Person: Srivastava, Sanjay
Loading...
Name
Sanjay Srivastava
Job Title
Faculty
Email Address
Telephone
079-68261547
Birth Date
Specialization
Internet of Things, Protocol Modelling and Analysis, Simulation
10 results
Search Results
Now showing 1 - 10 of 10
Publication Metadata only A Review of Distributed Scheduling Algorithms for Tree based Wireless Sensor Networks(IJNGC, 01-11-2019) Vasavada, Tejas Mukesh; Srivastava, Sanjay; DA-IICT, GandhinagarPublication Metadata only Weather impact on macroscopic traffic stream variables prediction using recurrent learning approach(Taylor & Francis Publications, 02-01-2023) Srivastava, Sanjay; DA-IICT, Gandhinagar; Nigam, Archana (201421001)Accurate prediction of the macroscopic traffic stream variables is essential for traffic operation and management in an intelligent transportation system. Adverse weather conditions like fog, rainfall, and snowfall affect the driver�s visibility, vehicle mobility, and road capacity. The rainfall effect on traffic is not directly proportional to the distance between the weather station and the road because of terrain feature constraints. The prolonged rainfall weakens the drainage system, affects soil absorption capability, which causes waterlogging. The weather event has a spatiotemporal correlation with traffic stream variables, as waterlogging on the road due to rainfall affects the traffic on adjacent roads. The spatiotemporal and prolonged impact of rainfall is not studied in the literature. In this research, we examine whether the inclusion of the rainfall variable improves the traffic stream variables prediction of a deep learning model or not. We use the RNN and LSTM models to capture the spatiotemporal correlation between traffic and rainfall data using past and current traffic and weather information. To capture the prolonged impact of rainfall more extended past sequence of rainfall data than traffic data is used in this study. The roads prone to waterlogging are more affected due to rainfall compared to freeways. Thus we examine the effect of rain on traffic stream variables prediction for different types of roads. The test experiments show that the inclusion of weather data improves the prediction accuracy of the model. The LSTM outperforms other models to capture the spatiotemporal relationship between the rainfall and traffic stream variables.Publication Metadata only Comparison of SUMO and SiMTraM for Indian Traffic Scenario Representation(Elsevier, 01-12-2016) Patel, Viral; Chaturvedi, Manish; Srivastava, Sanjay; DA-IICT, GandhinagarA simulation model enables abstraction of a real system and permits one to focus on interesting phenomena. At the same time, the results obtained using simulation study are eloquent and useful only if the simulation model closely represents the real world scenario. Many traffic simulators are available in literature (e.g. SUMO, VISSIM, VanetMobiSim, PARAMICS, etc. to name a few) that permit microscopic traffic simulation. However, most of them model lane based traffic movement and homogeneous vehicles. SUMO supports customized vehicle types like two wheelers, three wheelers, cars, etc. However, it enforces lane discipline on all the vehicles, including smaller vehicles (like two wheelers) which is very unrealistic. On Indian roads, large fraction of vehicles are two wheeler and traffic is less lane disciplined. To overcome the limitations of SUMO, and for better modeling of Indian traffic scenario, SiMTraM was developed by the transportation research group at IIT Bombay. SiMTraM divides lane width into multiple strips and assigns road space to vehicles in terms of number of strips, thereby permitting multiple small vehicles to share the lane width. However, SiMTraM was developed by adapting older version of SUMO (version 0.12) and after its development, was not adapted for later versions of SUMO which had many better features for more realistic traffic scenario representation. In our study we redeveloped SiMTraM by adapting SUMO 0.17 version and made it available as open source to the user community. We carried out detailed performance analysis of SUMO 0.17 and the adapted SiMTraM (now onwards called upgraded SiMTraM) with respect to the edge level vehicle flow, speed and flow-speed relationship for various traffic scenarios. The simulation results show that upgraded SiMTraM permits more realistic representation of Indian traffic scenario.Publication Metadata only Multi-Modal Design of an Intelligent Transportation System(IEEE, 01-08-2017) Chaturvedi, Manish; Srivastava, Sanjay; DA-IICT, Gandhinagar; Chaturvedi, Manish (201021007)This paper proposes a novel intelligent transportation system (ITS) using the cellular network, GPS probes, and limited ITS infrastructure for edge-level speed estimation under heterogeneous traffic condition. The erroneous vehicle position data taken from cellular network are processed in real time to compute edge level vehicle flow, space