Search

Mobile ad-hoc networks (MANETs) comprise a large number of mobile wireless nodes that can move in a random fashion with the capability to join or leave the network anytime. Due to the rapid growth of devices on the Internet of Things (IoT), a large number of messages are transmitted during information exchange in dense areas. It can cause congestion that results in increasing transmission delay and packet loss. This problem is more severe in larger networks with more network traffic and high mobility that enforces dynamic topology. To resolve these issues, we present a bandwidth aware routing scheme (BARS) that can avoid congestion by monitoring residual bandwidth capacity in network paths and available space in queues to cache the information. The amount of available and consumed bandwidth along with residual cache must be worked out before transmitting messages. The BARS utilizes the feedback mechanism to intimate the traffic source for adjusting the data rate according to the availability of bandwidth and queue in the routing path. We have performed extensive simulations using NS 2.35 on Ubuntu where TCL is used for node configuration, deployment, mobility and message initiation, and C language is used for modifying the functionality of AODV. The results are extracted from trace files using Perl scripts to prove the dominance of the BARS over preliminaries in terms of packet delivery ratio, throughput and end-to-end delay, and the probability of congested node for static and dynamic topologies.

Abstract Human activity monitoring in the video sequences is an intriguing computer vision domain which incorporates colossal applications, e.g., surveillance systems, human-computer interaction, and traffic control systems. In this research, our primary focus is in proposing a hybrid strategy for efficient classification of human activities from a given video sequence. The proposed method integrates four major steps: (a) segment the moving objects by fusing novel uniform segmentation and expectation maximization, (b) extract a new set of fused features using local binary patterns with histogram oriented gradient and Harlick features, (c) feature selection by novel Euclidean distance and joint entropy-PCA-based method, and (d) feature classification using multi-class support vector machine. The three benchmark datasets (MIT, CAVIAR, and BMW-10) are used for training the classifier for human classification; and for testing, we utilized multi-camera pedestrian videos along with MSR Action dataset, INRIA, and CASIA dataset. Additionally, the results are also validated using dataset recorded by our research group. For action recognition, four publicly available datasets are selected such as Weizmann, KTH, UIUC, and Muhavi to achieve recognition rates of 95.80, 99.30, 99, and 99.40%, respectively, which confirm the authenticity of our proposed work. Promising results are achieved in terms of greater precision compared to existing techniques.

The significance of automated SLR (Sign Language Recognition) proved not only in the deaf community but in various other spheres of life. The automated SLR are mainly based on the machine learning methods.PSL (Pakistani Sign Language)is an emerging area in order to benefit a big community in this region of the world. This paper presents recognition of PSL using machine learning methods. We propose an efficient and invariant method of classification of signs of PSL. This paper also presents a thorough empirical analysis of signature-based classification methods. Six different signatures are analyzed for two distinct groups of signs having total of forty five signs. Signs of PSL are close enough in terms of inter-sign similarity distancetherefore, it is a challenge to make the classification. Methodical empirical analysis proves that proposed method deals with these challenges adequately and effectively

National security has gained vital importance due to increasing number of suspicious and terrorist events across the globe. Use of different subfields of information technology has also gained much attraction of researchers and practitioners to design systems which can detect main members which are actually responsible for such kind of events. In this paper, we present a novel method to predict key players from a covert network by applying a hybrid framework. The proposed system calculates certain centrality measures for each node in the network and then applies novel hybrid classifier for detection of key players. Our system also applies anomaly detection to predict any terrorist activity in order to help law enforcement agencies to destabilize the involved network. As a proof of concept, the proposed framework has been implemented and tested using different case studies including two publicly available datasets and one local network.