Sukant Bisoy - Academia.edu (original) (raw)
Papers by Sukant Bisoy
Cognitive Big Data Intelligence with a Metaheuristic Approach
2023 1st International Conference on Circuits, Power and Intelligent Systems (CCPIS)
International Journal of Communication Networks and Distributed Systems, 2018
In this work, a stable active queue management (AQM) controller named SAQM is proposed to control... more In this work, a stable active queue management (AQM) controller named SAQM is proposed to control the congestion in the network and improve the stability of the queue length at internet router. It is an integrated rate and queue-based AQM technique which is stable and robust under dynamic environment where a number of TCP connections, bottleneck bandwidth, round trip time (RTT), and target queue length keeps changing. The proposed AQM controller is implemented in NS2 network simulator. In this paper, the control theory is used for stability analysis of TCP/SAQM system and finally validated through MATLAB Simulink model. The simulation results show that the SAQM outperforms other existing AQM controllers in terms of achieving queue stability, less oscillatory with faster response and at the same time it responds quickly to traffic change.
TCP is a connection oriented and reliable protocol of transport layer. TCP receiver generates one... more TCP is a connection oriented and reliable protocol of transport layer. TCP receiver generates one acknowledgment denoted as ACK for each data packet that received to acknowledges the sender. If the data packet and ACK uses the same path, then there may be collision between them due to channel contention. Finally TCP throughput is reduced due to huge ACK packet generation. TCP performance is very poor due to wireless channel characteristics. Since the packet loss is always related to path length, choosing appropriate delayed window size is an important parameter in wireless network. This paper, analyzes the impact of delayed acknowledgment on TCP variants say Newreno and Vegas in multi hop wireless networks. However the issues of congestion and loss recovery techniques for these two mechanisms are different. In this paper we study how well the TCP variants react with delayed acknowledgement strategy in multi hop network with different topologies and various flow patterns with respect to throughput, average delay and number of retransmission. The simulation result shows that TCP-Vegas achieve higher throughput, fewer retransmissions and less delay than Newreno. The performance of TCP-Vegas with delayed ack gets best performance among TCP variants.
Advances in intelligent systems and computing, 2017
In this work, we analyzed the performance of TCP variant protocols in wired-cum-wireless networks... more In this work, we analyzed the performance of TCP variant protocols in wired-cum-wireless networks considering active queue management (AQM) techniques such as random exponential marking (REM) and adaptive virtual queue (AVQ) along with Droptail. For analysis, we consider Reno, Newreno, Sack1, and Vegas as TCP variants and proposed a network model for wired-cum-wireless scenario. Then, the performance of TCP variants is analyzed using delayed acknowledgement (DelACK) technique. The simulation results show that Newreno performs better than others when DelACK is not used. However, when DelACK is used, the performance of Vegas is better than others irrespective of AQM techniques.
Computational Intelligence and Neuroscience, Jan 17, 2022
Understanding the situation is a critical component of any self-driving system. Accurate real-tim... more Understanding the situation is a critical component of any self-driving system. Accurate real-time visual signal processing to create pixelwise classed pictures, also known as semantic segmentation, is critical for scenario comprehension and subsequent acceptance of this new technology. Due to the intricate interaction between pixels in each frame of the received camera data, such efficiency in terms of processing time and accuracy could not be achieved prior to recent advances in deep learning algorithms. We present an effective approach for semantic segmentation for self-driving automobiles in this study. We combine deep learning architectures like convolutional neural networks and autoencoders, as well as cutting-edge approaches like feature pyramid networks and bottleneck residual blocks, to develop our model. e CamVid dataset, which has undergone considerable data augmentation, is utilised to train and test our model. To validate the suggested model, we compare the acquired findings to various baseline models reported in the literature.