occupancy, and congestion with a mean error of less than 10%. For edge-level speed estimation, two models of ITS infrastructure deployment are proposed: the COngestion COverage MOdel (COCOMO) and the Edge COverage MOdel (ECOMO). The GPS Probes' speed data are used to extrapolate speed estimations from an infrastructure edge to the associated infrastructureless edge(s). The infrastructure requirement of COCOMO is constant, whereas that of ECOMO depends upon diversity in the congestion profile of edges. The COCOMO and ECOMO permit edge-level speed estimation with the 90 percentile error of 10%-22% and 10%-13%, respectively. The communication and storage requirement of the proposed ITS and the utility of generated traffic information are analyzed.Publication Metadata only Algorithm for fairness in schedule lengths of sink-rooted trees in multi-sink heterogeneous wireless sensor networks(Springer, 01-12-2020) Vasavada, Tejas; Srivastava, Sanjay; DA-IICT, GandhinagarSensor networks are used for observing some region of interest. Sensors sense different physical quantities and send to base station or sink. The tree based networks with TDMA as MAC protocol are preferable because of simplicity of tree and guaranteed data access in TDMA. Often sensor networks are multi-sink and multi-attribute networks. Multi-sink network means more than one sinks are present and so multiple sink-rooted trees are formed. When more than one types of nodes are deployed in the network, the network is known as heterogeneous network or multi-attribute network. Sometimes node density or heterogeneity is not uniform across entire network. As a result of non-uniform node distribution or non-uniform heterogeneity distribution, schedule lengths of sink-rooted trees are very different. Nodes part of trees with small schedule length will get more frequent transmission turns compared to those which belong to trees with large schedule length. To ensure fairness in terms of transmission opportunities, it is desired that schedule lengths should be balanced. In this work, an algorithm named as Schedule Length Balancing for Multi-sink HeTerogeneous networks (SLBMHT) is presented to balance schedule lengths. The SLBMHT algorithm is evaluated through simulations. It is found that the SLBMHT algorithm results in 8�56% reduction in schedule length difference of trees. It also results in 2�17% energy savings during data transmission phase. Only demerit is increase in control overhead. But as resulting increase in energy consumption is not much, it is overcome by savings in data energy consumption. Thus network lifetime is likely to increase.Publication Metadata only On effectiveness of cooperation enforcement mechanisms in wireless ad hoc networks(Inderscience Enterprises, 01-07-2012) Chaturvedi, Manish; Srivastava, Sanjay; DA-IICT, Gandhinagar; Chaturvedi, Manish (201021007)Wireless ad hoc networks are resource constrained, infrastructureless peer to peer networks where nodes are responsible for performing routing activity for other nodes to establish end to end communication. Due to limited resources (energy and bandwidth), in certain networks, nodes may behave selfishly and not forward packets for other nodes. Many cooperation enforcement mechanisms are proposed in literature to enforce packet forwarding on resource constrained nodes, and are shown to perform better than the defenseless dynamic source routing (DSR) protocol under their own set of assumptions. However, they do not consider the effect on network lifetime and energy consumption. Here we analyse routing overhead, end to end packet delivery ratio (PDR), and per packet energy cost of DSR and a representative set of cooperation enforcement mechanisms using various network scenario to show that in absence of infrastructure support cooperation enforcement mechanisms in their present form are not effective in dealing with node selfishness.Publication Metadata only Community aware heterogeneous human mobility (cahm): Model and analysis(Elsevier, 01-08-2015) Narmawala, Zunnun; Srivastava, Sanjay; DA-IICT, GandhinagarCommunity Aware Heterogeneous Human Mobility Model (CAHM) is based on Heterogeneous Human Walk (HHW) Yang et al. (2010) mobility model. CAHM achieves heterogeneous local popularity as observed in real mobility traces which HHW fails to achieve. It also incorporates following additional properties of human mobility: preference of nearby locations, speed as a function of distance to be traveled and power-law distributed pause time. We show that these properties make significant impact on routing protocols� performance. We also propose methods based on mathematical models to identify popular nodes within community (hubs) and in entire network (gateways) from overlapping community structure itself without doing message flooding.Publication Metadata only Hybrid deep learning models for traffic stream variables prediction during rainfall(Elsevier, 