Nikhil Sharma - Academia.edu (original) (raw)

Papers by Nikhil Sharma

Research paper thumbnail of Performance Analysis of Multi-User Uplink NOMA With Low Complex RS-OSIC Detector

2022 13th International Conference on Information and Communication Technology Convergence (ICTC)

Research paper thumbnail of Water discharge variations control fluvial stratigraphic architecture in the Middle Eocene Escanilla formation, Spain

Ancient fluvial deposits typically display repetitive changes in their depositional architecture ... more Ancient fluvial deposits typically display repetitive changes in their depositional architecture such as alternating intervals of coarse-grained highly amalgamated (HA), laterally-stacked, channel bodies, and finer-grained less amalgamated (LA), vertically-stacked, channels encased in floodplain deposits. Such patterns are usually ascribed to slower, respectively higher, rates of base level rise (accommodation). However, “upstream” factors such as water discharge and sediment flux also play a potential role in determining stratigraphic architecture, yet this possibility has never been tested despite the recent advances in the field of palaeohydraulic reconstructions from fluvial accumulations. Here, we chronicle riverbed gradient evolution within three Middle Eocene (~ 40 Ma) fluvial HA-LA sequences in the Escanilla Formation in the south-Pyrenean foreland basin. This work documents, for the first time in a fossil fluvial system, how the ancient riverbed systematically evolved from ...

Research paper thumbnail of A Survey on Energy harvested Cooperative Communication System

It is a review article on the topic of energy harvesting.

Research paper thumbnail of Black Hole Attack and Its Security Measure in Wireless Sensors Networks

Advances in Intelligent Systems and Computing, 2020

Computers are being reasonably important part of our daily lives. Different solutions were introd... more Computers are being reasonably important part of our daily lives. Different solutions were introduced for its connectivity such as wired approach which existed for longer duration of time. But due to advancement in field of technology wireless connectivity came into picture for connectivity to internet, exchanging information etc. One of suited network based on wireless standard is wireless sensor network (WSN). WSN comprises of a number of small motes that are distributed in random manner in any environment. The motes communicate with one another in their resource domain using low power communication medium. WSN possesses many characteristics but difficulty arises in its energy management, security aspect and deployment. Communication take place over radio environment and are prone to various attacks. Some of the attacks include Black hole attack, Gray hole attack, sinkhole attack etc. In this paper we are discussing network layer Black hole attack along with its security measure to decrease its effect in the network.

Research paper thumbnail of A Green 6G Network Era: Architecture and Propitious Technologies

Data Analytics and Management, 2021

Research paper thumbnail of Transaction Privacy Preservations for Blockchain Technology

Advances in Intelligent Systems and Computing, 2020

Blockchain is defined as a technology comprising of logic unit of work done in form of transactio... more Blockchain is defined as a technology comprising of logic unit of work done in form of transactions where any read or write operation can be carried out and stored in form of blocks where any financial accounts can be accessed by different servers. This is widely used technology due to its tremendous applicability such as independent verification without the involvement of any third party, no alteration or deletion of data, use of cryptography for secure financial accounts, sequential recording of transactions, pure transparent process of transaction. Blockchain follows both private and public type of architecture. In public model, anyone can read and write while in private only authorized users have the access. Security plays a vital role where secure transactions are to be carried out. Main principles of security such as confidentiality which implies the secrecy of send message, integrity which implies the consistency of send and received message and availability which implies information required must be available for reference, which must be preserved in any security-based model. In order to reduce the impact of attacks, various security algorithms such as cryptography, ring signature, Non-interactive Zero- Knowledge Proof, CryptoNote and Monero, Homomorphic Cryptosystem system are introduced in the paper.

Research paper thumbnail of Abstract T P344: Clopidogrel Resistance in Stroke Patients (The CRISP Trial)

Stroke, 2014

Background: Clopidogrel is a thienopyridine with ADP antagonising effects that is widely used in ... more Background: Clopidogrel is a thienopyridine with ADP antagonising effects that is widely used in the secondary prevention of ischemic strokes and transient ischemic attacks (TIAs). Despite this some patients will go on to have recurrent strokes or TIAs. Cardiology trials have identified patients with clopidogrel resistance, defined as a vasodilator-stimulated phosphoprotein (VASP) index of >50% after at least 7 days of clopidogrel treatment. This was associated with recurrent cardiac events. The VASP-index measures phosphorylation of VASP in the resting state and activated states using flow cytometry techniques. As phosphorylation of VASP is dependent upon activation of the platelet P2Y12 receptor that is targeted by clopidogrel the difference in phosphorylation between the two states provides an indirect measure of platelet inhibition. Previous trials have shown a prevalence of clopidogrel resistance in stroke patients of between 8 and 18%, but no study to date has shown whether...

