Tobias Ziermann

Run time mapping of adaptive applications onto homogeneous NoC-based reconfigurable architectures

Tile-based NoC architectures have emerged as promising field-programmable hardware architectures which provide arrays of programmable processing units to support run time adaptation. However, efficient real-time support is required for executing adaptive applications which offline design approaches may not provide. In this paper, running dynamic master/slave applications on homogeneous NoC architectures is investigated. Such applications have the characteristic that their program structure may be adapted at run time by inserting or removing tasks to react to changing requirements that are not known a priori. A heuristic is presented for an energy-and performance-aware assignment of dynamically created tasks to the processing units. The provided experimental results give evidence of the benefits of the proposed methods for synthetic test-cases as well as an adaptive image processing application.

Interprocedural Placement-Aware Configuration Prefetching for FPGA-Based Systems

One of the major impediments to deploying partially run-time reconfigurable FPGAs as hardware accelerators is the time overhead involved in loading the hardware modules. While configuration prefetching is an effective method that can be employed to reduce this overhead, mispredicted prefetches may worsen the situation by increasing the number of reconfigurations needed. In this paper, we present a static algorithm for configuration prefetching in partially reconfigurable FPGAs that minimizes the reconfiguration overhead. By making use of profiling, the interprocedural control flow graph, and the placement information of hardware modules, our algorithm predicts hardware execution and tries to prefetch hardware modules as early as possible while minimizing the risk of mis-predictions. Our experiments show that our algorithm performs significantly better than current state-of-the-art prefetching algorthms for control-bound applications.

Game-theoretic analysis of decentralized core allocation schemes on many-core systems

Many-core architectures used in embedded systems will contain hundreds of processors in the near future. Already now, it is necessary to study how to manage such systems when dynamically scheduling applications with different phases of parallelism and resource demands. A recent research area called invasive computing proposes a decentralized workload management scheme of such systems: applications may dynamically claim additional processors during execution and release these again, respectively. In this paper, we study how to apply the concepts of invasive computing for realizing decentralized core allocation schemes in homogeneous many-core systems with the goal of maximizing the average speedup of running applications at any point in time. A theoretical analysis based on game theory shows that it is possible to define a core allocation scheme that uses local information exchange between applications only, but is still able to provably converge to optimal results. The experimental evaluation demonstrates that this allocation scheme reduces the overhead in terms of exchanged messages by up to 61.4% and even the convergence time by up to 13.4% compared to an allocation scheme where all applications exchange information globally with each other.

DynOAA — Dynamic offset adaptation algorithm for improving response times of CAN systems

CAN bus systems are used in many industrial control applications, particularly automotive. Due to growing system and functional requirements, the low capacity of the CAN bus and usually strict conditions under which it is used in realtime applications, applicability of CAN bus is severely limited. The paper presents an approach for achieving high utilization and breathes new life to CAN bus based systems by proposing a dynamic offset adaptation algorithm for scheduling messages and improving message response times without any changes to a standard CAN bus. This simple algorithm, which runs on all nodes of the system, results in excellent average response times at all loads and makes the approach particularly attractive for soft real-time systems. We demonstrate the performance improvement of the proposed approach by comparisons to other approaches and introduce a new performance measure in the form of a rating function.

Self-organizing Bandwidth Sharing in Priority-Based Medium Access

In this paper, we present an analysis of self-organizing bandwidth sharing in priority-based medium access. For this purpose, the priority-based Access Game is introduced. Analysis shows that a fair distribution of bandwidth cannot be achieved in this game. Therefore, we enhance this game by introducing a constraint that demands a small amount of the overall bandwidth being free. Fair bandwidth sharing is one Nash Equilibrium of this enhanced game, but not a unique one. Based upon this theoretical analysis, a multi-agent reinforcement learning algorithm is proposed, where each agent tries to maximize its success rate for accessing the medium, while avoiding to violate the bandwidth constraint. We experimentally evaluate this mechanism for a system comprised of selfish agents. Experimental results show that the system is able to self-organize itself towards a fair distribution of bandwidth in a totally decentralized way without the need of global information or coordination.

