Most evolutionary approaches to fault recovery in FPGAs focus on evolving alternative logic configurations as opposed to evolving the intra-cell routing. Since the majority of transistors in a typical FPGA are dedicated to interconnect, nearly 80% according to one estimate, evolutionary fault-recovery systems should benefit by accommodating routing. In this paper, we propose an evolutionary fault-recovery system employing a genetic representation that takes into account both logic and routing configurations. Experiments were run using a software model of the Xilinx Virtex FPGA. We report that using four Virtex combinational logic blocks, we were able to evolve a 100% accurate quadrature decoder finite state machine in the presence of a stuck-at-zero fault.

Abstract—We have developed a genetic algorithm (GA) approach to the problem of task scheduling for multiprocessor systems. Our approach requires minimal problem specific information and no problem specific operators or repair mechanisms. Key features of our system include a flexible, adaptive problem representation and an incremental fitness function. Comparison with traditional scheduling methods indicates that the GA is competitive in terms of solution quality if it has sufficient resources to perform its search. Studies in a nonstationary environment show the GA is able to automatically adapt to changing targets. Index Terms—Genetic algorithm, task scheduling, parallel processing. 1

We hypothesize that evolutionary algorithms can effectively schedule coordinated fleets of Earth observing satellites. The constraints are complex and the bottlenecks are not well understood, a condition where evolutionary algorithms are often effective. This is, in part, because evolutionary algorithms require only that one can represent solutions, modify solutions, and evaluate solution fitness.

Abstract not found

Most evolutionary approaches to fault recovery in FPGAs focus on evolving alternative logic configurations as opposed to evolving the intra-cell routing. Since the majority of transistors in a typical FPGA are dedicated to interconnect, nearly 80% according to one estimate, evolutionary faultrecovery systems should benefit by accommodating routing. In this paper, we propose an evolutionary fault-recovery system employing a genetic representation that takes into account both logic and routing configurations. Experiments were run using a software model of the Xilinx Virtex FPGA. We report that using four Virtex combinational logic blocks, we were able to evolve a 100% accurate quadrature decoder finite state machine in the presence of a stuck-at-zero fault. Evolutionary experiments with the hardware in the loop have begun and we discuss the preliminary results.

of the work presented in this thesis has been previously published as listed below. Although some of these papers have co-authors, the work appearing in this thesis is entirely my own, with the exception of parts of chapter 3, which presents work jointly carried out by myself and Adrian Thompson. The respective contributions to this work will be explicitly stated at the beginning of the chapter. List of Previous Publications Kuntz, P., Layzell, P., & Snyers, D. (1997). A Colony of Ant-like Agents for Partitioning

We present results from a study comparing a recently developed coevolutionary genetic algorithm (CGA) against a set of evolutionary algorithms using a suite of multiobjective optimization benchmarks. The CGA embodies competitive coevolution and employs a simple, straightforward target population representation and fitness calculation based on developmental theory of learning. Because of these properties, setting up the additional population is trivial making implementation no more di#- cult than using a standard GA. Empirical results using a suite of two-objective test functions indicate that this CGA performs well at finding solutions on convex, nonconvex, discrete, and deceptive Pareto-optimal fronts, while giving respectable results on a nonuniform optimization. On a multimodal Pareto front, the CGA yields poor coverage across the Pareto front, yet finds a solution that dominates all the solutions produced by the eight other algorithms.

Yagi-Uda antennas are known to be di#cul to design and optimize due to their sensitivity at high gain, and theincl#wY4 of numerous parasitic elVYy ts. We present a genetic alzyzw4Iw#VWy4 automated antenna optimization system that uses a fixed Yagi-Uda topolwY and a byte-encoded antenna representation. The fitnesscals4YWBWw al l ws the implxBB relBB4IwVBB between power gain andsidely e/backlc e lw# to emergenaturalV , a technique that isl esscomply than previous approaches. The genetic operators used are alY simplYY OurresulB s inclYz Yagi-Uda antennas that have excelfz t bandwidth and gain properties with very good impedance characteristics. ResulY exceeded previous Yagi-Uda antennas produced via evolYwzzy4I alYWBzz4I by at lw#V 7.8% in mainlY e gain. WealW present encouraging prelaging4 resula where a coevolVwfWW4I geneticaltic4Yx is used.

The following paper describes a cooperative coevolutionary algorithm which incorporates a novel collaboration formation mechanism.

We describe a novel GA approach to partition programs to be executed on a parallel system. Two unique features distinguish this GA from traditional GA programs. First, this GA uses a dynamically incremental fitness function which starts out rewarding for simpler goals, gradually increasing the difficulty of the desired fitness values or goals until a full solution is found. Second, this GA uses a flexible representation style which allows the GA itself more control over both the structure and the value of the evolved solutions.