Research paper thumbnail of An Intelligent Differential Protection Scheme for DC Microgrid

2021 9th IEEE International Conference on Power Systems (ICPS), 2021

This paper proposes a differential current based intelligent fault detection technique for DC mic... more This paper proposes a differential current based intelligent fault detection technique for DC microgrids. The currents of both ends of the faulted line are utilized for the computation of differential current. This differential current is then processed through the machine learning (ML) technique named as support vector machine (SVM) for fault detection in low voltage DC (LVDC) microgrids. The SVM based protection technique is tested on a LVDC microgrid integrated with different types of renewable energy sources (RES) modelled in the SIMULINK platform. The presented protection scheme is examined for pole-ground (PG) and pole-pole (PP) faults with varying fault resistance, fault location and RES penetration in grid-connected and islanding mode. The proposed methodology is verified for its capability to discriminate the transient no-fault condition from fault cases.

Research paper thumbnail of Upstream versus downstream changes in a natural sediment routing system from source to sink

Research paper thumbnail of Plankton activity may explain aerobic methane production in meromictic Lake Cadagno

Research paper thumbnail of In silico analysis of some microbial amidases for their amino acid and physiochemical parameters

International Journal of Bioassays, 2013

Amino acid sequences of amidases were retrieved from respective databases and in silico analysis ... more Amino acid sequences of amidases were retrieved from respective databases and in silico analysis for different physiochemical properties and substrate specificity has been done. Multiple Sequence Alignment (MSA) and statistical analysis of amino acid sequences has revealed significant differences among aliphatic and signature amidases in terms of conserved motif that plays a vital role in substrate binding and catalytic function. MSA has revealed that in aliphatic amidases the conserved amino acid residues involved in catalytic function are position specific which remains within catalytic traid of Cys-166, Glu-59, Lys-134, while signature amidase contains signature motif GGSS (S/G) GS and catalytic traid of Ser-171, Ser-195, Lys-96 which are not position specific. Statistical analysis has revealed that these two groups also differ in physiochemical properties. In contrast to aliphatic amidases, signature amidase have significantly higher number of amino acid residues & molecular mas...

Research paper thumbnail of Miscellaneous Affections of Knee

Research paper thumbnail of A cyclic backbone curve for squat reinforced concrete shear walls

by Siamak Epackachi, Nikhil Sharma, Andrew Whitaker, Ronald O. Hamburger and Ayse Hortacsu

Research paper thumbnail of Analysis of Contributors to Diversity in South India

The poster presents the analysis of the trends in various fields in south India mainly concerning... more The poster presents the analysis of the trends in various fields in south India mainly concerning the states Kerala, Tamil Nadu, Andhra Pradesh, Karnataka and Maharashtra. District level analyses of these states have been undertaken for concentrating on spatial distribution of data. 2001 census data is the source for this analysis. The variations in Per Capita Income and National District Domestic Product (NDDP) of Agriculture are of major concern. Besides this literacy, sex ratio, population of the minorities, NDDP for infrastructure development & trade has also been considered & their overall effect on the growth of the state has been studied. Visualization of this data in ArcGIS using regression, cluster and hot spot analysis was performed which revealed the spatial patterns in the data. GIS was highly helpful in finding the anomalies within data which unless otherwise would have been left unnoticed. The metro cities like Chennai, Bangalore & Hyderabad which were supposed to show...

