Q-Learning is an important class of stochastic optimization which has recently been used in the area of dynamic adaptive streaming over HTTP (DASH). Though DASH is very popular method of video delivery in recent years it is plagued with problems when multiple players share a bottleneck link. Thus, this area has become a very active area of research. Two works which implement Q-Learning in DASH are selected and their performances compared against the Conventional DASH player. It is shown that Q-Learning works well for various network conditions.

Sensor cloud is a new technological development and improvement of numerous day to day life applications. It is in the form to embed for cloud computing with wireless Sensor network. As wireless no more furnishes with processing, information storage and exploring but it all done by the use of cloud and also it help to overcome problems like efficiency, lifespan and energy supply. The proposed methodology works on push pull mechanisms to save the node’s energy and to improve the network lifetime. The two way communication from sensor to cloud is made possible along with making nodes in active and sleep mode to preserve the node’s energy and prolong the lifespan of network. Push mechanism is responsible for sending the active nodes data to cloud server and push is operated for two ways communication i.e. user request are executed by communicating with sensor nodes from cloud server and intern sensor network provide the required data to cloud server. The result shows that there is an improvement in saving node energy, less end to end delay and hence prolonging network lifetime.

In current days an effective design of WSN has became a leading area of research with the wide application of WSN, data transmission in network becoming a more hot research topic. The biggest challenge is to transmit the data from source to sink node securely. The main disadvantage in WSN is they have limited resources, cluster network have been proposed by many researchers to reduce the energy load & reduces the overhead in the network to increase the network life time. The SLEACH is the modified version of LEACH is a clustering selection method that provides an effective way to minimize the energy utilization along with providing security to the WSN, but still energy efficient for SLEACH is not up to the mark and it doesn’t guarantee confidentiality, availability and secure transmission of data from one end to another end. This paper introduces a lightweight Secure data authentication scheme (LWSDAS) which is based on Elliptic curve cryptography (ECC) to provide integrity, confidentiality, authentication & data aggregation and it is also capable to protect against attacks. The proposed work is more energy efficient over SLEACH. The results of simulation shows that the proposed work is more advantageous than SLEACH, considering parameter like end to end delay ,energy overhead, packet delivery ratio and overall it is up to 7-10% far better than SLEACH algorithm.

The IoT is the current revolt in Internet, mobile and more technologies. By adding intellect to the physical objects the objects can communicate each other to perform the desired job which reduces the human intervention. In this paper we propose a context aware load balancing in IoT. Load balancing is the splitting of traffic from one path to the other path to avoid congestion in the network. Load balancing is done by computing multipath routing, and multipath routing is computed based on context of data such as weight factor and quality of link. In this paper as the source mode initiates the route discovery by flooding the route request message to the destination in the network and waits for route reply message while migrated route packet collects the information such as node id, energy, distance, neighbor node count and mobility of nodes from all the intermediate nodes and gives it to the destination node. Once the destination receives the packet, it computes the node disjoint paths and assigns the priorities based on the type of request received. The proposed scheme is simulated for its effectiveness and the performance parameters used are path discovery time, load balancing time, throughput and packet delivery ratio.

Wireless Sensor Networks (WSN) is composed of small sensor nodes which may be hundreds or multi hundreds or thousands in number. Sensor nodes, also known as mote, are small, lightweight and portable devices equipped with a transducer, microcomputer, transceiver, and power source. Based on the network topology, routing protocols in sensor networks can be classified as flat-based routing, hierarchical-based routing and location-based routing. This paper studied some hierarchical-based routing protocols and evaluated these protocols with different cluster head probability in medium network with 200 nodes number. Protocols like Low Energy Adaptive Clustering Hierarchy (LEACH), Distributed Energy-Efficient Clustering (DEEC), Threshold sensitive Energy Efficient sensor Network protocol (TEEN) and Stable Election Protocol (SEP) are used for our comparisons. We evaluate the performance of these protocols for a cluster head probability sensing application. Cluster head Probability effects on throughput, overhead, packet delivery ratio, alive nodes and dead nodes, as parameters used to measure the performance of these protocols. We observed new results and different comparisons for hierarchical protocols in WSN.

