The primary benefit of the cloud is that it elastically scales to meet variable demand and it provides the environment which scales up and scales down instantly according to the demand, so it needs great protection from DDoS attacks to tackle downtime effects of DDoS Attacks. Distribute Denial of Service attacks fall on the category of critical attacks that compromise the availability of the network. These attacks have become sophisticated and continue to grow at a rapid pace so to detect and counter these attacks have become a challenging task. This work was carried out on the owncloud environment using Tor Hammer as an attacking tool and a new dataset was generated with Intrusion Detection System. This work incorporates various machine learning algorithms: Support Vector Machine, Naive Bayes, and Random Forest for classification and overall accuracy was 99.7%, 97.6% and 98.0% of Support Vector Machine, Random Forest and Naive Bayes respectively.

Delay Tolerant Network is a buzz word in Mobile Adhoc Network. Research in this direction started with the approach for interplanetary communication. Delay Tolerant Network (in short) DTN is an evolving research area. A way to store, carry and forward messages over the underlying existing infrastructure. The intermittent connectivity and limited resources of a DTN node make the routing difficult to implement in real world application. It is a constant endeavor of any network community to look for security and privacy inclusions in routing protocols. This paper proposes a Light Weight Security algorithm that forwards messages only to the trusted nodes, thus prevents the malicious or selfish nodes from affecting the entire DTN communication. In this paper, the trust value is computed based on their one to one association of mobile nodes in a DTN. Researchers have contributed by proposing a novel solution for detecting hostile mobile nodes. The computed trust value is further used to prevent malicious nodes to compromise the network. The proposed algorithm is simulated and its performance is analyzed with existing based on the performance metrics like overhead ratio, relayed, removed, delivery ratio and hop count. DTN is still evolving and to the best of our knowledge; the proposed work is first of its kind.

Delay Tolerant Networks fall into the category of networks that works over the underlying TCP/IP infrastructure. DTN is capable of using the stationery as well as mobile routers for communication. Although the delay involved is not known. Due to such challenged class of networks, in the environment that is characterized by frequent node movement, low energy devices and limited storage, devices are able to communicate. In such challenged classes of networking effective optimization of computation, storage and communication resources is very essential. In order to make the DTN routing feasible in real world applications, it is essential to critically address the question whether what is the sufficient buffer size for each routing protocol? The researchers have simulated and analyzed the impact of variable buffer size for various routing protocol in objective manner. To the best of our knowledge the proposed work is first to provide comparative results with varying buffer size on various routing protocols. Further the performance results are important for practical implementation of the DTN Routing Protocols in real life scenarios.

India is a country where major source of living and economy is through agriculture and agricultural industry Prediction of prices in agricultural commodity has always been a major problem for the farmers. Farmers aren't able to get desired price for their crops and that's why the suicide rate is increasing with every passing year. This paper mainly focuses on proposing a decision making support model for prediction of prices in agricultural commodities. This paper also includes techniques of data mining in agriculture that will help the farmers to predict the agricultural commodity prices. The objective of this project is to build a system which provides efficient and effective price prediction features. The aim is to propose a new framework and develop a system to make some advances towards more efficient price prediction.