Research paper thumbnail of Predictive Analysis of Type 2 Diabetes Using Hybrid ML Model and IoT

IoT Security Paradigms and Applications, 2020

Research paper thumbnail of Analysis of Various Machine Learning Algorithm for Cardiac Pulse Prediction

2019 International Conference on Computing, Communication, and Intelligent Systems (ICCCIS), 2019

By applying machine learning algorithms, patterns are identified or recognized in the process of ... more By applying machine learning algorithms, patterns are identified or recognized in the process of Pattern Recognition. On the grounds of prior knowledge, the data is collected and sorted. In this method, the raw data is transformed into a susceptible form which can be used by the machine. Electrocardiogram (ECG) Pattern Recognition is the main focus of this paper. ECG keeps a track of heart’s electrical activity. In the field of biometric it is used as a robust biometric. On the person, off the person and in the person, are the three categories for tracking and capturing signals. Only Off-the-person category in which there is no or minimal skin contact, is included in this paper. To analyze and implement data, six baseline methods are utilized. These baseline methods are applied two publicly available databases-CYBHi and UofT. Raw signals and spectrogram of heartbeat are used for studying about representing features. Various machine learning algorithms are also discussed. Implementation for predicting heartbeat as normal or abnormal and heart diseases, is performed.

Research paper thumbnail of Time-Resolved Endoscopic Evaluation of Spatial Temperature and Soot Distribution in a Butanol-Diesel Blend Fueled Direct Injection Compression Ignition Engine

Journal of Energy Resources Technology, 2021

In situ spatial soot and temperature distributions were investigated experimentally for B20 (20% ... more In situ spatial soot and temperature distributions were investigated experimentally for B20 (20% v/v butanol and balance mineral diesel blend) vis-a-vis mineral diesel using endoscopic visualization technique. Endoscope captured the in-cylinder combustion images in a production-grade direct injection compression ignition (DICI) engine at varying engine operating conditions. A comparative combustion data analysis using pressure-crank angle history and the captured endoscopic images was performed. An attempt was made to correlate the results of these two experimental investigations. Combustion duration (CD) obtained from the endoscopic images was relatively longer than the CD calculated from the thermodynamic analysis. Most research on soot and NOx emitted from the engine using a raw exhaust gas emission analyzer provides bulk, time-averaged, and cycle-averaged information about the pollutant formation. This investigation is unique wherein the spatial or time-resolved soot and NOx for...

Research paper thumbnail of Quality Initiative for the Administration of Vancomycin Prophylaxis in Penicillin-Allergic Neurosurgery Patients

Research paper thumbnail of Spectrally efficient M‐ary QAM based multi‐antenna cooperative system over TWDP fading channel

Transactions on Emerging Telecommunications Technologies, 2021

To cater the ever‐increasing demand of wireless high‐speed data access and mobile users, one of t... more To cater the ever‐increasing demand of wireless high‐speed data access and mobile users, one of the potential enablers toward the sustainable improved network capacity is the network densification at different layers of the radio access system. In this article, a possible solution to achieve a spectrally efficient, ultradense network (UDN) for cellular communication is analyzed. In particular, the performance of the generalized M‐ary quadrature amplitude modulation (M‐QAM) scheme for multiantenna cooperative decode‐and‐forward (DF) system over two‐wave with diffuse power (TWDP) fading is presented. The TWDP fading is evaluated using the moment generating function (MGF) based approach on spectral efficient QAM transmission. In this article, novel closed‐form expressions for the exact average symbol error rate (ASER) for the generalized rectangular‐QAM (R‐QAM) and square‐QAM (S‐QAM) schemes are presented. The obtained expressions are in the form of Appell's and Laureicella's confluent hypergeometric function which can be evaluated mathematically either in the series or the integral‐form. Further, the asymptotic analysis is also presented for the considered system to get better insights into the system. The analytical results derived are corroborated by simulations to validate the accuracy of the analysis.

Research paper thumbnail of Depression Anatomy Using Combinational Deep Neural Network

Advances in Intelligent Systems and Computing, 2020

Depression is a temperament syndrome that causes a tenacious emotion of wretchedness and loss of ... more Depression is a temperament syndrome that causes a tenacious emotion of wretchedness and loss of interest in any activity. It is a supreme root of mental illness, which has established the growth in the risk of early death and economic burden to a country. Traditional clinical analysis procedures are subjective, complex and need considerable contribution of professionals. The turn of the century saw incredible progress in using deep learning for medical diagnosis. Though, prediction and implementation of mental state can be remarkably hard. In this paper, we present Combinational Deep Neural Network (CDNN) for automated depression detection with facial images and text data using amalgamation of Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN). Thereafter, combining the prediction scores of both CNN and RNN model and level of depression is decided on the basis of the range of the predefined depression-level scores. Simulation outcomes based on real-field channel measurements show that the proposed model can significantly predict depression with superior performance.