01-03-2023) Nigam, Archana; Srivastava, Sanjay; DA-IICT, Gandhinagar; Nigam, Archana (201421001)Adverse weather conditions like fog, rainfall, and snowfall affect the driver�s visibility, mobility of vehicle, and road capacity. Accurate prediction of the macroscopic traffic stream variables such as speed and flow is essential for traffic operation and management in an Intelligent Transportation System (ITS). The accurate prediction of these variables is challenging because of the traffic stream�s non-linear and complex characteristics. Deep learning models are proven to be more accurate for predicting traffic stream variables than shallow learning models because it extracts hidden abstract representation using layerwise architecture. The impact of weather conditions on traffic is dependent on various hidden features. The rainfall effect on traffic is not directly proportional to the distance between the weather station and the road because of terrain feature constraints. The prolonged rainfall weakens the drainage system, affects soil absorption capability, which causes waterlogging. Therefore, to capture the spatial and prolonged impact of weather conditions, we proposed a soft spatial and temporal threshold mechanism. To fill out the missing weather data spatial interpolation techniques are used. The traffic condition on a target road depends on the surrounding area�s traffic and weather conditions and relies on its own traffic characteristics. We designed the hybrid deep learning models, CNN-LSTM and LSTM-LSTM. The former model in the hybrid model extracts the spatiotemporal features and the latter model uses these features as memory. The latter model predicts the traffic stream variables depending upon the passed features and temporal input. We perform multiple experiments to validate the deep learning model�s performance. The experiments show that a deep learning model trained with traffic and rainfall data gives better prediction accuracy than the model trained without rainfall data. The performance of the LSTM-LSTM model is better than other models in extracting long-term dependency between the traffic and weather data.Publication Metadata only Heuristics Based Tree Switching in Two-sink Sensor Networks(INFOCOMP, 10-12-2019) Vasavada, Tejas Mukesh; Srivastava, Sanjay; DA-IICT, GandhinagarIn sensor networks, tree is a well-known topology formation method and TDMA is a desirable MAC protocol due to guaranteed channel access and no collisions. Many times node distribution across the region is not uniform. If finer observations are required in a region, node density is kept high. But in other regions where accurate readings are not needed, network may be sparse. Often multiple sinks are deployed in WSNs. Use of multiple sinks provides fault tolerance and load balancing. When multiple sinks are deployed, more than one sink-rooted trees are formed. The trees with dense node deployment would have higher schedule lengths than the trees with sparse node deployment. Thus trees part of the same network have different schedule lengths. In other words, schedule lengths are not balanced. As a result, nodes of some trees (with higher schedule length) have to wait for longer duration for transmission turn compared to the nodes of the other trees (with lower schedule length). As all the nodes belong to the same network, it is desirable that the waiting time for transmission turn should not be very different. So, schedule length balancing is required to ensure fairness. In this work, an algorithm known as HTSTSN (Heuristics based Tree Switching in Two-sink Sensor Networks) algorithm for two-sink network is proposed. It helps every node to decide which sink (i.e. tree) to join such that schedule lengths of trees remain balanced. The HTSTSN algorithm executes before actual scheduling algorithm. It is shown through simulations that the proposed algorithm results in average 13% �to 74% reduction in schedule length difference and maximum 12% increase in energy consumption. It is found that the HTSTSN algorithm balances schedule length without much affecting the network lifetime.Publication Metadata only A study on creating energy efficient cloud-connected user applications using the RMVRVM paradigm(Elsevier, 01-07-2024) Singh, Lavneet; Tiwari, Saurabh; Srivastava, Sanjay; DA-IICT, GandhinagarMany applications that run on smartphones are heavy on User Interface (UI) and depend on back-end services deployed on the cloud to fetch the required data through REST-based API. Because of the large number of devices actively being used, their collective energy consumption is very significant. Saving energy on these devices is beneficial not only for reducing the carbon footprint but also for the end users as it results in longer battery life. The data fetched by these applications using the REST API is generally processed, filtered and made compliant with the user interface through a paradigm called the Model View View-Model (MVVM).