Ravi Shanker ¹, Ashish Kr. Luhach ¹, Amit Sardar ¹

Affiliations
1 Department of Computer Science and Engineering, Lovely Professional University, Phagwara - 144411, Punjab, IN

Abstract

Objectives: Wireless technologies enable devices to communicate without any guided medium. It uses radio frequency for transmitting data, where the wired technologies use cables. It is mostly used to refer a telecommunications network that is interconnected between nodes and implemented without any kind of wires like as a computer network. Due to unguided media it’s very tedious to detect any intrusion during communication. The aim of this work is to develop an approach which will detect and prevent black hole attacks so the any unauthorized or unwanted nodes if participates in the communication process will be detected in the real time. Methods/Statistical Analysis: In our work we are going to propose IDS which will helps to provide security to detect attacks and to prevent the attacks. Centralize and synchronized IDS node will work on the basis of anomaly detection technique. It will create pattern for malicious activates and do event detection on the basis of anomaly detection. Findings: The proposed method will enhance the detection of anomaly detection techniques with synchronized IDS within the cluster as well as between the clusters with the help of DRI table. Application/Improvement: This technique will helps to enhance security in Wireless network and helps to make more reliable network for user.

Keywords

Attacks, Anomaly, Cluster, Intrusion Detection System, Malicious.

Full Text

   

Isha ¹, Ashish Kr. Luhach ¹

Affiliations
1 Lovely Professional University, Jalandhar - 144411, Punjab, IN

Abstract

Internet of Things (IoT) enables physical things to communicate, compute and take decisions based on any network activity. This calls for a secure solution for communication among heterogeneous devices. With the development in Information and Communication Technology (ICT), a unique impact of smart things is observed in our everyday life. IoT can consider users that interact in heterogeneous environment. In heterogeneous environment motive of each user in IoT can be different in form of communication and computation and is difficult to be judged. A malicious user can destroy the security and privacy of the network. This study gives a detailed analysis of existing security solutions for IoT. Firstly, a comparison of lightweight cryptography algorithms is made on basis of block size, key size, number of rounds, and probable attacks. Later, the various security issues in IoT are discussed along with possible solution. Security solutions in IoT will improve the trust over IoT. A secure solution that will require less computational power and is less vulnerable to existing attacks is desired.

Keywords

Internet of Things (IoT), Lightweight Algorithms, Radio Frequency Identification (RFID), Wireless Sensor Networks (WSN).

Full Text

   

Shabnam Sharma ¹, Ashish Kr. Luhach ¹, Sinha Sheik Abdhullah ¹

Affiliations
1 School of Computer Science and Engineering, Lovely Professional University, Jalandhar-Delhi, G.T. Road, National Highway 1, Phagwara - 144411, Punjab, IN

Abstract

Background/Objectives: In this research, load balancing algorithm based on Bat Algorithm is implemented to achieve the goals of load balancer. Methods/Statistical Analysis: The proposed algorithm is implemented using “Parallel Processing toolbox” in Mat lab and the results are compared with other techniques, namely, Round Robin and Fuzzy GSO. Findings: Results are discussed in the context of response time and based on the number of tasks. While balancing the load among the virtual machines, job migration to another virtual machine will affect the response time. In future, further work is on the development of job migration algorithm.

Keywords

BAT Algorithm, Cloud Computing, Load Balancing.