Research paper thumbnail of Performance Analysis of Multi-User Uplink NOMA With Low Complex RS-OSIC Detector

2022 13th International Conference on Information and Communication Technology Convergence (ICTC)

Research paper thumbnail of Water discharge variations control fluvial stratigraphic architecture in the Middle Eocene Escanilla formation, Spain

Ancient fluvial deposits typically display repetitive changes in their depositional architecture ... more Ancient fluvial deposits typically display repetitive changes in their depositional architecture such as alternating intervals of coarse-grained highly amalgamated (HA), laterally-stacked, channel bodies, and finer-grained less amalgamated (LA), vertically-stacked, channels encased in floodplain deposits. Such patterns are usually ascribed to slower, respectively higher, rates of base level rise (accommodation). However, “upstream” factors such as water discharge and sediment flux also play a potential role in determining stratigraphic architecture, yet this possibility has never been tested despite the recent advances in the field of palaeohydraulic reconstructions from fluvial accumulations. Here, we chronicle riverbed gradient evolution within three Middle Eocene (~ 40 Ma) fluvial HA-LA sequences in the Escanilla Formation in the south-Pyrenean foreland basin. This work documents, for the first time in a fossil fluvial system, how the ancient riverbed systematically evolved from ...

Research paper thumbnail of A Survey on Energy harvested Cooperative Communication System

It is a review article on the topic of energy harvesting.

Research paper thumbnail of Black Hole Attack and Its Security Measure in Wireless Sensors Networks

Advances in Intelligent Systems and Computing, 2020

Computers are being reasonably important part of our daily lives. Different solutions were introd... more Computers are being reasonably important part of our daily lives. Different solutions were introduced for its connectivity such as wired approach which existed for longer duration of time. But due to advancement in field of technology wireless connectivity came into picture for connectivity to internet, exchanging information etc. One of suited network based on wireless standard is wireless sensor network (WSN). WSN comprises of a number of small motes that are distributed in random manner in any environment. The motes communicate with one another in their resource domain using low power communication medium. WSN possesses many characteristics but difficulty arises in its energy management, security aspect and deployment. Communication take place over radio environment and are prone to various attacks. Some of the attacks include Black hole attack, Gray hole attack, sinkhole attack etc. In this paper we are discussing network layer Black hole attack along with its security measure to decrease its effect in the network.

Research paper thumbnail of A Green 6G Network Era: Architecture and Propitious Technologies

Data Analytics and Management, 2021

Research paper thumbnail of Transaction Privacy Preservations for Blockchain Technology

Advances in Intelligent Systems and Computing, 2020

Blockchain is defined as a technology comprising of logic unit of work done in form of transactio... more Blockchain is defined as a technology comprising of logic unit of work done in form of transactions where any read or write operation can be carried out and stored in form of blocks where any financial accounts can be accessed by different servers. This is widely used technology due to its tremendous applicability such as independent verification without the involvement of any third party, no alteration or deletion of data, use of cryptography for secure financial accounts, sequential recording of transactions, pure transparent process of transaction. Blockchain follows both private and public type of architecture. In public model, anyone can read and write while in private only authorized users have the access. Security plays a vital role where secure transactions are to be carried out. Main principles of security such as confidentiality which implies the secrecy of send message, integrity which implies the consistency of send and received message and availability which implies information required must be available for reference, which must be preserved in any security-based model. In order to reduce the impact of attacks, various security algorithms such as cryptography, ring signature, Non-interactive Zero- Knowledge Proof, CryptoNote and Monero, Homomorphic Cryptosystem system are introduced in the paper.

Research paper thumbnail of Abstract T P344: Clopidogrel Resistance in Stroke Patients (The CRISP Trial)

Stroke, 2014

Background: Clopidogrel is a thienopyridine with ADP antagonising effects that is widely used in ... more Background: Clopidogrel is a thienopyridine with ADP antagonising effects that is widely used in the secondary prevention of ischemic strokes and transient ischemic attacks (TIAs). Despite this some patients will go on to have recurrent strokes or TIAs. Cardiology trials have identified patients with clopidogrel resistance, defined as a vasodilator-stimulated phosphoprotein (VASP) index of >50% after at least 7 days of clopidogrel treatment. This was associated with recurrent cardiac events. The VASP-index measures phosphorylation of VASP in the resting state and activated states using flow cytometry techniques. As phosphorylation of VASP is dependent upon activation of the platelet P2Y12 receptor that is targeted by clopidogrel the difference in phosphorylation between the two states provides an indirect measure of platelet inhibition. Previous trials have shown a prevalence of clopidogrel resistance in stroke patients of between 8 and 18%, but no study to date has shown whether...

