Dual AGC Model Implementation of the Inner
Hair Cell and Auditory Nerve IC in euromorphic VLSI

Dr.J.Raja; Saranya.SA; Suresh Babu.T.N

Dual AGC Model Implementation of the Inner Hair Cell and Auditory Nerve IC in euromorphic VLSI

Dr.J.Raja, Saranya.SA ,Suresh Babu.T.N

Professor, Department of ECE, Adhiparasakthi Engineering College, Melmaruvathur, Tamil Nadu, India
PG Student (VLSI Design), Department of ECE, Adhiparasakthi Engineering College, Melmaruvathur, Tamil Nadu, India
Assistant Professor, Department of ECE, Adhiparasakthi Engineering College, Melmaruvathur, Tamil Nadu, India

Related article at Pubmed, Scholar Google

Visit for more related articles at International Journal of Innovative Research in Science, Engineering and Technology

Abstract

ABSTRACT-An analog inner hair cell and auditory nerve has been implemented for the persons with hearing disabilities. The designed circuit uses fully balanced circuits to reduce the mismatch of the signals that enters through the hearing aid. Ultra low power consumption of the circuit in the machine save the battery life to be used for years. CMOS current mode is being implemented to the computationally developing hearing aid IC. The designed IC can be used to obtain a result in spectrogram by measuring the frequencies from articulogram in future. The power consumption of the circuit is reduced dramatically.

Keywords

Auditory Nerve ,dual AGC model Biomimetics, Inner hair cell, Neuromorphic, Very Large Scale Integration (VLSI), Hearing Aid

INTRODUCTION

Neuromorphic aims for the increased performance of the analog electronic circuit and for the artificial intelligence like robotics, smartphones, nano technology etc.,the growing technology results in advances in minimturized systems and also nano systems where the tie up between both is power. Power consumption has become the thrist for the modern technology. In biomedical VLSI plays the major role for developing instruments for major and minor operations. Medical applicances like octolography ,electroenceplograghy, electro cardiogram measures the signals obtained for the patients, the signals whose frequencies are measured through medical devices to obtain the result. An analogue VLSI is adapted to implement in inner hair cell and auditory nerve.Normally human ear consist of outer hair cell(OHC),inner hair cell (IHC) and middle hair cell, the damage of inner hair cell results hearing loss, injection to cochlea in order to replace the damaged inner hair cell results in regeneration hair growth but this cause serious injury. Cochlear implants are to make the disabilities to hear the sound in environment directly without earing aid ,perhaps implanting artificial cochlea produces waxy substances in the ear, these are done in order to have spotless hearing aid. The human ear is a highly sensitive sound receptor in which pressure fluctations in the outer ear are transformed into vibrations of small bones (the ossicles) in the middle ear that are ultimately communicated to the cochlea located in the inner ear ,where the vibrations are further transformed by stereocillia (hair cells) into neural impulses distributed by frequency.

Natural human voice spans a frequency range from 20Hz to 20KHz, however conventional microphone system passes frequencies from 400Hz to 3.5KHz. Therefore phone conversation differs from face-to-face conversation and also the heared by the deaf.

THRESHOLD OF HEARING

The threshold pressure level of a sound is the lowest level at which an observer can discriminate between the desired sound and the noise background always present in the auditory system.The base of each audio logical examination is the determination of the hearing threshold.It is necessary to know whether the hearing lossis due to disorder of the hearing organ (cochlea) and connecting nerves with the brain(perceptive loss) or the loss because of reduced transmission of sound vibrations through the middle ear mechanism (conductive loss).

MEASUREMENT OF SOUND

Sound intensity is the amount of energy flow per unit time through a unit area perpendicular to the direction of energy flow. The common receivers are microphones which do not measure sound intensity directly. They are sensitive to sound pressure therefore it is pertinent to measure the sound in terms of sound pressure. Sound pressure is related to both amplitude and frequency and sound intensity is proportional to the square of sound pressure. The convenient way for expressing sound pressure and sound intensity in terms of decibel.

CURRENT MODE IMPLEMENTATION

The hearing aid consisting the integrated chip which should be used for a period of years with less consumption of power which can be achieved by Very Large Scale integration (VLSI).