Full Text

   

Nishita Arora ¹, Ashish Kr. Luhach ¹, Parampreet Kaur ¹

Affiliations
1 School of Computer Sciences and Engineering, Lovely Professional University, Phagwara – 144411, Punjab, IN

Abstract

Background/Objectives: An extensive literature survey has been carried out and focused on variety of application and different research area for many data mining techniques in Multinational Company (MNC). Methods/Statistical Analysis: In this research paper we find the best cluster to extract meaningful data from large and complex pattern of neural network. This work is implemented on MATLAB Tool. By applying all the functions of research methodology we found the best cluster. This research work is compared with SOM Topology and can be used in so many applications such as biometric etc. Findings: Variety of application and different research area are identified which will be helpful and marked important for many data mining techniques in Multinational Company (MNC) and large organization for decision maker for all research resources. A methodology is proposed that focuses on zoning of neural network via membership function of Fuzzy Logic. This method removes the complexity of pattern extract the best sample easily. Application/Improvements: In future, this method will work on biometric like it will identify finger prints easily. It will divide the pattern and keep them into the zones and find the best one easily.

Keywords

Membership Function, Neurons, Zones.

Full Text

   

Virrat Devaser ¹, Ashish Kr. Luhach ¹

Affiliations
1 School of Computer Science and Engineering, Lovely Professional University, Jalandhar-Delhi G.T. Road, National Highway 1, Phagwara – 144411, Punjab, IN

Abstract

Background/Objective: Crop acreage and production estimate is of great challenge and requires strategic decisions to be taken by government agencies regarding controlling import and export duties and also to control inflation as a whole. Methods/Statistical Analysis: A lot of techniques exist to measures the yield and production for a given area and that too crop wise. We estimate the crop acreage with the help of remote sensing data and geographic information system techniques. In remote sensing, we acquire information related to the earth surface and to determine the crop acreage with the help of remote sensing images. We extract the useful information related to the yield estimation. Findings: In image processing, it is used to analyze or manipulate the features of an image. We used K-mean clustering and Self Organizing Map algorithms here for the crop Acreage and production estimation as a comparative approach to measure out acreage for a specific area image. By applying different image processing techniques and using the remote sensing techniques we estimated the production estimates for crops and compared the results with actual data obtained from statistical sources. Geographic Information System is widely used in crop acreage and production estimation to find out the locations of a particular image and helps in improving the accuracy of estimates. The challenge still remained regarding the crop segregation which requires lot of ground truth validation and hence makes the approach less viable as it reduces accuracy. Self organizing maps were found to be giving better results than other approaches. Applications/Improvement: It can be used by government agencies for taking decisions regarding import/export duty control and by agencies interested in procurement of crops and can be used by RBI to tackle Inflation as well.

Keywords

Remote Sensing, Geographic Information System, Un Supervised Classification, K Means, SOM Algorithm, Supervised Classification.

Full Text

   

Ravishanker ¹, Ashish Kr. Luhach ¹, Richa Sharma ¹

Affiliations
1 Lovely Professional University, Phagwara - 144411, Punjab, IN

Abstract

Background/Objectives: The community overlapping is the process by which number of nodes within the mesh network share common resources. The shared resources could lead to the conflict such as inconsistent analysis problem. Study of these problems is the objective of the paper. Methods/Statistical Analysis: In order to analyze the problem Modified K-Clique with sink node elimination technique is suggested. K-Clique method used detects the nodes in the mesh network having more than one connection. The modification to K-Clique enhance speed since sink node is eliminated prior to calculation of cliques. The adjacency matrix is used in order to detect the sink nodes. The Simulation is conducted in MATLAB. The MATLAB provides tools of network programming in terms of plots and graphs. The existing K-Clique is compared against the modified K-Clique and result obtained is better for Modified k-Clique. Findings: The speed is enhanced almost by 5% and number of cliques of distinct sizes discovered is also increased by 5%. The speed can further be enhanced by following hop count mechanism to reach destination quickly in addition to sink node elimination. Application/Improvement: Enhancement of performance using community overlapping detection in wireless mesh network through which it is possible to transfer the data towards multiple destinations with the help of community overlapping detection. Multiple destination towards which is to be transferred can be detected. Time will be less consumed in this case. The distance vector routing can be merged in the supposed system to further enhance the scope of the system.

Keywords

Community Overlapping, Complex Network, Distance Vector, K-Clique, Sink Nodes.

Full Text