Research paper thumbnail of An Intelligent Differential Protection Scheme for DC Microgrid

2021 9th IEEE International Conference on Power Systems (ICPS), 2021

This paper proposes a differential current based intelligent fault detection technique for DC mic... more This paper proposes a differential current based intelligent fault detection technique for DC microgrids. The currents of both ends of the faulted line are utilized for the computation of differential current. This differential current is then processed through the machine learning (ML) technique named as support vector machine (SVM) for fault detection in low voltage DC (LVDC) microgrids. The SVM based protection technique is tested on a LVDC microgrid integrated with different types of renewable energy sources (RES) modelled in the SIMULINK platform. The presented protection scheme is examined for pole-ground (PG) and pole-pole (PP) faults with varying fault resistance, fault location and RES penetration in grid-connected and islanding mode. The proposed methodology is verified for its capability to discriminate the transient no-fault condition from fault cases.

Research paper thumbnail of Upstream versus downstream changes in a natural sediment routing system from source to sink

Research paper thumbnail of Plankton activity may explain aerobic methane production in meromictic Lake Cadagno

Research paper thumbnail of In silico analysis of some microbial amidases for their amino acid and physiochemical parameters

International Journal of Bioassays, 2013

Amino acid sequences of amidases were retrieved from respective databases and in silico analysis ... more Amino acid sequences of amidases were retrieved from respective databases and in silico analysis for different physiochemical properties and substrate specificity has been done. Multiple Sequence Alignment (MSA) and statistical analysis of amino acid sequences has revealed significant differences among aliphatic and signature amidases in terms of conserved motif that plays a vital role in substrate binding and catalytic function. MSA has revealed that in aliphatic amidases the conserved amino acid residues involved in catalytic function are position specific which remains within catalytic traid of Cys-166, Glu-59, Lys-134, while signature amidase contains signature motif GGSS (S/G) GS and catalytic traid of Ser-171, Ser-195, Lys-96 which are not position specific. Statistical analysis has revealed that these two groups also differ in physiochemical properties. In contrast to aliphatic amidases, signature amidase have significantly higher number of amino acid residues & molecular mas...

Research paper thumbnail of Miscellaneous Affections of Knee

Research paper thumbnail of A cyclic backbone curve for squat reinforced concrete shear walls

by Siamak Epackachi, Nikhil Sharma, Andrew Whitaker, Ronald O. Hamburger and Ayse Hortacsu

Research paper thumbnail of Analysis of Contributors to Diversity in South India

The poster presents the analysis of the trends in various fields in south India mainly concerning... more The poster presents the analysis of the trends in various fields in south India mainly concerning the states Kerala, Tamil Nadu, Andhra Pradesh, Karnataka and Maharashtra. District level analyses of these states have been undertaken for concentrating on spatial distribution of data. 2001 census data is the source for this analysis. The variations in Per Capita Income and National District Domestic Product (NDDP) of Agriculture are of major concern. Besides this literacy, sex ratio, population of the minorities, NDDP for infrastructure development & trade has also been considered & their overall effect on the growth of the state has been studied. Visualization of this data in ArcGIS using regression, cluster and hot spot analysis was performed which revealed the spatial patterns in the data. GIS was highly helpful in finding the anomalies within data which unless otherwise would have been left unnoticed. The metro cities like Chennai, Bangalore & Hyderabad which were supposed to show...

