Papers by Sudipta Mahapatra
Traffic jams and accidents with increasing number of vehicles on the road cause significant socia... more Traffic jams and accidents with increasing number of vehicles on the road cause significant social and financial losses. Minor accidents on the road are handled by locals and other drivers. However, attention must be paid to the serious accidents that happen more frequently. Hence, an intelligent accident detection system is necessary to detect accident and save the life of injured. Now-a-days, vehicular ad hoc network (VANET) has been used significantly to solve the issue. In this work, we have developed a prototype hardware VANET model for detecting accidents using fuzzy logic controller (FLC) or fuzzy logic decision support. The proposed model detects accidents and sends the messages to a central server through a low range radio (LoRa) transceiver network. The server after receiving the message from a vehicle estimates the shortest path using Dijkstra’s algorithm and send messages to a nearby ambulance to take the victim to a nearby hospital. Ambulance is also a node in the propo...
With the increasing usage of mobile devices to ubiquitously access heterogeneous applications in ... more With the increasing usage of mobile devices to ubiquitously access heterogeneous applications in wireless Internet, the measurement and analysis of Internet traffic has become a key research area. In this paper, we present the results of our measurements for VBR traffic over UDP in 802.11g and GPRS networks. We focus on Inter-Packet Arrival Time (IPRT) and Inter-Packet Transmission Delay (IPTD) and observe that the later has a significant impact on the round trip delay. Numerical parameters for Weibull, Exponential and Normal distribution in order to represent such traffic are also presented.
AI & SOCIETY, 2013
Big data analytics has been the focus for large scale data processing. Machine learning and Big d... more Big data analytics has been the focus for large scale data processing. Machine learning and Big data has great future in prediction. Churn prediction is one of the sub domain of big data. Preventing customer attrition especially in telecom is the advantage of churn prediction. Churn prediction is a day-today affair involving millions. So a solution to prevent customer attrition can save a lot. This paper propose to do comparison of three machine learning techniques Decision tree algorithm, Random Forest algorithm and Gradient Boosted tree algorithm using Apache Spark. Apache Spark is a data processing engine used in big data which provides in-memory processing so that the processing speed is higher. The analysis is made by extracting the features of the data set and training the model. Scala is a programming language that combines both object oriented and functional programming and so a powerful programming language. The analysis is implemented using Apache Spark and modelling is done using scala ML. The accuracy of Decision tree model came out as 86%, Random Forest model is 87% and Gradient Boosted tree is 85%.
Intelligent Traffic Quantification System
IOP Conference Series: Materials Science and Engineering, 2017
Currently, city traffic monitoring and controlling is a big issue in almost all cities worldwide.... more Currently, city traffic monitoring and controlling is a big issue in almost all cities worldwide. Vehicular ad-hoc Network (VANET) technique is an efficient tool to minimize this problem. Usually, different types of on board sensors are installed in vehicles to generate messages characterized by different vehicle parameters. In this work, an intelligent system based on fuzzy clustering technique is developed to reduce the number of individual messages by extracting important features from the messages of a vehicle. Therefore, the proposed fuzzy clustering technique reduces the traffic load of the network. The technique also reduces congestion and quantifies congestion.
A Novel Two Pass Rate Control Scheme for Variable Bit Rate Video Streaming
2015 IEEE International Symposium on Multimedia (ISM), 2015
In this paper, a novel two-pass rate control scheme is proposed to achieve a consistent visual qu... more In this paper, a novel two-pass rate control scheme is proposed to achieve a consistent visual quality media for variable bit rate (VBR) video streaming. The rate-distortion (RD) characteristics of each frame is used to establish a frame complexity model, which is later used along with statistics collected in the first-pass to derive an optimal quantization parameter for encoding the frame in the second-pass. The experimental results demonstrate that the proposed rate control scheme significantly outperforms the existing rate control mechanism in the Joint Model (JM) reference software in terms of the Peak Signal to Noise Ratio (PSNR) and consistent perceptual visual quality while achieving the target bit rate. Further, the proposed scheme is validated through implementation on a miniature test-bed.
An analytical framework for evaluating MIPv6 protocols applying transport engineering concepts
Proceedings of the 7th ACM workshop on Performance monitoring and measurement of heterogeneous wireless and wired networks - PM2HW2N '12, 2012
ABSTRACT As the future generation networks are envisioned to be heterogeneous in nature, seamless... more ABSTRACT As the future generation networks are envisioned to be heterogeneous in nature, seamless mobility in such networks is an important issue. While IETF work groups have standardised various mobility management protocols, such as Mobile IPv6 (MIPv6), Fast Handovers for Mobile IPv6 (Predictive FMIPv6, and Reactive FMIPv6), Hierarchical Mobile IPv6 (HMIPv6), Proxy Mobile IPv6 (PMIPv6) and Fast Handovers for PMIPv6 (Predictive FPMIPv6, and Reactive FPMIPv6), the decision regarding which protocol suits the future networks is still a research issue. A good analytical or evaluation model, in terms of various performance metrics like handover latency, packet density, and signalling cost during the handover, is needed to compare various mobility management protocols. In this paper, a novel analytical model in terms of the handover latency, as well as the packet density, and packet arrival rate during the handover time is developed for comparison of various mobility management protocols by applying transport engineering principles in the field of telecommunication. The model enables us to make a few important observations regarding the performance of these mobility management protocols.
