A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
Chandraprakash, V.
- Finding The Effectiveness Of Prioritized Test Suite Using Neural Networks
Authors
1 Department of Computer Science and Engineering, K. L. University, Green Fields, Vaddeswaram, Guntur - 522502, Andhra Pradesh, IN
Source
Indian Journal of Science and Technology, Vol 9, No 44 (2016), Pagination:Abstract
Objective: To find out whether a trained neural network can predict the Average percentage of Faults Detected (APFD) value for the given prioritized test suite without having the knowledge of computing APFD formula. Methods: To generate a test suite, A fault matrix containing faults and test cases is considered and for each possible permutation of test cases. The APFD value is computed for each of such test suite. The test suites with their respective APFD values are given to the neural network during training. During testing, a prioritized test suite is fed to the neural network. The APFD value predicted by the neural network is noted down. Findings: The neural network has learnt to predict APFD value of the given prioritized test suite. The predicted APFD value is compared with computed APFD value using the Root Mean Square Deviation. The deviation is found to be very low, showing that the values are very nearer. Applications/Improvements: Number of hidden layers can be increased in order to reduce the deviation further.Keywords
Artificial Neural Network, Average Percentage of Faults Detected, Test Case Prioritization.Full Text
- A Soft Computing Approach to Improve the Network Performance
Authors
1 Department of Computer Science and Engineering, K. L. University, Vaddeswaram, Guntur - 522502, Andhra Pradesh, IN
2 Department of Computer Engineering, R. H. Sapat College of Engineering, Nasik - 422005, Maharashtra, IN
Source
Indian Journal of Science and Technology, Vol 9, No 2 (2016), Pagination:Abstract
Objectives: Nowadays the Internet has become vital to each and every one. It is sensitive to link failures and node failures due to many reasons in the network connectivity. Any change in the node or link may change the routing table of many nodes. Due to this change, the routing table of many nodes may be unstable. These failures lead to increase the convergence time of network. The paper focused on reducing the convergence time of network. Methods-In this research paper, we have proposed a novel approach that keeps the value of Minimum Route Advertisement Interval (MRAI) timer fickle. In this approach, depending on the position of receiver, the value of the MRAI timer is varying. In this research, if more than one route is available to reach the destination then we have used the soft computing approach so that the load is balanced over all the paths that are reaching to the destination to avoid the congestion. Result: The approach used in this paper focuses on fast network convergence and maximum link utilization. The Fickle MRAI used in the network for reducing the convergence time up to 5 second, also the use of load balancing approach to improve link utilization of the network. Conclusion: The proposed work improves fast network convergence and utilization of bandwidth as the traffic flows from the entire shortest paths in the network.
Keywords
Autonomous System, Border Gateway Protocol, Congestion Control, Load BalancingFull Text