International Journal of Engineering and Advanced Technology (IJEAT) ISSN: 2249 – 8958, Volume-9 Issue-2, December, 2019 1361 Published By: Blue Eyes Intelligence Engineering & Sciences Publication Retrieval Number: A5089119119/2019©BEIESP DOI: 10.35940/ijeat.A5089.129219 Abstract: An important issue incurred by users that limits the use of internet is the long web access delays. Most efficient way to solve this problem is to use “Prefetching”. This paper is an attempt to dynamically monitor the network bandwidth for which a neural network-based model has been worked upon. Prefetching is an effective and efficient technique for reducing users perceived latency. It is a technique that predicts & fetches the web pages in advance corresponding to the clients’ request, that will be accessed in future. Generally, this prediction is based on the historical information that the sever maintains for each web page it serves in a chronological order. This is a speculative technique where if predictions are incorrect then prefetching adds extra traffic to the network, which is seriously negating the network performance. Therefore, there is critical need of a mechanism that could analyze the network bandwidth of the system before prefetching is done. Based on network conditions, this model not only guides if the prefetching should be done or not but also tells number of pages which are to be prefetched in advance so that network bandwidth can be effectively utilized. Proposed control mechanism has been validated using NS-2 simulator and thus various adverse effects of prefetching in terms of response time and bandwidth utilization have been reduced. Keywords: Network Bandwidth, Neural Network, Prediction, Prefetching I. INTRODUCTION Due to enormous information present on the World Wide Web, users have been experiencing long delays while accessing files from World Wide Web. Prefetching is the solution to render these delays. The intent behind prefetching is to take benefit of the idle time between two network accesses i.e. when users are viewing the web pages which are just downloaded. In this idle period, prefetching estimates and fetches the additional web pages which will be accessed in near future based on some intelligence added to the applications so that users‟ waiting time can be reduced and thus experience of using Internet could be improved. If the prefetched web pages are indeed requested, these can be accessed with negligible delay. If the system could exactly predict those web pages which a user will request next, we will fetch only those web pages in advance and user will enjoy zero latency. Unfortunately, some prefetched web pages may never be used which results in wastage of network bandwidth and adds to the principal cost of prefetching. In literature, there are a lot of prefetching techniques discussing prediction Revised Manuscript Received on December 15, 2019. Sonia Setia, Computer Science, YMCA, Faridabad, India. Email: setiasonia53@gmail.com Jyoti, Computer Science, YMCA, Faridabad, India. Email: justjyoti.verma@gmail.com Neelam Duhan, Computer Science, YMCA, Faridabad, India. Email: neelam.duhan@gmail.com algorithms, their accuracy, precision and hit ratio etc. which are mainly its host aspects. Second aspect is networking aspect of the prefetching i.e. how to determine the number of web pages to prefetch to reduce its adverse effects on the network. Though, prefetching is taking advantage of users‟ idle time, however, it is also necessary to consider whether network is idle at prefetching time or not. Based on these two aspects, prefetching scheme basically consists of two modules: A. Prediction Module After a users‟ current request is satisfied, prediction module immediately starts working and predicts the future requests of the user by computing the probability with which the web pages will be accessed in near future. Different types of prediction algorithms have been used in literature for this module. B. Threshold Module Based on network conditions, this module takes decision for Prefetching. If it allows for prefetching then it computes value of prefetching threshold i.e. how many numbers of documents which are to be prefetched to achieve optimum performance. This module is independent of the prediction module i.e. same threshold algorithm can be applicable with different prediction algorithms. This paper focused on second aspect of prefetching i.e. Threshold module which determines the prefetch threshold based on network conditions in real time. In this view, a control mechanism has been proposed which uses the ping‟s ICMP (Internet control message protocol) messages to compute the RTT (round trip time) and network bandwidth is also measured to control the prefetch threshold so that network performance can be optimized. It employs a Neural Network model over the RTT and Network Bandwidth basis which it tells if the system is ready for prefetching or not and if yes, how many web pages to be prefetched to optimize the network usage. The remainder of the paper is organized as follows. Brief review of literature work has been given in section 2. Section 3 presented the proposed work in which an algorithm has been developed to determine the prefetch threshold based on network conditions. In section 4, summarized results of evaluation of the proposed work through trace-driven simulations have been shown. Finally, conclusion has been presented in section 5. Neural Network Based Prefetching Control Mechanism Sonia Setia, Jyoti, Neelam Duhan