2021 International Conference on Recent Trends on Electronics, Information, Communication & Technology (RTEICT), Aug 27, 2021
In this paper, a frontal facial expressions recognition classifier is developed based on a deep c... more In this paper, a frontal facial expressions recognition classifier is developed based on a deep convolutional model with a few numbers of learning parameters that make the proposed model applicable for mobile applications. In this, a high rate of accuracy and fast recognition is achieved. Firstly, contrast enhancement of the input images of the FER+ dataset is applied using the adaptive histogram equalization method. Then, the frontal facial images are only selected by using Haar-Cascade descriptors. Moreover, three different training methods are examined and the best recognition outcome of the presented CNN model is chosen, which outperforms the compared CNN models in terms of correctness and reduced number of parameters. The average accuracy of the proposed model reaches 94 % on validation and test sets. The acquired result can be considered promising in the domain of real-time communication with machines. Stating that feature selection and image enhancement have an effective impact on the precision of the emotion classifier. However, the shortage of datasets in any emotion class can defect the learning process and cause an instant decrease in the classifier accuracy.
Energy Efficiency and the Scheduling Policies are the key parameters of Cloud specialist organiza... more Energy Efficiency and the Scheduling Policies are the key parameters of Cloud specialist organizations. Consistently immense measures of electrical vitality devour by Cloud server farm which prompts more cost in expenses and discharge of CO2 to nature which is unfortunate for us. For this situation the need of Green Cloud tactics for limit emanation of carbon impressions just as operational expenses is the most extreme want. In our experimental work we have executed four distinctive power models, for example, Linear model, Cubic model, Square model and Square Root model with three different schedulers i.e. Time-Shared scheduling, Space-Shared scheduling and Dynamic-Workload scheduling on an IaaS Cloud condition to discover the best combination. Here we thought about the CPU uses and power utilization by empowering virtual machine migration. We discovered that the Cubic polynomial model under Time-Shared scheduling consumes less resource utilization (CPU, RAM and Bandwidth). Hence this combination is the most proficient one and expends less power.
Advances in intelligent systems and computing, 2014
Mobile ad hoc networks (MANETs) form a random network by consists of mobile nodes where node shar... more Mobile ad hoc networks (MANETs) form a random network by consists of mobile nodes where node share information with each other while moving. Due to presence of mobility in the MANET, the interconnections between nodes are likely to change, resulting in frequent changes of network topology. Therefore there is a need of identifying efficient dynamic routing protocol to provide call services in such network. In this paper, the effect of node velocity and unresponsive traffic volume is explored on performance of three routing protocols i.e. DSR, AODV and DYMO via NS2 simulator of ad hoc network of 100 mobile nodes. The performance is measured based on traffic admission ratio, packet delivery ratio (PDR), routing overhead, and average end-to-end delay. We observe that DYMO performs better in terms of PDR and traffic admission ratio and DSR has least overhead than others irrespective of node velocity, traffic volume and number of connection.
Lecture notes on data engineering and communications technologies, 2022
Wireless Communications and Mobile Computing, Jun 15, 2022
Credit scoring analysis has gained tremendous importance for researchers and the financial indust... more Credit scoring analysis has gained tremendous importance for researchers and the financial industries around the globe. It helps the financial industries to grant credits or loans to each deserving applicant with zero or minimal risks. However, developing an accurate and effective credit scoring model is a challenging task due to class imbalance and the presence of some irrelevant features. Recent researches show that ensemble learning has achieved supremacy in this field. In this paper, we performed an extensive comparative analysis of ensemble algorithms to bring further improvements in the algorithm oversampling, and feature selection (FS) techniques are implemented. The relevant features are identified by utilizing three FS techniques, such as information gain (IG), principal component analysis (PCA), and genetic algorithm (GA). Additionally, a comparative performance analysis is performed using 5 base and 14 ensemble models on three credit scoring datasets. The experimental results exhibit that the GA-based FS technique and CatBoost algorithm perform significantly better than other models in terms of five metrics, i.e., accuracy (ACC), area under the curve (AUC), F1-score, Brier score (BS), and Kolmogorov-Smirnov (KS).
