Novel hysteretic noisy chaotic neural network for broadcast scheduling problems in packet radio networks

  • Authors:
  • Ming Sun;Lin Zhao;Wei Cao;Yaoqun Xu;Xuefeng Dai;Xiaoxu Wang

  • Affiliations:
  • College of Computer and Control Engineering, Qiqihar University, Qiqihar, Heilongjiang Province, China;College of Automation, Harbin Engineering University, Harbin, Heilongjiang Province, China;College of Computer and Control Engineering, Qiqihar University, Qiqihar, Heilongjiang Province, China;Institute of System Engineering, Harbin University of Commerce, Harbin, Heilongjiang Province, China;College of Computer and Control Engineering, Qiqihar University, Qiqihar, Heilongjiang Province, China;College of Automation, Northwestern Polytechnical University, Xi'an, Shanxi Province, China

  • Venue:
  • IEEE Transactions on Neural Networks
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Noisy chaotic neural network (NCNN), which can exhibit stochastic chaotic simulated annealing (SCSA), has been proven to be a powerful tool in solving combinatorial optimization problems. In order to retain the excellent optimization property of SCSA and improve the optimization performance of the NCNN using hysteretic dynamics without increasing network parameters, we first construct an equivalent model of the NCNN and then control noises in the equivalent model to propose a novel hysteretic noisy chaotic neural network (HNCNN)- Compared with the NCNN, the proposed HNCNN can exhibit both SCSA and hysteretic dynamics without introducing extra system parameters, and can increase the effective convergence toward optimal or near-optimal solutions at higher noise levels. Broadcast scheduling problem (BSP) in packet radio networks (PRNs) is to design an optimal time-division multiple-access (TDMA) frame structure with minimal frame length, maximal channel utilization, and minimal average time delay. In this paper, the proposed HNCNN is applied to solve BSP in PRNs to demonstrate its performance. Simulation results show that the proposed HNCNN with higher noise amplitudes is more likely to find an optimal or near-optimal TDMA frame structure with a minimal average time delay than previous algorithms.