Research paper thumbnail of Predictive Analysis of Type 2 Diabetes Using Hybrid ML Model and IoT

IoT Security Paradigms and Applications, 2020

Research paper thumbnail of Analysis of Various Machine Learning Algorithm for Cardiac Pulse Prediction

2019 International Conference on Computing, Communication, and Intelligent Systems (ICCCIS), 2019

By applying machine learning algorithms, patterns are identified or recognized in the process of ... more By applying machine learning algorithms, patterns are identified or recognized in the process of Pattern Recognition. On the grounds of prior knowledge, the data is collected and sorted. In this method, the raw data is transformed into a susceptible form which can be used by the machine. Electrocardiogram (ECG) Pattern Recognition is the main focus of this paper. ECG keeps a track of heart’s electrical activity. In the field of biometric it is used as a robust biometric. On the person, off the person and in the person, are the three categories for tracking and capturing signals. Only Off-the-person category in which there is no or minimal skin contact, is included in this paper. To analyze and implement data, six baseline methods are utilized. These baseline methods are applied two publicly available databases-CYBHi and UofT. Raw signals and spectrogram of heartbeat are used for studying about representing features. Various machine learning algorithms are also discussed. Implementation for predicting heartbeat as normal or abnormal and heart diseases, is performed.

Research paper thumbnail of Time-Resolved Endoscopic Evaluation of Spatial Temperature and Soot Distribution in a Butanol-Diesel Blend Fueled Direct Injection Compression Ignition Engine

Journal of Energy Resources Technology, 2021

In situ spatial soot and temperature distributions were investigated experimentally for B20 (20% ... more In situ spatial soot and temperature distributions were investigated experimentally for B20 (20% v/v butanol and balance mineral diesel blend) vis-a-vis mineral diesel using endoscopic visualization technique. Endoscope captured the in-cylinder combustion images in a production-grade direct injection compression ignition (DICI) engine at varying engine operating conditions. A comparative combustion data analysis using pressure-crank angle history and the captured endoscopic images was performed. An attempt was made to correlate the results of these two experimental investigations. Combustion duration (CD) obtained from the endoscopic images was relatively longer than the CD calculated from the thermodynamic analysis. Most research on soot and NOx emitted from the engine using a raw exhaust gas emission analyzer provides bulk, time-averaged, and cycle-averaged information about the pollutant formation. This investigation is unique wherein the spatial or time-resolved soot and NOx for...

Research paper thumbnail of Quality Initiative for the Administration of Vancomycin Prophylaxis in Penicillin-Allergic Neurosurgery Patients

Research paper thumbnail of Spectrally efficient M‐ary QAM based multi‐antenna cooperative system over TWDP fading channel

Transactions on Emerging Telecommunications Technologies, 2021

To cater the ever‐increasing demand of wireless high‐speed data access and mobile users, one of t... more To cater the ever‐increasing demand of wireless high‐speed data access and mobile users, one of the potential enablers toward the sustainable improved network capacity is the network densification at different layers of the radio access system. In this article, a possible solution to achieve a spectrally efficient, ultradense network (UDN) for cellular communication is analyzed. In particular, the performance of the generalized M‐ary quadrature amplitude modulation (M‐QAM) scheme for multiantenna cooperative decode‐and‐forward (DF) system over two‐wave with diffuse power (TWDP) fading is presented. The TWDP fading is evaluated using the moment generating function (MGF) based approach on spectral efficient QAM transmission. In this article, novel closed‐form expressions for the exact average symbol error rate (ASER) for the generalized rectangular‐QAM (R‐QAM) and square‐QAM (S‐QAM) schemes are presented. The obtained expressions are in the form of Appell's and Laureicella's confluent hypergeometric function which can be evaluated mathematically either in the series or the integral‐form. Further, the asymptotic analysis is also presented for the considered system to get better insights into the system. The analytical results derived are corroborated by simulations to validate the accuracy of the analysis.

Research paper thumbnail of Depression Anatomy Using Combinational Deep Neural Network

Advances in Intelligent Systems and Computing, 2020

Depression is a temperament syndrome that causes a tenacious emotion of wretchedness and loss of ... more Depression is a temperament syndrome that causes a tenacious emotion of wretchedness and loss of interest in any activity. It is a supreme root of mental illness, which has established the growth in the risk of early death and economic burden to a country. Traditional clinical analysis procedures are subjective, complex and need considerable contribution of professionals. The turn of the century saw incredible progress in using deep learning for medical diagnosis. Though, prediction and implementation of mental state can be remarkably hard. In this paper, we present Combinational Deep Neural Network (CDNN) for automated depression detection with facial images and text data using amalgamation of Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN). Thereafter, combining the prediction scores of both CNN and RNN model and level of depression is decided on the basis of the range of the predefined depression-level scores. Simulation outcomes based on real-field channel measurements show that the proposed model can significantly predict depression with superior performance.