Computational-Complexity Comparison of Artificial Neural Network and Volterra Series Transfer Function for Optical Nonlinearity Compensation

Otsuka, Yuta; Fukumoto, Yuta; Owaki, Shotaro; Nakamura, Moriya

doi:10.1109/oecc.2018.8730100

Cited by 7 publications

(9 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From computational aspect, the computation time of an ANN depends on the number of layers and node counts at each layer. For a three‐layered feed‐forward NN, computational complexity is given as the following 53,54 :

O (i \cdot j + j \cdot k),

…”

Section: Neural Networkmentioning

confidence: 99%

Neural network‐based energy management of multi‐source (battery/UC/FC) powered electric vehicle

2020

View full text Add to dashboard Cite

Due to increased environmental pollution and global warming concerns, the use of energy storage units that can be supported by renewable energy resources in transportation becomes more of an issue and plays a vital role in terms of clean energy solutions. However, utilization of multiple energy storage units together in an electric vehicle makes the powertrain system more complex and difficult to control. For this reason, the present study proposes an advanced energy management strategy (EMS) for range extended battery electric vehicles (BEVs) with complex powertrain structure. Hybrid energy storage system (HESS) consists of battery, ultra-capacitor (UC), fuel cell (FC) and the vehicle is propelled with two complementary propulsion machines. To increase powertrain efficiency, traction power is simultaneously shared at different rates by propulsion machines. Propulsion powers are shared by HESS units according to following objectives: extending battery lifetime, utilizing UC and FC effectively. Primarily, to optimize the power split in HESS, a convex optimization problem is formulated to meet given objectives that results 5 years prolonged battery lifetime. However, convex optimization of complex systems can be arduous due to the excessive number of parameters that has to be taken into consideration and not all systems are suitable for linearization. Therefore, a neural network (NN)-based machine learning (ML) algorithm is proposed to solve multi-objective energy management problem. Proposed NN model is trained with convex optimization outputs and according to the simulation results the trained NN model solves the optimization problem within 92.5% of the convex optimization one.

show abstract

O (i \cdot j + j \cdot k),

…”

Section: Neural Networkmentioning

confidence: 99%

Neural network‐based energy management of multi‐source (battery/UC/FC) powered electric vehicle

2020

View full text Add to dashboard Cite

show abstract

“…The non‐linearity of the optical fibre can be approximately represented using a third‐order VSTF and expressed as

\begin{matrix} right leftthickmathspace.5em \\ y (n) = & {false\sum}_{m_{1} = - N}^{N} h_{m_{1}} x false(n - m_{1} false) \\ + {false\sum}_{m_{1} = - N}^{N} \sum_{m_{2} = - N}^{N} h_{m_{1} m_{2}} x (x - m_{1}) x (x - m_{2}) \\ + {false\sum}_{m_{1} = - N}^{N} \sum_{m_{2} = m_{1}}^{N} {false\sum}_{m_{3} = - N}^{N} h_{m_{1} m_{2} m_{3}} x false(n - m_{1} false) x false(n - m_{2} false) x^{*} false(n - m_{3} false), \end{matrix}

where x ( n ) and y ( n ) are the input and output of the VSTF at time index n , respectively [1, 2];

h_{m_{1}}

h_{m_{1} m_{2}}

and

h_{m_{1} m_{2} m_{3}}

are the first‐, second‐ and third‐order Volterra kernels, respectively; and L = 2 N + 1 expresses the tap length of the VSTF, as shown in Fig. 1 b [6, 7].…”

Section: Ann/vstf and Overestimation Problemmentioning

confidence: 99%

“…With this scheme, ANN‐based digital signal processing is employed in the time domain [3, 4] or the frequency domain [5] to compensate for the optical non‐linearity. We reported the advantages of ANN over VSTF in terms of the computational complexity [6, 7]. Recently, however, the problem of overestimation in an ANN has been reported; specifically, it has been pointed out that an ANN can potentially learn to predict the pseudo‐random binary sequence (PRBS) pattern, resulting in overestimation of the system performance [8, 9].…”

Section: Introductionmentioning

confidence: 99%

“…ANN/VSTF and overestimation problem: Fig. 1a shows the construction of a three-layer ANN which had been used in our investigations of ANN-based non-linear equalisers [3,6,7]. The input layer of the ANN has a feedforward tapped delay line.…”

mentioning

confidence: 99%

“…where x(n) and y(n) are the input and output of the VSTF at time index n, respectively [1,2]; h m1 , h m1m2 and h m1m2m3 are the first-, second-and third-order Volterra kernels, respectively; and L = 2N + 1 expresses the tap length of the VSTF, as shown in Fig. 1b [6,7].…”

mentioning

confidence: 99%

See 2 more Smart Citations

Overestimation problem with ANN and VSTF in optical communication systems

et al. 2019

View full text Add to dashboard Cite

The authors investigated the problem of overestimation with the Volterra series transfer function (VSTF) and an artificial neural network (ANN), which are used for non-linear equalisers in optical communication systems. The results revealed that the risk of predicting a pseudo-random binary sequence (PRBS) pattern, which causes overestimation of the equaliser performance, occurs not only with an ANN but also with the VSTF. When using PRBS9, PRBS11 and PRBS15, the number of taps of a feedforward tapped delay line, which is required in the VSTF to predict the PRBS pattern, was the same as that with the ANN. When the second-order Volterra kernels were omitted, a larger number of taps was required in the VSTF to observe the overestimation. a EVM versus the number of taps in ANN-based non-linear equaliser b EVM versus the number of taps in VSTF-based non-linear equaliser using first-, second-and third-order Volterra kernels c EVM versus the number of taps in VSTF with first-, second-and third-order Volterra kernels and VSTF with firstand third-order Volterra kernels (the VSTFs were trained on PRBS9) Conclusion: We investigated the problem of overestimation in the ANN-and VSTF-based non-linear equalisers in PRBS-based signal quality evaluation. The results revealed that a VSTF can predict PRBS patterns, and the overestimation problem occurs not only with the ANN but also with the VSTF. In our investigation using PRBS9, PRBS11 and PRBS15, the number of taps that the VSTF required to predict the PRBS pattern was the same as that with the ANN.

show abstract

Overfitting of ANN-based Nonlinear Equalizer for Multilevel Signals in Optical Communication Systems

Ikuta

Otsuka

Nakamura

2020

2020 Opto-Electronics and Communications Conference (OECC)

View full text Add to dashboard Cite

Computational-Complexity Comparison of Artificial Neural Network and Volterra Series Transfer Function for Optical Nonlinearity Compensation

Cited by 7 publications

References 6 publications

Neural network‐based energy management of multi‐source (battery/UC/FC) powered electric vehicle

Neural network‐based energy management of multi‐source (battery/UC/FC) powered electric vehicle

Overestimation problem with ANN and VSTF in optical communication systems

Overfitting of ANN-based Nonlinear Equalizer for Multilevel Signals in Optical Communication Systems

Contact Info

Product

Resources

About