Mathias Pacher

An Agressive Strategy for an Artificial Hormone System to Minimize the Task Allocation Time

We present an aggressive task allocation strategy for an Artificial Hormone System (AHS). The AHS is a completely decentralized operation principle for a middleware which can be used to allocate tasks in a system of heterogeneous processing elements (PEs) or cores. Tasks are scheduled according to suitability of the heterogeneous PEs, current PE load and task relationships. In addition, the AHS provides properties like self-configuration, self-optimization and self-healing by task allocation. The AHS is able to guarantee real-time bounds regarding these self-X-properties. The aggressive task allocation strategy presented in this paper allows to halve the worst case execution times for the self-X-properties compared to previous strategies thus improving the suitability of the AHS for hard real-time systems.

The Artificial Hormone System—An Organic Middleware for Self-organising Real-Time Task Allocation

This article presents an artificial hormone system for a completely decentralised realisation of self-organising task allocation. We show tight upper bounds for the real-time behaviour of self-configuration and self-healing. We also present stability criteria and a AHS implementation coded in pure ANSI C together with some real-world scenario test series and results.

Real-Time Distribution of Time-Dependant Tasks in Heterogeneous Environments

In this paper we present a new approach to distribute tasks connected by causal dependencies within a heterogeneous environment, e.g. several resources communicating with each other or a processor grid. Our approach uses an artificial hormone system for task distribution which is able to meet real-time constraints. Several enhancements of the artificial hormone system are made such as partial suppressing of tasks and distributed task termination.

A Generalized Model to Control the Throughput in a Processor for Real-Time Applications

In this paper we present a control theory approach to stabilize the throughput of threads for real-time applications on a multithreaded processor. We use a statistical model of a super scalar, multi-threaded processor as transfer function to calculate the resulting IPC rate. Our control theory approach is not limited to a specific processor and can be adapted to different microprocessor architectures. We are able to guarantee a minimum IPC rate within a defined convergence interval. Furthermore our results provide a method to improve WCET analysis, because inaccuracies of the processor model are soften by the use of our control theory approach.

Evaluation and superiority analysis of a decentralized task control mechanism for dependable real-time SoC architectures

This article presents our concept of an artificial hormone system for realizing a completely decentralized self-organizing and real-time capable task control mechanism using self-X properties. Besides the fundamentals of the prior hormone concept and the implementation model, we present latest results of our research: evaluation and superiority analysis of a AHS-controlled SoC towards other approaches in centralized or partly decentralized manner like feedback controllers and complex multi-agents. Furthermore we validate and compare the overheads for size, communication and computation in relation to the improvement in system reliability.

Towards a Statistical Model of a Microprocessor's Throughput by Analyzing Pipeline Stalls

In this paper we model a threads throughput, the instruction per cycle rate (IPC rate), running on a general microprocessor as used in common embedded systems. Our model is not limited to a particular microprocessor because our aim is to develop a general model which can be adapted thus fitting to different microprocessor architectures. We include stalls caused by different pipeline obstacles like data dependencies, branch misprediction etc. These stalls involve latency clock cycles blocking the processor. We also describe each kind of stall in detail and develop a statistical model for the throughput including the entire processor pipeline.

Implementation and evaluation of a self-organizing artificial hormone system to assign time-dependent tasks

We present a new method to assign tasks, which are connected by time dependencies within a heterogeneous environment, (e.g., several resources communicating with each other). Our approach for task distribution uses an artificial hormone system, which is self‐organizing and able to meet real‐time constraints. Several enhancements to the artificial hormone system have been made, such as partial suppressing of tasks and distributed task termination.Copyright

A Theoretical Examination of a Self-Adaptation Approach to Improve the Real-Time Capabilities in Multi-Threaded Microprocessors

Our aim is to investigate the possibility to control the throughput (IPC rate) of a thread running on a multithreaded microprocessor by a closed feedback loop. We showed in previous experimental studies the practability of this approach. In this paper we discuss the control theory approach from a mathematical point of view. We develop a formal model of a general purpose multi-threaded microprocessor enhanced with a closed feedback controller and use control theory methods to investigate properties like stability and settling time.

Detection and Defense Strategies against Attacks on an Artificial Hormone System Running on a Mixed Signal Chip

The Artificial Hormone System (AHS) is a self organizing system which allocates tasks to processing elements. It works in a distributed way, is able to hold real-time conditions and can run in a mixed-signal chip environment significantly increasing system reliability. Yet the hormone mechanisms offer new ways for malicious attacks which can affect the correct functioning of the AHS. Such attacks may cause severe damage if the AHS is used in an embedded (maybe real-time) environment. Therefore, this paper deals with analyzing malicious attacks on the AHS. We present several ways of attacking the AHS and resulting detection and defense strategies. We also evaluate these strategies in the paper and demonstrate that they can help to protect the AHS from attacks.

Two-level extensions of an artificial hormone system

The ARTIFICIAL HORMONE SYSTEM (AHS) is a decentralized software that is able to allocate tasks in a system of heterogeneous processing elements (PEs). Tasks are allocated according to their suitability for the heterogeneous PEs, the current PE load, and task relationships. The AHS also provides properties like self‐configuration, self‐optimization, and self‐healing in the context of task allocation. In addition, it is able to guarantee real‐time bounds for such self‐X properties. However, using self‐organization principles introduces increased system complexity such as control of system parameters for self‐organization and additional communication effort. In this contribution, we address these problems by using two different two‐level extensions of the AHS: we consider the choice and control of hormone parameters by an observer/controller architecture as first extension of the AHS and a HIERARCHICAL AHS (HAHS) to save communication bandwidth as second extension of the AHS. For the first extension, we use a machine learning approach for gradually learning the hormone values of different tasks. This is a major advance because expert knowledge is needed to configure the AHS up to now. We present an observer/controller architecture as an extension of the AHS to monitor and control its behavior. The user has to provide a simple set of initial rules, and the observer/controller is able to generate new rules if needed. The evaluation of our approach using a benchmark (containing six different types of tasks) shows that the observer/controller is able to match the goals provided by the user, and we discuss it in detail. The second extension is the hierarchical AHS where the PEs of the system consist of several different clusters each of them having its own communication infrastructure, for example, a bus system. The HAHS is able to save communication effort because the broadcast communication of the hormones is limited to the clusters. Nevertheless, it provides the same self‐X properties as the AHS, even the time for self‐configuration is shorter than for the AHS. We present evaluations that show that the HAHS performs better for applications with large numbers of PEs than the AHS. Copyright