Electronics, Aug 29, 2022
Advances in intelligent systems and computing, Aug 12, 2018
TCP is a transport layer protocol used for reliable transmission of data packets from one end to ... more TCP is a transport layer protocol used for reliable transmission of data packets from one end to other. One of the important functionality of TCP is to control congestion in the network. Congestion can be controlled through the window based mechanism of TCP. TCP comes with many variants Newreno, Vegas and FullTCP (two-way TCP). In this, TCP variants such as Newreno and Vegas including FullTCP protocol is analyzed using ad hoc on demand distance vector (AODV), dynamic source routing (DSR) and destination sequenced distance vector (DSDV) routing protocols using grid topology. The simulation result using NS2 shows that performance of FullTCP protocol is better than Newreno and Vegas irrespective routing protocol used.
Arabian journal for science and engineering, Aug 9, 2017
Journal of King Saud University - Computer and Information Sciences, Sep 1, 2022
Localization forms the heart of various autonomous mobile robots. For efficient navigation, these... more Localization forms the heart of various autonomous mobile robots. For efficient navigation, these robots need to adopt effective localization strategy. This paper, presents a comprehensive review on localization system, problems, principle and approaches for mobile robots. First, we classify the localization problems in to three categories based on the information of initial position of the robot. Next, we discuss on robot position update principles. Then, we discuss key techniques to localize the mobile robot such as: probabilistic approach, autonomous map building and radio frequency identification (RFID) based scheme. In the probabilistic localization section, we discuss the Markov localization and Kalman filter along with its extended versions. Autonomous map building focuses on the widely used simultaneous localization and mapping (SLAM) approach. This section also discusses on applying SLAM to localize braincontrolled mobile robots. Next, we discuss on applying evolutionary approaches to estimate optimal position. The RFID scheme addresses on effective utilization of RFID tags to track objects and position the robot. We then analyze on position and orientation errors occurred by different localization strategies. We conclude this paper by highlighting future research possibilities.
International Journal of Reasoning-based Intelligent Systems, 2020
Cognitive Big Data Intelligence with a Metaheuristic Approach
2023 1st International Conference on Circuits, Power and Intelligent Systems (CCPIS)
International Journal of Communication Networks and Distributed Systems, 2018
In this work, a stable active queue management (AQM) controller named SAQM is proposed to control... more In this work, a stable active queue management (AQM) controller named SAQM is proposed to control the congestion in the network and improve the stability of the queue length at internet router. It is an integrated rate and queue-based AQM technique which is stable and robust under dynamic environment where a number of TCP connections, bottleneck bandwidth, round trip time (RTT), and target queue length keeps changing. The proposed AQM controller is implemented in NS2 network simulator. In this paper, the control theory is used for stability analysis of TCP/SAQM system and finally validated through MATLAB Simulink model. The simulation results show that the SAQM outperforms other existing AQM controllers in terms of achieving queue stability, less oscillatory with faster response and at the same time it responds quickly to traffic change.
TCP is a connection oriented and reliable protocol of transport layer. TCP receiver generates one... more TCP is a connection oriented and reliable protocol of transport layer. TCP receiver generates one acknowledgment denoted as ACK for each data packet that received to acknowledges the sender. If the data packet and ACK uses the same path, then there may be collision between them due to channel contention. Finally TCP throughput is reduced due to huge ACK packet generation. TCP performance is very poor due to wireless channel characteristics. Since the packet loss is always related to path length, choosing appropriate delayed window size is an important parameter in wireless network. This paper, analyzes the impact of delayed acknowledgment on TCP variants say Newreno and Vegas in multi hop wireless networks. However the issues of congestion and loss recovery techniques for these two mechanisms are different. In this paper we study how well the TCP variants react with delayed acknowledgement strategy in multi hop network with different topologies and various flow patterns with respect to throughput, average delay and number of retransmission. The simulation result shows that TCP-Vegas achieve higher throughput, fewer retransmissions and less delay than Newreno. The performance of TCP-Vegas with delayed ack gets best performance among TCP variants.
