M. M. Manohara Pai

Traffic monitoring and routing in VANETs — A cluster based approach

Vehicular ad hoc networks (VANETs) involves vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communication and has received a greater attention in the recent years. Though safety is the primary motive for the development of vehicular ad-hoc network, they also facilitate applications like managing the traffic flow, monitoring the road conditions, dissemination of data and many more. Efficiency of all these applications depends on density estimation. Of the various approaches clustering is demonstrated to be an effective approach. In this paper a clustering approach for traffic monitoring and routing is proposed where the Cluster Head(CH) election is done based on distance and direction information. Since clusters are formed all along the road, CHs will take the responsibility of routing the packet to destination. The simulation results show better stability, accurate density estimation in the cluster, better End-to-End delay and good delivery ratio. The protocol used is 802.11p with Wave Short Message (WSM) which is proved to be best technology designed for VANETs environment. The performance of the algorithm is analyzed using NCTUns simulator.

Density aware routing using road hierarchy for vehicular networks

Inter-vehicular communication (or Vehicular Adhoc Networks, VANET) mirrors as one of the emerging technologies for road safety and comfort applications. However the use of this technology is limited due to the lack of an efficient routing protocol. In this paper we propose Density Aware Routing using Road Hierarchy (DAR-RH) which is based on realistic road topology. DAR-RH exploits the available road hierarchy information to transfer data from source to destination. In the proposed solution two routes are maintained; in case of failure of the primary route, an alternative route is maintained at a higher hierarchy under sparse network conditions. The proposed scheme is found to exhibit significant performance improvement in comparison with a geographic routing protocol under similar conditions.

Transport and traffic rule violation monitoring service in ITS: A secured VANET cloud application

Vehicular Ad-hoc Network (VANET) cloud, a hybrid technology, provides several computational services to minimize traffic congestion, travelling time, accidents, and environmental pollution. In the proposed work, the concept of VANET cloud is used for helping the regulatory authorities in identifying the vehicles violating the traffic rules through sensors included as part of On Board Unit (OBU). When the vehicle is on fly the sensor values are periodically transferred to the cloud, controlled by the traffic police. A novel concept called Transient Ticket (TT) has been used to minimize the time and the cost of distributing Certificate Revocation List (CRL) to the vehicles. The proposed scheme also ensures utmost verification of identity, authenticity, confidentiality and integrity of the communication parties and messages exchanged. The work has been simulated using NS3 network simulator and Google App Engine (GAE). All the generated keys, TTs and the exchanged messages have been securely stored in the GAE for the ease of accessibility and processing. The results show that the proposed approach consumes very less time with respect to the generation of keys, exchange of messages, verification of authenticity and the generation of TT without compromising security.

Cloud enabled air quality detection, analysis and prediction — A smart city application for smart health

The work proposes a cloud based air quality detection system that analyzes the data for providing atmospheric quality to the user in real time. The data, collected through various gas sensors deployed outdoors in strategic locations across the city of Manipal, Karnataka is sent to the cloud via an adaptive interface that supports 2G/3G/4G infrastructure. Also, a live feed of Closed-circuit Television (CCTV) footage of some strategic locations of Manipals road traffic is sent to the cloud for analysing the density of pollutants in air with respect to the road traffic. In addition to this, the database from the Regional Transport Office (RTO) and the Computerized Pollution Check Centres of Manipal provide a basis for a comparative analysis of the variances in the spectrum of emissions from vehicles and sensor-based data coming from strategic locations. This combination along with the knowledge of the geographic and industrial properties of the area will help analyse the data for finding patterns in air quality in a particular time interval. The proposed model will then be able to predict the air quality for future days. A web and mobile application interface will help the users to check and understand the air quality at their current location. The mobile application will also notify the user about severe toxicity. People with respiratory problems will be able to get personalized notifications for poor conditions.

Mitigation of insider and outsider DoS attack against signature based authentication in VANETs

Authentication is an essential framework for safe and secure communication of messages in VANETs. For authenticating messages the IEEE 1609.2 standard uses ECDSA as the standard digital signature algorithm. But the verification time for an ECDSA signature is very high. As a result an inside or an outside attacker could use a fraction of bandwidth and flood the network with invalid signatures resulting in Denial of Service (DoS) attack. Thus in this work a two phase scheme is proposed to mitigate inside and outside DoS attackers in VANETs. In the first phase HMAC signatures computed from private and public key pair are used for authenticating the communicating entity. As only authentic users can compute the HMAC signature, DoS attack due to outside attackers is mitigated. If the entity is authentic and subjects other vehicle to DoS attack, the second phase is designed to detect the insider attackers. In this phase based on the number of invalid signatures flooded by the attacker, it is compared against a threshold value to identify the inside attacker. Hence DoS attack is mitigated due to inside as well as outside attackers. Experimental results show that the proposed scheme not only alleviates DoS attack but also performs better with negligible computational overhead.