HIGH SPEED DIGITAL SIGNAL PROCESSING

The chip designed controls the level of noise.The digital chip was integrated into number of hearing aid with high speed signal processing or DSP. The acoustic signals are sent to signal processing chips to be implemented into the hearing aids. These integrated circuits were produced in hearing aids. One of the major contributions of these chips was the ability to process both speech and other types of noises in real time application. One major down fall of these chips was that they were massive and used up a lot of battery, which made them to nearly impossible to be worn and thus the use of CMOS technology to minimize the consumption of power.

The above fig 5 shows the process of mammalian hear working, the envelope follower consists of half wave rectifier this is buffered with fully balanced current mirror. The function of fully balanced circuit for the is to act as a buffer. The buffered signal enters the next two stages in the dual AGC model. Low pass filtering here filters the unwanted noise. The dual AGC model receives the incoming signal and amplifies the sound that enters the ear. The FBCM provides impedance matching for both multiplier and low pass filter. Reference current level are K1 and K2 are provided by the subtraction which the filtered signal is compared with the normal speech. In hearing aid the circuit IHC and AN changes the intensity of the sounds magnitude of 120 dB in sound pressure level. The dynamic response of the IHC circuit separates the binaural cues from the noisy environment. The alternative IHC is designed in dual AGC model. The model does the working of the normal human ear does of amplification, the hair cells remain unvibrated the sound entering simulates the hair cells generating ion channels within the hair cells.T he transconductance of the vibrated cells changes as the input sound changes.

FULLY BALANCED CURRENT MIRROR

The above Fig.6shows, FBCM consisting of two halves, positive and negative it is composed of Wilson current mirror whose output is given to the complementary current mirror. With a low supply voltage of 1.8V,keeping current 1 (I inN ) as low as 0 and current 2 ( I inP ) as high, keeping the gain to be one. The transistors used in this design, allows consumption of power to be reduced.The collector current does not remain constant with variation in temperature or constant. „S‟ should be small as possible to have better stability.The parametric analysis for the capacitor, this analysis helps to choose the exact value for the parameter here the value is choosen as C=3.258pf in order to obtain the stabilized output current for the circuit. The collector current does not remain constant with variation in temperature or power supply voltage. Therefore the operating point will be unstable.

CMOS CURRENT MIRROR

STABILITY

Fig 8.shows the CMOS current mirror allows the combination of both PMOS and NMOS, this circuit is used in inner hair cell circuit because it rectifies the signal before passing it through the filter. This adaptation has been modelled in but it was consideredtoo complex and too large for inclusion on the current chip of IHC. The analog and digital circuits run on separate power supplies to reduce digital noise and the analog signals. However the proposed IHC with current mirror has overcome this problem and this circuit is intended to include in future versions.

INNER HAIR CELL CIRCUIT

SIMULATION AND RESULTS

and the circuitresponse becomes nonlinear when the input exceeds 1.8A and the outputis connected to a 100kΩ load. The transistor dimensions were optimized forminimizing the area mismatch without significant concern for the DC operating point within the limits. Thus, the PMOS and NMOS dimensions have a equal widths and lengths, assuring their transconductance would be different. The output wave obtained is smaller than the input given thereby the power of the existing FBCM is 122Nw

thus both the combination increases the circuit complexity and power.The power wave rises to a particular level of 2W and reduces at 1.6W.This circuit when combined with IHC increases power.

CONCLUSION

The designed IHC circuit and FBCM is used as a buffer circuit in the chip designed for the hearing aid. Power consumption is the major requirement of the electronic circuit designed in this modern technology. An analog VLSI implementation of the dual AGC model which performs the function as the normal human ear is developed as a chip to place in the ear machine for the person with hearing disabilities this is used inmonaural cue, binaural cue, or multiple binaural systems. Using current-mode circuits the ability totune the system to remove the CMOS mismatch is achieved.Theseresults have been evaluated and established by analog CADENCE Virtuoso environment. The circuits are designed to obtain the power analysis, andto be applied in low-power envelope, real-time processing. And the behavior of flexibility from various levels make the AIHCAN IC a practical module for use in biomimetic processing, especially in smart phones ,battery powered applications and biological systems where the AIHCAN IC can be easily integrated into existing and future systems.