Object detection, classification and tracking are prime components in all computer vision application. The research problem here is to detect, classify and track small objects (such as ships, boats etc.) on maritime scenario. Main purpose of object detection in maritime is to secure the country from various rocket launchers, sea side firearms. According to today’s scenario, security is very important in maritime applications. This paper showcases experiment results for object detection using Speeded Up Robust Features (SURF), Binary Robust Invariant Scalable Key points (BRISK), Lukas-kanade on standard dataset IPATCH PETS 2016. However, the state of the art methods limits in handle camouflage scenario. This paper proposes an adaptive Lucas-kanade approach that handles such scenarios. The proposed approach utilizes interaction of arithmetic mean and histogram equalization with optical flow (Lucas-Kanade) approach to resolve camouflage. Finally, the proposed approach is evaluated using standard parameters such as recall, precision and f1 score. The performance measures depict that the proposed approach outperforms state of the art trackers.

The techniques of data mining are very popular of Diseases. The advancement in health analysis has been improved by technical advances in computation, automation and data mining. Nowadays, data mining is getting used in a vast area .The Nature of the medical field is made with the knowledge wherever there’s a spread of data but untapped during a correct. and thus, the foremost serious challenge facing this area is the quality of service provided which suggests to create the diagnose during a correct manner in a timely manner and supply acceptable medications to patients. Thus Health information technology has emerged as a replacement technology within the health care sector in a short amount by utilizing Business Intelligence ‘BI’ that could be a data-driven Decision Support System. The various techniques of data mining are used and compared during this analysis.

The performance of today’s adaptive video streaming players (DASH) is severely hindered by overlapping ON-OFF patterns that occurs during a streaming session. High switching rates, freezing and skipping annoy users resulting in poor user-QoE. This paper attempts to overcome the ON-OFF problem by keeping players aware of each other’s downloads and inter-request times. Using this a player is able to better predict future player actions and reschedule their downloads to produce least conflicts with others. If a player’s start segment download overlaps with the end of one or more players’ current downloads within a certain time t1 it waits until the download of others completes before starting its own download. Conversely if too many players are sharing the bottleneck link a player will hold of its download by a time t2 enabling others to move towards completion of their downloads. This reduces the overlap of its download with other players. Both these methods help reduce competition at the bottleneck and produces a more balanced sharing of network resources.

Wireless sensor network (WSN) consists of small devices, which are called sensors. It is capable of sensing the environmental events, make processing of them and send data to the base station (BS), which needs high energy for its usage. This network which is limited to iterate the dead nodes, bring by energy depletion and to maximize the life-span of the system. Many routing protocols have been proposed and the efficiency of WSN declines as changing of the parameters of sensor nodes. The protocols in WSN are classified to heterogeneous or homogeneous. In this paper, we test the effects of node density and network area on various distributed energyefficient clustering based on protocols such as Distributed Energy-Efficient Clustering (DEEC), Developed Distributed Energy-Efficient Clustering (DDEEC) and Threshold Distributed Energy-Efficient Clustering (TDEEC) as multilevel heterogeneous protocols, and MODLEACH protocols as an example of homogeneous routing protocols. Threshold Distributed Energy-Efficient Clustering protocol has better performance than Distributed Energy-Efficient Clustering protocol, Distributed Energy-Efficient Clustering protocol and Enhanced Distributed Energy-Efficient Clustering protocol (EDEEC) but Modified Low Energy Adaptive Clustering Hierarchy protocol (MODLEACH) is lengthy the stable period other than protocols. The sent packet to BS and the received one from BS are increased with increasing of nodes number and decreased with increasing of network area. The life time of network decreases conversely with increasing the area of transmission. These parameters will increase the performance of the entire network. Especially in real-time applications that use the WSNs, which, are expected to work in fields such as industry, rubout or battle tracking.

The Intrusion detection from the last some decades are very important for current as well as for present networks. In the recent times many new methods have been used for IDS with machine learning technique and analysis for huge data is very much suitable. But the techniques like WMV (weighted majority voting) which have large dataset will take much more amount of time with this there is degradation of results whenever increasing the dataset. For this problem this paper focuses on the Extreme learning machine would be the best suitable for IDS with the analysis of big data and improving the accuracy. The proposed technique will integrate Mutual information ranking filter and attribute ranking feature selection with ELM technique. The Mutual information technique will implement attribute selection and will analyse proposed technique performance MI-ELM technique with algorithms like Modified Naïve Bayes, Support vector machine, LP Boosting and also hybrid of these three algorithms with respect to Precision, Recall. F-measure, accuracy, the (KS) Kappa statistic, Incorrectly and Correctly (CI) classified instances, RMSE (Root of Mean square erratum or error) and RRE (Root relative of error). There will be analysis of the dataset on according to basis of traffic is it normal or abnormal and the experimental results has shown that there will be increased accuracy in comparison with the classifiers.