Thriveni J

Cloud enabled 3D tablet design for medical applications

The prime objective of any technological innovation is to improve the life of people. Technological innovation in the field of medical devices directly touches the lives of millions of people; not just patients but doctors and other technicians as well. Serving these care givers is serving humanity. Growth of Mobil Devices and Cloud Computing has changed the way we live and work. We try to bring the benefits of these technological innovations to the medical field via equipment which can improve the working efficiencies and capabilities of the medical professionals and technicians. The improvements in the camera and image processing capabilities of the Mobile Devices coupled with their improved processing power and an infinite processing and storage offered by Cloud Computing infrastructure opens up a window of opportunity to use them in the specialized field like microsurgery. To enable microsurgery, surgeons use optical microscope to zoom into the working area to get better visibility and control. However, these devices suffer from various drawbacks and are not comfortable to use. We build a Tablet with large stereoscopic screen allowing glasses free 3D display enabled by cameras capable of capturing 3D video and enhanced by an image processing pipeline, greatly improves the visibility and viewing comfort of the surgeon. Moreover using the capabilities of Cloud computing, these surgeries can be recorded and streamed live for education, training and consultation. An expert sitting in a geographically remote location can guide the surgeon performing the surgery. All vital parameters of the patient undergoing surgery can be shown as an overlay on the Tablet screen so that the surgeon is alerted of any parameter going beyond limit. Developing this kind of complex device involves engineering skills in hardware and software and huge amount of investments in terms of time, resources and money. To accelerate the development, we make use of open source hardware and software and demonstrate how we can accelerate the development using these open source resources.

Mobile cloud computing for medical applications

Mobile Devices like Smartphones and Tablets are getting ever increasing processing power. This makes them powerful enough to do heavy duty realtime 3D video processing. Tablets capable of recording 3D video using stereo camera can have applications in various fields. In the medical field these tablets can be used for micromanipulations such as microsurgery. We describe the design of a tablet capable of 3D vision and a glasses free stereoscopic display, which can be used to enhance the working of the technicians performing micromanipulations. We also illustrate how by providing a deep integration with the cloud, we can reduce the processing overheads and enable valuable new services. For example, in case of microsurgeries, the process is very specialized and is done in a few specialized medical centers only. When these powerful mobile devices are coupled with the cloud, it adds whole new dimension in the services which can be offered. With our deep cloud integration, the surgery being performed can be streamed in realtime from the Tablet itself. This live feed can be viewed by a remote expert to provide guidance to the surgeon. The students at remote locations can also view this live feed. This would provide them a unique learning opportunity. Moreover, the recordings of these surgeries can be backed up on cloud and provided for on demand viewing for learning purposes.

QoS Preemptive Routing with Bandwidth Estimation for Improved Performance in Ad Hoc Networks

Ad hoc networks are collection of mobile nodes which communicate using wireless media without any fixed infrastructure. Each mobile host acts both as a network router and an end point as packets are forwarded over multiple hops. Node mobility being the distinguishing feature of the mobile ad hoc network, makes routing an important challenge. Hence, Quality of Service (QoS) is not easily achieved in ad hoc networks. To provide QoS routing in Ad hoc networks, it is not just sufficient to provide a basic routing functionality where only a feasible route is found, other aspects like the bandwidth constraints due to shared media, dynamic topology due to continuously changing topology and the power consumption due to limited battery powers must also be considered. The nodes in the Ad hoc network move randomly, hence, the topology changes continuously resulting in route breakage. Thus, the route breakage must first be detected and a new route should be found to the destination in advance. This paper proposes a QoS Preemptive Routing protocol with Bandwidth estimation (QPRB) that computes the available bandwidth in the route and then sets up the route based on the network traffic and maintains the route using preemptive routing procedure. This protocol provides QoS support to the real time applications by providing a feedback about the network status. The algorithm improves network performance and performs well in route breakage conditions as better routes are found in advance to route breakage. The cost incurred to detect the route breakage and to find a new route is avoided. Simulation results show that the protocol significantly improves the packet delivery ratio and reduces the end-to-end delay in the route breakage conditions.

Efficient video transfer using LAN caching assisted by cloud computing

There is a good probability of accessing same video content multiple times from a cloud based Video Streaming Server by same peer or different peers of a given LAN, effectively increasing Internet bandwidth or data flow for same content from server to client, thereby over loading routers between server and client and also resulting in higher power consumption at routers. This proposed concept tries to avoid multiple streaming of high volume video files from Server by caching first successful streamed data on to LAN peer which is currently viewing the video data and subsequently the same LAN peer streaming the video to other desiring peers when demanded for. Proposed implementation model retains all other server activities with server except for allowing an available LAN peer copy of video to be streamed to another peer of the same LAN when requested for.

