Parallel scalable hardware implementation of asynchronous discrete particle swarm optimization

Farmahini-Farahani, Amin; Vakili, Shervin; Fakhraie, S.M.; Safari, Saeed; Lucas, Caro

doi:10.1016/j.engappai.2009.12.001

Cited by 29 publications

(14 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although this approach has more flexibility, improvement in the execution speed is less. Later, an architecture for PSO algorithm was proposed in which an on-chip multiprocessing based on system-on-a-programmable-chip (SOPC) methodology is used to realize the PSO algorithm [16]. Later a hardware accelerator for PSO algorithm was reported [18] and validated its [20].…”

Section: Literature Surveymentioning

confidence: 99%

Comparative study of system on chip based solution for floating and fixed point differential evolution algorithm

Peesapati

Anumandla

Kudikala

et al. 2014

Swarm and Evolutionary Computation

View full text Add to dashboard Cite

Section: Literature Surveymentioning

confidence: 99%

Comparative study of system on chip based solution for floating and fixed point differential evolution algorithm

Peesapati

Anumandla

Kudikala

et al. 2014

Swarm and Evolutionary Computation

View full text Add to dashboard Cite

“…The digital circuits act as the controller that provides a time sequence of signals for initiating the operations in the data path and also determine the next state of the control sub system. A pipelined architecture is required for computational operations of algorithms [28]. The pipeline level and the schedules of neural-controller computation process flow are given in Table 4 and the pipeline stages are marked (lower case letter) in Fig.…”

Section: Design Approach and Behavioral Modeling Of Neural-controllermentioning

confidence: 99%

Design of Cognitive Decision Making Controller for Autonomous Online Adaptive Beam Steering in Free Space Optical Communication System

Raj¹,

Selvi²,

Kumar

et al. 2015

Wireless Pers Commun

View full text Add to dashboard Cite

The beam motion is one of the main causes for major power loss which severely degrades the performance of the Free Space Optical Communication (FSOC) system. Designing a suitable controller to correct the beam motion to increase the beam stability at a point becomes significant. This paper presents an investigation on the performance of two types of controller designed for aiming a laser beam to be at a particular spot under dynamic disturbances generated at the transmitter. The experimental nonlinear input-output data mapping is used as the principal components for the controller design. The first design is based on the Taguchi method while the second is the artificial neural network-cognitive method. These controllers process the beam pointing (beam location) information from an opto-electronic position detector and then generate the necessary outputs to steer the beam with fast steering mirror. Conceptually, the controller design problem is addressed with looking the position error information to decide the level of nonlinear static output map for generating the correction control information. The pipelined-parallel architecture of both controllers are developed in a field programmable gate array (FPGA) and installed at the receiver station. Evidence of the suitability and the effectiveness of the proposed controller in terms of prediction exactness, response to impulse, scintillation, Q-factor and bit error rate are provided through experimental results obtained from the 155 Mbps FSOC data link set-up established for the horizontal range of 0.5 km at an altitude of 15.25 m.

show abstract

“…It also contains some memory blocks used to keep vectors and values of PSO such as particle location, particle velocity, and pbest vectors. The internal structure of the core can be found in [5,6]. Location vectors of particles (weights) are updated using this core.…”

Section: Pso Corementioning

confidence: 99%

Scalable Architecture for on-Chip Neural Network Training using Swarm Intelligence

Farmahini-Farahani¹,

Fakhraie²,

Safari³

2008

2008 Design, Automation and Test in Europe

Self Cite

View full text Add to dashboard Cite

This paper presents a novel architecture for on-chip neural network training using particle swarm optimization (PSO). PSO is an evolutionary optimization algorithm with a growing field of applications which has been recently used to train neural networks. The architecture exploits PSO algorithm to evolve network weights as well as a method called layer partitioning to implement neural networks. In the proposed method, a neural network is partitioned into groups of neurons and the groups are sequentially mapped to available functional units. Thus, the architecture is reconfigurable for training and implementing different multilayer feedforward neural networks without the need for modifying the architecture. The implementation is intended for real-time applications regarding hardware cost and speed. The results show that the proposed system provides a trade-off between resource requirements and speed.

show abstract

Parallel scalable hardware implementation of asynchronous discrete particle swarm optimization

Cited by 29 publications

References 33 publications

Comparative study of system on chip based solution for floating and fixed point differential evolution algorithm

Comparative study of system on chip based solution for floating and fixed point differential evolution algorithm

Design of Cognitive Decision Making Controller for Autonomous Online Adaptive Beam Steering in Free Space Optical Communication System

Scalable Architecture for on-Chip Neural Network Training using Swarm Intelligence

Contact Info

Product

Resources

About