Advances in intelligent systems and computing, 2017
In this work, we analyzed the performance of TCP variant protocols in wired-cum-wireless networks... more In this work, we analyzed the performance of TCP variant protocols in wired-cum-wireless networks considering active queue management (AQM) techniques such as random exponential marking (REM) and adaptive virtual queue (AVQ) along with Droptail. For analysis, we consider Reno, Newreno, Sack1, and Vegas as TCP variants and proposed a network model for wired-cum-wireless scenario. Then, the performance of TCP variants is analyzed using delayed acknowledgement (DelACK) technique. The simulation results show that Newreno performs better than others when DelACK is not used. However, when DelACK is used, the performance of Vegas is better than others irrespective of AQM techniques.
Computational Intelligence and Neuroscience, Jan 17, 2022
Understanding the situation is a critical component of any self-driving system. Accurate real-tim... more Understanding the situation is a critical component of any self-driving system. Accurate real-time visual signal processing to create pixelwise classed pictures, also known as semantic segmentation, is critical for scenario comprehension and subsequent acceptance of this new technology. Due to the intricate interaction between pixels in each frame of the received camera data, such efficiency in terms of processing time and accuracy could not be achieved prior to recent advances in deep learning algorithms. We present an effective approach for semantic segmentation for self-driving automobiles in this study. We combine deep learning architectures like convolutional neural networks and autoencoders, as well as cutting-edge approaches like feature pyramid networks and bottleneck residual blocks, to develop our model. e CamVid dataset, which has undergone considerable data augmentation, is utilised to train and test our model. To validate the suggested model, we compare the acquired findings to various baseline models reported in the literature.
2021 International Conference on Recent Trends on Electronics, Information, Communication & Technology (RTEICT), Aug 27, 2021
In this paper, a frontal facial expressions recognition classifier is developed based on a deep c... more In this paper, a frontal facial expressions recognition classifier is developed based on a deep convolutional model with a few numbers of learning parameters that make the proposed model applicable for mobile applications. In this, a high rate of accuracy and fast recognition is achieved. Firstly, contrast enhancement of the input images of the FER+ dataset is applied using the adaptive histogram equalization method. Then, the frontal facial images are only selected by using Haar-Cascade descriptors. Moreover, three different training methods are examined and the best recognition outcome of the presented CNN model is chosen, which outperforms the compared CNN models in terms of correctness and reduced number of parameters. The average accuracy of the proposed model reaches 94 % on validation and test sets. The acquired result can be considered promising in the domain of real-time communication with machines. Stating that feature selection and image enhancement have an effective impact on the precision of the emotion classifier. However, the shortage of datasets in any emotion class can defect the learning process and cause an instant decrease in the classifier accuracy.
Energy Efficiency and the Scheduling Policies are the key parameters of Cloud specialist organiza... more Energy Efficiency and the Scheduling Policies are the key parameters of Cloud specialist organizations. Consistently immense measures of electrical vitality devour by Cloud server farm which prompts more cost in expenses and discharge of CO2 to nature which is unfortunate for us. For this situation the need of Green Cloud tactics for limit emanation of carbon impressions just as operational expenses is the most extreme want. In our experimental work we have executed four distinctive power models, for example, Linear model, Cubic model, Square model and Square Root model with three different schedulers i.e. Time-Shared scheduling, Space-Shared scheduling and Dynamic-Workload scheduling on an IaaS Cloud condition to discover the best combination. Here we thought about the CPU uses and power utilization by empowering virtual machine migration. We discovered that the Cubic polynomial model under Time-Shared scheduling consumes less resource utilization (CPU, RAM and Bandwidth). Hence this combination is the most proficient one and expends less power.