A Self-Organizing Distributed Reinforcement Learning Algorithm to Achieve Fair Bandwidth Allocation for Priority-Based Bus Communication

Due to the raising complexity in distributed embedded systems, a single designer will not be able to plan and organize the communication for such systems. Therefore, it will get more and more important to relieve the designer in that task. Our idea is a communication system that is capable to organize itself to satisfy predefined properties. In this paper, we want to solve the problem of establishing fair bandwidth sharing on priority-based buses by using simple local rules on the distributed system to avoid a single point of failure and cope with online system changes. Based on a game theoretical analysis, a multi-agent reinforcement learning algorithm is proposed that establishes fair bandwidth distribution. The main idea is to penalize nodes that claim too much bandwidth by the other nodes. We experimentally evaluated the algorithm with different parameter settings. The algorithm showed to converge to a fair solution in any experiment. This means the system is able to completely self-organize without global information for our assumptions. In addition, we could figure out that we can configure a trade-off between convergence speed and computation effort. We hope this is a small first step towards totally self-organizing real-time systems.

Self-organizing multi-cue fusion for FPGA-based embedded imaging

Self-organization is a natural concept that helps complex systems to adapt themselves autonomically to their environment. In this paper, we present a self-organizing framework for multi-cue fusion in embedded imaging. This means that several simple image filters are used in combination to lead to a more robust system behavior. Human motion tracking serves as a show case. The system adapts to changes in the environment while tracking a person. Besides this, system customization can be simplified. The designer just has to select a desired set of image filters for a given task. The system then finds the appropriate parameters, e.g., the weighting of different cues. With the option of partial re-configuration, FPGAs support this type of customization. An FPGA-based prototype implementation demonstrates the feasibility of this approach. Tracking and adaptation work in real-time with 25 FPS and a resolution of 640×480.

Distributed self-organizing bandwidth allocation for priority-based bus communication

The raising complexity in distributed embedded systems makes it necessary that the communication of such systems organizes itself automatically. In this paper, we tackle the problem of sharing bandwidth on priority‐based buses. Based on a game theoretic model, reinforcement learning algorithms are proposed that use simple local rules to establish bandwidth sharing. The algorithms require little computational effort and no additional communication. Extensive experiments show that the proposed algorithms establish the desired properties without global knowledge ortextita priori information. It is proven that communication nodes using these algorithms can co‐exist with nodes using other scheduling techniques. Finally, we propose a procedure that helps to set the learning parameters according to the desired behavior. Copyright

An FPGA implementation of a threat-based strategy for Connect6

In this paper, we present a strategy and an FPGA implementation of a Connect6 player submitted to the FPT 2011 Design Competition. Connect6 is a two-player strategy board game. The winner of the game is the player who first gets six pieces of his color in a connected horizontal, vertical or diagonal line. We assign a strategic value to each potential move depending on the current board configuration. Our approach uses a minimal amount of situation dependent game logic in order to take full advantage of the available compute resources and parallelism. The FPGA implementation of this strategy always wins against the software opponent provided for the competition. Additionally, our implementation wins on average against different software AIs from [1], as long as no sophisticated game-tree search is performed by the software.

Adaptive traffic scheduling techniques for mixed real-time and streaming applications on reconfigurable hardware

With the ongoing development of new FPGA generations, the reconfiguration time decreases and therefore the benefit of runtime reconfiguration increases. In this paper, we describe how to use runtime reconfiguration to improve the efficiency of transmitting streaming data on a communication channel shared with real-time applications. This means, the bandwidth that the streaming data has available is dynamically changing. To use the bandwidth effectively, different modules can be loaded on the reconfigurable hardware. These modules have a tradeoff between bandwidth and area requirements. The target now is to find an optimal reconfiguration schedule that minimizes an objective function consisting of two conflicting objectives: reducing the average area needed and providing a certain quality of transmission. In this paper, a model for this scheduling problem is presented and an Integer Linear Programming (ILP) formulation is introduced to calculate an optimal offline solution for benchmarking. In addition, an online scheduling system is presented. It uses the current delay of the streaming application to calculate the schedule. Extensive simulations have been made to show the benefits of the proposed solution.