|
|
Concurrent evolutionary design of hardware and software
Minařík, Miloš ; Sekaj, Ivan (oponent) ; Squillero, Giovanni (oponent) ; Sekanina, Lukáš (vedoucí práce)
Genetic programming (GP) can, to some extent, automatically generate desired programs without asking the user to specify how to do it. It has been used to solve a wide range of practical problems and produce a number of human-competitive results in different fields. An interesting and practically untouched question is whether for a given problem, GP can generate a highly optimized programmable computational model (platform) together with a program running on the platform, solving the problem and satisfying all constrains such as on the area on a chip and speed. In a multi-objective scenario, the user would obtain a set of non-dominated solutions showing various tradeoffs between resources (the area, power consumption) and performance (the speed of execution). This problem can be seen as a concurrent development of hardware and software, simply, HW/SW codesign. This thesis explores the ways how to evolve hardware platforms together with programs in the case that the specification is given in terms of a set of input-output vectors. The initial model of the architecture was created and the evolutionary framework capable of maintaining and evolving the population of such architectures was implemented. Candidate microprogrammed architectures were evolved together with programs using extended linear genetic programming. Several simple experiments were carried out and the framework proved competitive with state-of-the-art methods. The framework was subsequently extended addressing the weak points identified during the initial experiments. The extended framework was validated by means of more complex experiments. One of them focused on an effective implementation of sigmoid function approximation. Various implementations of sigmoid approximation were evolved (sequentional as well as purely combinational). The proposed framework provided several well-known solutions and even optimized some of them for the particular input domain chosen for the experiment. The next set of experiments was supposed to evolve an image filter reducing salt-and-pepper impulse noise. The framework was able to evolve the concept of switching-based filter and even the variation of a switching-based median filter comparable to the filters commonly used. This thesis proved that small-size HW/SW systems can be designed and optimized by means of genetic programming. Moving to an automated evolutionary design of more complex HW/SW systems is an open research problem waiting for a future research.
|
|
|
Evolutionary design and optimization of components used in high-speed computer networks
Grochol, David ; Sekaj, Ivan (oponent) ; Jašek, Roman (oponent) ; Sekanina, Lukáš (vedoucí práce)
The research presented in this thesis is directed toward the evolutionary optimization of selected components of network applications intended for high-speed network monitoring systems. The research started with a study of current network monitoring systems. As an experimental platform, the Software Defined Monitoring (SDM) system was chosen. Because traffic processing is an important part of all monitoring systems, it was analyzed in greater detail. For detailed studies conducted in this thesis, two components were selected: the classifier of application protocols and the hash functions for network flow processing. The evolutionary computing techniques were surveyed with the aim to optimize not only the quality of processing but also the execution time of evolved components. The single-objective and multi-objective versions of evolutionary algorithms were considered and compared. A new approach to the application protocol classifier design was proposed. Accurate and relaxed versions of the classifier were optimized by means of Cartesian Genetic Programming (CGP). A significant reduction in Field-Programmable Gate Array (FPGA) resources and latency was reported.Specialized, highly optimized network hash functions were evolved by parallel Linear Genetic Programming (LGP). These hash functions provide better functionality (in terms of quality of hashing and execution time) than the state-of-the-art hash functions. Using multi-objective LGP, we even improved the hash functions evolved with the single-objective LGP. Parallel pipelined hash functions were implemented in an FPGA and evaluated for purposes of network flow hashing. A new reconfigurable hash function was developed as a combination of selected evolved hash functions. Very competitive general-purpose hash functions were also evolved by means of multi-objective LGP and evaluated using representative data sets. The multi-objective approach produced slightly better solutions than the single-objective approach. We confirmed that common LGP and CGP implementations can be used for automated design and optimization of selected components; however, it is important to properly handle the multi-objective nature of the problem and accelerate time-critical operations of GP.
|
|
|
Evolutionary design and optimization of components used in high-speed computer networks
Grochol, David ; Sekaj, Ivan (oponent) ; Jašek, Roman (oponent) ; Sekanina, Lukáš (vedoucí práce)
The research presented in this thesis is directed toward the evolutionary optimization of selected components of network applications intended for high-speed network monitoring systems. The research started with a study of current network monitoring systems. As an experimental platform, the Software Defined Monitoring (SDM) system was chosen. Because traffic processing is an important part of all monitoring systems, it was analyzed in greater detail. For detailed studies conducted in this thesis, two components were selected: the classifier of application protocols and the hash functions for network flow processing. The evolutionary computing techniques were surveyed with the aim to optimize not only the quality of processing but also the execution time of evolved components. The single-objective and multi-objective versions of evolutionary algorithms were considered and compared. A new approach to the application protocol classifier design was proposed. Accurate and relaxed versions of the classifier were optimized by means of Cartesian Genetic Programming (CGP). A significant reduction in Field-Programmable Gate Array (FPGA) resources and latency was reported.Specialized, highly optimized network hash functions were evolved by parallel Linear Genetic Programming (LGP). These hash functions provide better functionality (in terms of quality of hashing and execution time) than the state-of-the-art hash functions. Using multi-objective LGP, we even improved the hash functions evolved with the single-objective LGP. Parallel pipelined hash functions were implemented in an FPGA and evaluated for purposes of network flow hashing. A new reconfigurable hash function was developed as a combination of selected evolved hash functions. Very competitive general-purpose hash functions were also evolved by means of multi-objective LGP and evaluated using representative data sets. The multi-objective approach produced slightly better solutions than the single-objective approach. We confirmed that common LGP and CGP implementations can be used for automated design and optimization of selected components; however, it is important to properly handle the multi-objective nature of the problem and accelerate time-critical operations of GP.
|
|
|
Concurrent evolutionary design of hardware and software
Minařík, Miloš ; Sekaj, Ivan (oponent) ; Squillero, Giovanni (oponent) ; Sekanina, Lukáš (vedoucí práce)
Genetic programming (GP) can, to some extent, automatically generate desired programs without asking the user to specify how to do it. It has been used to solve a wide range of practical problems and produce a number of human-competitive results in different fields. An interesting and practically untouched question is whether for a given problem, GP can generate a highly optimized programmable computational model (platform) together with a program running on the platform, solving the problem and satisfying all constrains such as on the area on a chip and speed. In a multi-objective scenario, the user would obtain a set of non-dominated solutions showing various tradeoffs between resources (the area, power consumption) and performance (the speed of execution). This problem can be seen as a concurrent development of hardware and software, simply, HW/SW codesign. This thesis explores the ways how to evolve hardware platforms together with programs in the case that the specification is given in terms of a set of input-output vectors. The initial model of the architecture was created and the evolutionary framework capable of maintaining and evolving the population of such architectures was implemented. Candidate microprogrammed architectures were evolved together with programs using extended linear genetic programming. Several simple experiments were carried out and the framework proved competitive with state-of-the-art methods. The framework was subsequently extended addressing the weak points identified during the initial experiments. The extended framework was validated by means of more complex experiments. One of them focused on an effective implementation of sigmoid function approximation. Various implementations of sigmoid approximation were evolved (sequentional as well as purely combinational). The proposed framework provided several well-known solutions and even optimized some of them for the particular input domain chosen for the experiment. The next set of experiments was supposed to evolve an image filter reducing salt-and-pepper impulse noise. The framework was able to evolve the concept of switching-based filter and even the variation of a switching-based median filter comparable to the filters commonly used. This thesis proved that small-size HW/SW systems can be designed and optimized by means of genetic programming. Moving to an automated evolutionary design of more complex HW/SW systems is an open research problem waiting for a future research.
|
host ::
RSS.