Abstract
The Particle Swarm Optimization (PSO) is a heuristic search method inspired by different biological populations on their swarming or collaborative behavior. This novel work has implemented PSO for the Travelling Salesman Problem (TSP) in high-level synthesis to reduce the computational time latency. The high-level synthesis design generates an estimation of the hardware resources needed to implement the PSO algorithm for TSP on FPGA. The targeted FPGA of this algorithm is the Xilinx Zynq family. The algorithm has been implemented for getting the best route between 5 given cities with given distances. The research has used 7 number of particles for a different number of iterations for generating the best route between those 5 cities. The overall latency has been reduced due to the applied optimization techniques. This paper also implemented and parallelized the same algorithm on CPU Intel I7 Processor; the result shows the FPGA implementation gives better results than CPU on the comparison of performance.
Highlights
Particle Swarm Optimization (PSO) is an algorithm which is been adapted from the unpredictable flight of the bird’s flock
The Zynq 7000 architecture consists of Processing System (PS), which is programmable dual-core ARM Cortex A9 Processor and Programmable Logic (PL), which is the 7 series Xilinx Field Programmable Gate Array (FPGA) Core with resources as look up table (LUT), FIFO, BRAM, DSP and IO’s [15]
The goal of this research is on solving the simple traveling salesman problem (TSP) with the PSO in High Level Synthesis, HLS allows writing an algorithm on C/C++ or OpenCL language
Summary
Particle Swarm Optimization (PSO) is an algorithm which is been adapted from the unpredictable flight of the bird’s flock. Qang et al [3] have implemented PSO for job scheduling application They encoded each particle with a natural number vector and have developed an own approach to move particles in the solution space. They compared the genetic algorithm (GA) with the PSO for job scheduling application and they found that PSO is very competitive with the GA. This implementation uses simulated annealing (SA) method for slow down the degeneration of the swarm and increase the swarm diversity They compared the PSO with SA, basic Genetic Algorithm (GA) and two other algorithms for solving TSP problem in which the PSO with SA gives the superior results than the other methods. This will help in finding the best route in a high speed
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