Predictability of sugar futures: evidence from the Indian commodity market
Agricultural Finance Review, 2015
Purpose – The forecasting power of commodity futures is a matter of intensive research as evidenc... more Purpose – The forecasting power of commodity futures is a matter of intensive research as evidenced by a number of related publications. The purpose of this paper is to illustrate how advanced forecasting techniques improve the predictability of sugar futures in the Indian commodity market. Design/methodology/approach – The forward premium is estimated using ordinary least square regression technique. Different linear and nonlinear models are used to forecast the sugar future spot prices from the futures prices. The forecasting accuracy of each pair of models is then compared by estimating the corresponding Diebold-Mariano test statistics. Findings – From the estimated forward premiums, it is found that there is more volatility toward the date of maturity for a three-month horizon compared to six-month, and 12-month horizons. It is established that the futures prices of sugar, when used in a model, are able to generate better forecasts for the future spot prices. Moreover, the forec...
As the future generation networks are envisioned to be heterogeneous in nature, seamless mobility... more As the future generation networks are envisioned to be heterogeneous in nature, seamless mobility in such networks is an important issue. While IETF work groups have standardized various mobility management protocols, such as Mobile IPv6 (MIPv6), Fast Handovers for Mobile IPv6 (Predictive FMIPv6, and Reactive FMIPv6), Hierarchical Mobile IPv6 (HMIPv6), Proxy Mobile IPv6 (PMIPv6) and Fast Handovers for PMIPv6 (Predictive FPMIPv6, and Reactive FPMIPv6), out of which some are host based and some are network based, the decision regarding which protocol suits the future networks is still a research issue. The study of various mobility management protocols in terms handover latency and the number of hops is needed to evaluate these protocols. Even though much study has been done in literature in terms of handover latency, study still needs performance evaluation in terms of average hop delay. In this paper we study various mobility management protocols by applying simple numerical analysi...
Next Generation Wireless Internet will comprise heterogeneous networks connected to seamlessly ac... more Next Generation Wireless Internet will comprise heterogeneous networks connected to seamlessly access multitude of applications and services; managing traffic will become very crucial to successfully run those networks. In this paper, we report a framework to measure end-to-end traffic between Server and Client located in India and USA. The measured traffic is used to design Hidden Markov Model based forecasting algorithm and is validated for different packet size. Keywords-Hidden Markov Model; traffic measurement; traffic forecasting;LTE;Next Generation Wireless Internet
Minimal Path-Based Reliability Model for Wireless Sensor Networks With Multistate Nodes
IEEE Transactions on Reliability
Wireless sensor networks (WSNs) find application in various fields like environmental monitoring,... more Wireless sensor networks (WSNs) find application in various fields like environmental monitoring, health-care, land security, and many more. To ease our day-to-day activity, WSNs have become an integral tool for complex data gathering tasks. Monitoring a phenomenon by a WSN depends on the collective data provided by the sensor nodes. To ensure reliable operation of WSNs, it is important to quantify the performance of such networks in terms of network reliability measures. This article studies the reliability of WSNs with multistate nodes and proposes an approach to evaluate the flow-oriented network reliability of WSNs consisting of multistate sensor nodes. The proposed method takes into account the dynamic state of the network due to multistate sensor nodes. The proposed approach includes enumeration of shortest minimal paths from application-specific flow satisfying sensor nodes (source nodes) to the sink node. It then proposes a modified sum-of-disjoint products approach to evaluate WSN reliability in the presence of multistate nodes from the enumerated shortest minimal paths. Simulations are performed on WSNs of various sizes to show the applicability of the proposed approach on arbitrary WSNs.