Multi model Personal Authentication using Finger vein and Face Images (MPAFFI)

Biometric based identifications are widely adopted for personnel identification. The unimodal recognition systems currently suffer from noisy data, spoofing attacks, biometric sensor data quality and many more. Robust personnel recognition considering multimodal biometric traits can be achieved. This paper introduces the Multimodal Personnel Authentication using Finger vein and Face Images (MPAFFI) considering the Finger Vein and Face biometric traits. The use of Magnitude and Phase features obtained from Gabor Kernels is considered to define the biometric traits of personnel. The biometric feature space is reduced using Fischer Score and Linear Discriminate Analysis. Personnel recognition is achieved using the weighted K-nearest neighbor classifier. The experimental study presented in the paper considers the (Group of Machine Learning and Applications, Shandong University-Homologous Multimodal Traits) SDUMLA - HMT multimodal biometric dataset. The performance of the MPAFFI is compared with the existing recognition systems and the performance improvement is proved through the results obtained.

Cognition based self-organizing maps (CSOM) for Intrusion detection in wireless networks

Cognitive networks is the solution for the problems existing on the current networks. Users maintain integrity of the networks and user node activity monitoring is required for provision of security. Cognitive Networks discussed in this paper not only monitor user node activity but also take preventive measures if user node transactions are malicious. The intelligence in cognitive engine is realized using self organizing maps (CSOMs). Gaussian and Mexican Hat neighbor learning functions have been evaluated to realize CSOMs. Experimental study proves the efficiency of Gaussian Learning function is better for cognition engine. The cognition engine realized is evaluated for malicious node detection in dynamic networks. The proposed concept results in better Intrusion detection rate as compared to existing approaches.

EDSC: Efficient document subspace clustering technique for high-dimensional data

With the advancement in the pervasive technology, there is a spontaneous rise in the size of the data. Such data are generated from various forms of resources right from individual to organization level. Due to the characteristics of unstructured or semi-structuredness in data representation, the existing data analytics approaches are not directly applicable which leads to curse of dimensionality problem. Hence, this paper presents an Efficient Document Subspace Clustering (EDSC) technique for high-dimensional data that contributes to the existing system with respect to identification by eliminating the redundant data. The discrete segmentation of data points are used to explicitly expose the dimensionality of hidden subspaces in the clusters. The outcome of the proposed system was compared with existing system to find the effective document clustering process for high-dimensional data. The processing time of EDSC for subspace clustering is reduced by 50% as compared to the existing system.

Bimodal Biometric Verification Mechanism using fingerprint and face images(BBVMFF)

An increased demand of biometric authentication coupled with automation of systems is observed in the recent times. Generally biometric recognition systems currently used consider only a single biometric characteristic for verification or authentication. Researchers have proved the inefficiencies in unimodal biometric systems and propagated the adoption of multimodal biometric systems for verification. This paper introduces Bi-modal Biometric Verification Mechanism using Fingerprint and Face (BBVMFF). The BBVMFF considers the frontal face and fingerprint biometric characteristics of users for verification. The BBVMFF Considers both the Gabor phase and magnitude features as biometric trait definitions and simple lightweight feature level fusion algorithm. The fusion algorithm proposed enables the applicability of the proposed BBVMFF in unimodal and Bi-modal modes proved by the experimental results presented.

Cloud resource reduction evaluation by video caching and streaming in LAN environment

Presumably Internet happens to be the first choice for on-demand entertainment, and HD-video contributes to major share of Internet bandwidth consumed. Adding to this there is considerable number of audience who intend to watch particular content more than once. Given the best effort delivery mechanism of Internet along with clogged data centers, there is a huge load on cloud in terms of data bandwidth, throughput and power consumption. This paper evaluates Internet traffic reduction and network infrastructure power reduction gains of proposed video caching and streaming mechanism in LAN environment.

Reinforcement based Cognitive Algorithms to Detect Malicious Node in Wireless Networks

The growth of wireless communication technologies and its applications leads to many security issues. Malicious node detection is one among the major security issues. Adoption of cognition can detect and Prevent malicious activities in the wireless networks. To achieve cognition into wireless networks, we are using reinforcement learning techniques. By using the existing reinforcement techniques, we have proposed GreedyQ cognitive (GQC) and SoftSARSA cognitive (SSC) algorithms for malicious node detection and the performances among these algorithms are evaluated and the result shows SSC algorithm is best algorithm. The proposed algorithms perform better in malicious node detection as compared to the existing algorithms.