temporal trajectories for the robots in order to execute the space-time tasks. As side condition, collisions with static as well as dynamic obstacles must be avoided at all times. The scheduler consists of two components: a job scheduler that uses a heuristic and performs a coarse-grained scheduling and a trajectory planner that takes the output of the job scheduler and computes spatio-temporal trajectories.

The Swarm as a Service: Virtualization of Motion

A defining aspect of Cyber-physical systems are the sensors and actuators of a computing system, which interact with the physical world. While sensors and actuators can take on many different forms, one type of actuator, namely motors, usually has the highest influence on the lifetime due to the high energy consumption. Therefore, it is necessary to reduce physical movement of robots as much as possible without reducing the quality of the application. In this paper we introduce virtual movement of applications, meaning that parts of an application can be executed on arbitrary robots that can reach the locations defined by the application programmer with minimal effort. As we assume that robots will be shared by multiple applications, programming virtual movement within the application is not only cumbersome, but impossible, as the location and status of a robot may be changed by a different application. Therefore, we provide a middleware abstraction which takes care of virtual movement and hides it from the application programmer.

Applying wireless sensor networks to renaturation monitoring: From requirements to deployment

In Germany, renaturation of former coal mining regions is an important topic. One of these regions, called Chicken Creek, is used by ecologists to observe the development of an initial ecosystem without human interference - a scenario that seems perfect for a wireless sensor network. However, the sensor network applications are still mostly designed from scratch for every new scenario. The reason for this are strongly differing user requirements. This time-consuming problem can be overcome by providing a reusable module pool from which hard- and software components can be chosen. Based upon this module pool, a semi-automatic generation of sensor network applications becomes possible. In this paper we focus on the differences between the aforesaid theory and the practice during the initial deployment.