References

David.S Freedman, Member, IEEE, Howard I. CohenSocrates Deligeorges, Christian Karl, and Allyn E.Hubbard Ã¢ÂÂAn analog VLSI m implementation of inner hair cell and auditory nerve using dual AGC modelÃ¢ÂÂ IEEE Trans.Biomed.CircuitsSyst.,vol.5,no.4,pp.240-256,2014

K. H. Wee, L. Turicchia, and R. Sarpeshkar,Ã¢ÂÂAn articulatory silicon vocal tract for speech and hearing prostheses,Ã¢ÂÂ IEEE Trans.Biomed.CircuitsSyst., vol. 5, no. 4, pp. 339Ã¢ÂÂ346, 2011

A. Hubbard, H. Cohen, C. Karl, D. Freedman, D. Mountain, L.Ziph-Schatzberg, M. Nourzad Karl, S. Kelsall, T. Gore, Y. Pu,Z. Yang, X. Xing,and S.Deligeorges, Ã¢ÂÂBiologically inspiredcircuitry that mimics mammalian hearing,Ã¢ÂÂ in Proc. SPIEConf.Ser., 2009, vol.7321.

Y. Shao, Z. Jin, D. Wang, and S.Srinivasan, Ã¢ÂÂAn auditory-based feature for robust speech recognition,Ã¢ÂÂ in Proc. IEEE Int. Conf.Acoustics,Speech,and Signal Processing, 2009, pp. 4625Ã¢ÂÂ4628.

V. Chan, S.-C. Liu, and A. van Schaik, Ã¢ÂÂAER EAR: A matched silicon cochlea pair with address event representation interface,Ã¢ÂÂIEEETrans.Circuits Syst. I, Reg. Papers, vol. 54, no. 1, pp. 48Ã¢ÂÂ59, 2007.

R. Sarpeshkar, C. Salthouse, J.-J. Sit, M. Baker, S. Zhak, T.-T. Lu, L.Turicchia, and S. Balster, Ã¢ÂÂAn ultra-low-power programmable analog bionicear processor,Ã¢ÂÂ IEEE Trans. Biomed. Eng., vol. 52, no.4, pp.711Ã¢ÂÂ727,2005

M. Baker and R. Sarpeshkar, Ã¢ÂÂLow-power single-loop and dual-loopAGCs for bionic ears,Ã¢ÂÂ IEEE J. Solid-State Circuits, vol. 41, no. 9, pp.1983Ã¢ÂÂ1996, 2006.

A. Hubbard, H. I. Cohen, S. Deligeorges, D. Freedman, T. Gore, C. Karl, S. Kelsall, M. Nourzad, Y. Pu, and S. Xue, Ã¢ÂÂMiniaturization ofelectronics for a biomimetic acoustic direction finding system for use on multiple platforms,Ã¢ÂÂ in Proc. SPIE Conf. Ser., 2008, vol. 6963.

F. L. Wightman and D. J. Kistler, Ã¢ÂÂThe dominant role of low-frequency interaural time differences in sound localization,Ã¢ÂÂ J. Acoust. Soc. Amer.,vol. 91, no. 3, pp. 1648Ã¢ÂÂ1661, 1992.

L. A. Jeffress, Ã¢ÂÂA place theory of sound localization,Ã¢ÂÂ J. Compar Physiol. Psychol., vol. 41, no. 1, pp. 35Ã¢ÂÂ39, 1948.

S. T. Birchfield and R. Gangishetty, Ã¢ÂÂAcoustic localization by interaural level difference,Ã¢ÂÂ in Proc. IEEE Int. Conf. Acoustics, Speech,Signal and Processing, 2005.

J. Weng and K. Y. Guentchev, Ã¢ÂÂThree-dimensional sound localization from a compact non-coplanar array of microphones using tree-basedlearning,Ã¢ÂÂ J. Acoust. Soc. Amer., vol. 110, no. 1, pp. 310Ã¢ÂÂ323, 2001.