Scholarlink

<p>In visible light communication systems, nonlinear distortion and memory effects caused by light-emitting diodes have been a non-negligible factor limiting system performance. To eliminate such nonlinear and memory effects, a postdistortion receiver based on radial basis function extended (RBFE) interpolation has been proposed. However, this scheme requires the least squares to estimate coefficients, which involves high-dimensional matrix inversion, resulting in unacceptable circuit overhead and power consumption. Aiming at this problem, we propose an RBFE interpolation scheme based on recursive least squares (RLS) to reduce the complexity of training stage. By using RLS to iteratively update the weights, the proposed scheme reduces the computational complexity from <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(O\left( {{N^3}} \right)\)</EquationSource> </InlineEquation> to <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(O\left( {{N^2}} \right)\)</EquationSource> </InlineEquation>. Moreover, computer simulation results show that our proposed scheme achieves the comparable performance to RBFE scheme, so the proposed RBFE-RLS scheme could be applicable for practical use because of the low complexity and fast adaptive convergence speed.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Low Complexity Radial Basis Function Extension Interpolation Postdistortion Algorithm Based on Recursive Least Squares for Visible Light Communications

  • 期刊论文

  • 发表日期：

    2025年12月15日

查看全文
  • Zihan Kang,
  • Jieling Wang,
  • Yuancheng Dang,
  • Ni Chen,
  • Ba-Zhong Shen

摘要

In visible light communication systems, nonlinear distortion and memory effects caused by light-emitting diodes have been a non-negligible factor limiting system performance. To eliminate such nonlinear and memory effects, a postdistortion receiver based on radial basis function extended (RBFE) interpolation has been proposed. However, this scheme requires the least squares to estimate coefficients, which involves high-dimensional matrix inversion, resulting in unacceptable circuit overhead and power consumption. Aiming at this problem, we propose an RBFE interpolation scheme based on recursive least squares (RLS) to reduce the complexity of training stage. By using RLS to iteratively update the weights, the proposed scheme reduces the computational complexity from \(O\left( {{N^3}} \right)\) to \(O\left( {{N^2}} \right)\) . Moreover, computer simulation results show that our proposed scheme achieves the comparable performance to RBFE scheme, so the proposed RBFE-RLS scheme could be applicable for practical use because of the low complexity and fast adaptive convergence speed.