Advances in intelligent systems and computing, 2014
Mobile ad hoc networks (MANETs) form a random network by consists of mobile nodes where node shar... more Mobile ad hoc networks (MANETs) form a random network by consists of mobile nodes where node share information with each other while moving. Due to presence of mobility in the MANET, the interconnections between nodes are likely to change, resulting in frequent changes of network topology. Therefore there is a need of identifying efficient dynamic routing protocol to provide call services in such network. In this paper, the effect of node velocity and unresponsive traffic volume is explored on performance of three routing protocols i.e. DSR, AODV and DYMO via NS2 simulator of ad hoc network of 100 mobile nodes. The performance is measured based on traffic admission ratio, packet delivery ratio (PDR), routing overhead, and average end-to-end delay. We observe that DYMO performs better in terms of PDR and traffic admission ratio and DSR has least overhead than others irrespective of node velocity, traffic volume and number of connection.
Lecture notes on data engineering and communications technologies, 2022
Wireless Communications and Mobile Computing, Jun 15, 2022
Credit scoring analysis has gained tremendous importance for researchers and the financial indust... more Credit scoring analysis has gained tremendous importance for researchers and the financial industries around the globe. It helps the financial industries to grant credits or loans to each deserving applicant with zero or minimal risks. However, developing an accurate and effective credit scoring model is a challenging task due to class imbalance and the presence of some irrelevant features. Recent researches show that ensemble learning has achieved supremacy in this field. In this paper, we performed an extensive comparative analysis of ensemble algorithms to bring further improvements in the algorithm oversampling, and feature selection (FS) techniques are implemented. The relevant features are identified by utilizing three FS techniques, such as information gain (IG), principal component analysis (PCA), and genetic algorithm (GA). Additionally, a comparative performance analysis is performed using 5 base and 14 ensemble models on three credit scoring datasets. The experimental results exhibit that the GA-based FS technique and CatBoost algorithm perform significantly better than other models in terms of five metrics, i.e., accuracy (ACC), area under the curve (AUC), F1-score, Brier score (BS), and Kolmogorov-Smirnov (KS).
Electronics, Aug 29, 2022
Advances in intelligent systems and computing, Aug 12, 2018
TCP is a transport layer protocol used for reliable transmission of data packets from one end to ... more TCP is a transport layer protocol used for reliable transmission of data packets from one end to other. One of the important functionality of TCP is to control congestion in the network. Congestion can be controlled through the window based mechanism of TCP. TCP comes with many variants Newreno, Vegas and FullTCP (two-way TCP). In this, TCP variants such as Newreno and Vegas including FullTCP protocol is analyzed using ad hoc on demand distance vector (AODV), dynamic source routing (DSR) and destination sequenced distance vector (DSDV) routing protocols using grid topology. The simulation result using NS2 shows that performance of FullTCP protocol is better than Newreno and Vegas irrespective routing protocol used.
Arabian journal for science and engineering, Aug 9, 2017
Journal of King Saud University - Computer and Information Sciences, Sep 1, 2022
Localization forms the heart of various autonomous mobile robots. For efficient navigation, these... more Localization forms the heart of various autonomous mobile robots. For efficient navigation, these robots need to adopt effective localization strategy. This paper, presents a comprehensive review on localization system, problems, principle and approaches for mobile robots. First, we classify the localization problems in to three categories based on the information of initial position of the robot. Next, we discuss on robot position update principles. Then, we discuss key techniques to localize the mobile robot such as: probabilistic approach, autonomous map building and radio frequency identification (RFID) based scheme. In the probabilistic localization section, we discuss the Markov localization and Kalman filter along with its extended versions. Autonomous map building focuses on the widely used simultaneous localization and mapping (SLAM) approach. This section also discusses on applying SLAM to localize braincontrolled mobile robots. Next, we discuss on applying evolutionary approaches to estimate optimal position. The RFID scheme addresses on effective utilization of RFID tags to track objects and position the robot. We then analyze on position and orientation errors occurred by different localization strategies. We conclude this paper by highlighting future research possibilities.
International Journal of Reasoning-based Intelligent Systems, 2020