Brillouin Distributed Temperature Sensor Using Optical Time Domain Reflectometry Techniques
Smart Sensors, Measurement and Instrumentation, 2016
This paper presents the performance improvement of Brillouin distributed temperature sensor (BDTS... more This paper presents the performance improvement of Brillouin distributed temperature sensor (BDTS) using deconvolution algorithm. We have analysed three different OTDR techniques in terms of spatial resolution improvement using Fourier regularized deconvolution (FourRD) algorithm. The effects of coherent Rayleigh noise (CRN) on temperature and spatial resolution of the above system are being investigated. In this paper, using a light source of power 10 mW and 400 ns pulse widths; a spatial resolution of 25 and 20 m is observed for a pseudorandom coded BOTDR in the presence and absence of CRN respectively. Similarly, in case of conventional BOTDR and coherent BOTDR system in the presence of CRN, the spatial resolutions observed are 40 and 35 m respectively. Numerical simulation results indicate that the pseudorandom coded BOTDR is a better candidate for design of BDTS in terms of spatial resolution as compared to the other two schemes as discussed above.
Adaptive enhancement of compressed SAR images
Signal, Image and Video Processing, 2016
Synthetic aperture radar (SAR) imagery is an important global all-weather surveillance and mappin... more Synthetic aperture radar (SAR) imagery is an important global all-weather surveillance and mapping satellite imagery system. As space-borne systems have a limited storage capacity, it is imperative to heavily compress SAR images, possible with lossy compression schemes. As a result, SAR images need to be enhanced in earth stations. The work reported in this paper aims to address the issue of compression artefact removal of SAR images in an adaptive manner. The SAR images, compressed using the JPEG utility at significantly low bit rates, are enhanced by adaptively removing coding artefacts and speckle noise. As edges carry significant information in satellite imagery, a significant edge image is used for edge enhancement with selective removal of noisy edges. Further, an image sharpness metric is proposed in this work to serve as an objective no-reference metric for measuring the sharpness of SAR images.
In this paper, we present a novel extension technique to the Set Partitioning in Hierarchical Tre... more In this paper, we present a novel extension technique to the Set Partitioning in Hierarchical Trees (SPIHT) based image compression with spatial scalability. The present modification and the preprocessing techniques provide significantly better quality (both subjectively and objectively) reconstruction at the decoder with little additional computational complexity. There are two proposals for this paper. Firstly, we propose a pre-processing scheme, called Zero-Shifting, that brings the spatial values in signed integer range without changing the dynamic ranges, so that the transformed coefficient calculation becomes more consistent. For that reason, we have to modify the initialization step of the SPIHT algorithms. The experiments demonstrate a significant improvement in visual quality and faster encoding and decoding than the original one. Secondly, we incorporate the idea to facilitate resolution scalable decoding (not incorporated in original SPIHT) by rearranging the order of the...
Complexity reduction of test zonal search for fast motion estimation in uni-prediction of High Efficiency Video Coding
Journal of Real-Time Image Processing
Quality of Experience Driven Rate Adaptation for Adaptive HTTP Streaming
IEEE Transactions on Broadcasting
Indonesian Journal of Electrical Engineering and Computer Science
Road traffic congestion, a serious illness in developing regions, is one of the biggest problems ... more Road traffic congestion, a serious illness in developing regions, is one of the biggest problems in our day-today life, resulting in delays, wastage of fuel and money. In this paper, a new model is developed using Simulation of Urban Mobility (SUMO) simulator for simulating a realistic traffic scenario for a large city like Bhubaneswar where, traffic congestion is a critical issue. In a city, traffic congestion is characterised by many parameters such as rapid growth of population, number of four wheelers, inadequate and poor road infrastructures and shortage of physical plan to govern the developments, which are focused on enhancing the volume of the roads by raising the number of lanes, overpasses , underpasses and over-bridges at many junctions. However, for the success of these master plans to fully overcome the congestion issues, it is necessary to transmit the congestion information to vehicles coming towards a congestion area by using a Vehicular Ad-hoc Network. This paper analyzes clustering techniques in Vehicular Ad-hoc Networks to detect congestion in roads with the minimal infrastructural support. The raw data from vehicles are classified using cluster analysis. Out of a number of algorithms that are used to solve the congestion detection problem, three important algorithms such as Centroid based K-means, object based FCM and FKM algorithms are compared in this work on the basis of data points and number of clusters. The results of the algorithms are close to each other, but fuzzy techniques are preferable as the traffic situations are dynamic in nature.
Pattern Recognition Letters
It has come to the attention of the Editors-in-Chief of Pattern Recognition Letters (PRL) that an... more It has come to the attention of the Editors-in-Chief of Pattern Recognition Letters (PRL) that an article ''Invariance image analysis using modified Zernike moments'', N.K.
Modified Frei-Chen Operator-Based Infrared and Visible Sensor Image Fusion for Real-Time Applications
IEEE Sensors Journal
Search range reduction for uni-prediction and bi-prediction in HEVC
Journal of Real-Time Image Processing, 2016
A multi-stage temporal pooling mechanism for video quality assessment
2016 50th Asilomar Conference on Signals, Systems and Computers, 2016
Uploads
Papers by Sudipta Mahapatra