Modified C-DRIVE: Clustering based on direction in vehicular environment

Efficiency applications in VANETs are focused on increasing the productivity of the road resources by managing the traffic flow and monitoring the road conditions. The performance of most such applications is dependent on an effective density estimation of the vehicles in the surroundings. Of the various methods, clustering demonstrates to be an effective concept to implement this. However due to high mobility a stable cluster, within a vehicular framework, is difficult to implement. In this work, we propose a new clusterhead election policy for direction based clustering algorithm C-DRIVE. This policy facilitates to attain better stability and thus accurate density estimation within the clusters. Simulation results show that the C-DRIVE is rendered stability through new clusterhead election policy by electing fewer clusterheads in the network. This supports for a better accuracy in density estimation with fewer overheads.

Cloud enabled secure communication in Vehicular Ad-hoc Networks

Security is one of the existing challenge in cloud enabled Vehicular Ad-hoc Network (VANET) communication. To address the security issues and to reduce payload overhead a novel method has been proposed for classification of VANET messages along with a Multi-hashed Binary Tree (MuBT) Algorithm. This algorithm verifies the identity and authenticity of the vehicles and messages exchanged. In addition to this, the concept of public cloud is used to store the location of vehicles, keys, messages and certificates involved in VANET communication. This ensures any time data accessibility from anywhere and helps to track the messages when necessary. The messages exchanged and the Transient Certificates (TCs) used in VANET communication is encrypted using the geolocation key of the RSU. The use of geolocation key provides location confidentiality against vehicles outside the zone and also ensures that a message can be sent only if the vehicle is physically present in the location.

Scalable routing technique using road hierarchy for vehicular networks

In this paper, we describe a new route discovery technique known as Scalable Routing Technique. This technique enhances the performance of the routing protocol Density Aware Routing using Road Hierarchy (DAR-RH) [3], which involves a route discovery procedure using Dijkstra algorithm before transferring data from source to destination. When DAR-RH is used in large road network scenarios the time complexity increases as Dijkstra algorithm considers all road intersections for route discovery. The solution proposed in this paper minimizes the road intersections to be considered for route discovery by dividing the city in terms of road hierarchies. By restricting Dijkstra algorithm to act on necessary data instead of entire city data shows significant performance improvement and effectiveness of our proposed technique when incorporated in DAR-RH.

Combining temporal interpolation and DCNN for faster recognition of micro-expressions in video sequences

Micro-expressions are the hidden human emotions that are short lived and are very hard to detect them in real time conversations. Micro-expressions recognition has proven to be an important behavior source for lie detection during crime interrogation. SMIC and CASME II are the two widely used, spontaneous micro-expressions datasets which are available publicly with baseline results that uses LBP-TOP for feature extraction. Estimation of correct parameters is the key factor for feature extraction using LBP-TOP, which results in long computation time. In this paper, the video sequences are interpolated using temporal interpolation(TIM) and then the facial features are extracted using deep convolutional neural network(DCNN) on CUDA enabled General Purpose Graphics Processing Unit(GPGPU) system. Results show that the proposed combination of DCNN and TIM can achieve better performance than the results published in baseline publications. The feature extraction time is reduced due to the usage of GPU enabled systems.

QoS in mobile networks by assigning priorities to SCTP streams

A two level QoS architecture for network mobility is proposed. At the transport layer the Mobile Stream Controlled Transport Protocol (MSCTP) is modified by assigning priorities to multiple streams in a MSCTP association. The performance of the protocol is further improved at the interface level by modifying the existing Concurrent Multipath Transfer (CMT) implementation of SCTP. The data from the application layer are classified and added into their respective streams which are then processed based on the priority of the stream. Also the transmission of data packets over the existing interfaces is based on the priority of the data, thus reducing the delay experienced by high priority data. The evaluation showed that the proposed QoS architecture achieves around 15% increase in the processing of high priority data at the transport layer and significant 50% increase in the processing of high priority data at the interface level.