A state-constrained model for cellular nonlinear network optimization

Abstract

A G m C-style state constrained neuron (SCN) model for the design of processors in analog recurrent neural networks such as Hopfield neural networks, cellular nonlinear networks for combinatorial optimization is described. The unconstrained neurons which have the free state variable, could be stable at any arbitrary point in the solution space or trapped by un-intentional effects. These may introduce errors. For the unconstrained network, the solution could be different from the expected one due to the discrepancy in the energy function of the network and the objective function to be optimized. In addition, if the state variable is limited by some neighboring saturated transistors, un-desirable results may be obtained. GmC-style SCN model can ensure the convergence of the network and avoid discrepancy between the energy function of the network and the objective function. The state resistor is also eliminated in the G m C model so that high cell-density can be achieved. Simulation results show that the proposed model is effective in significantly reducing optimization error.

A compact neural network for VLSI PRML detectors: scalable architecture

“…The annealing gain controlling signal serves as an input to the transconductive multiplier. The gain controlling signal can be a sawtooth signal sweeping from 0 to 1 V. The governing equation for the state variable can be described as (7) Then, (8) where is the voltage on the state capacitor of th neuron, is the transconductance of the multiplier, is the voltage of the received data sample in the range of [ 1 V, 1 V], is the time constant of the neuron, and are the connection matrix coefficients and the input correlator coefficients, respectively. A transconductive processor is used to convert the voltage-mode state variable into the current mode, , with the saturation property of the sigmoidal characteristics.…”

“…A state constrained neuronal (SCN) model [7] which effectively solves the out-of-bound equilibrium problem and the inaccurate state resistance problem, is used to accelerate parallel-distributed optimization. Such a parallel architecture has the following important features:…”