Ijaz Ali Shoukat

Practical assessment of remotely communicated end-point servers of leading web crawlers

Exploration of future dreams inclined us in information quest over the global literature. Web crawlers enriched with technological phenomena for such purpose to millions of seekers. The usage of crawling tools implicates justification for their consequences as searching technology that facilitates initiative passion to target new information. Several web crawlers are available with different technological mechanism of remotely communicated servers; but their efficiency in terms of different performance metrics has been an open question. Previous studies provide inadequate comparison with limited analysis factors against these servers. Here we put an effort and provided new results by considering several performance metrics under practical estimation of technical factors with maximum capacity to evaluate foremost crawling technology based end-point servers. We extensively studied remote servers running behind Google, Yahoo, MSN, Gigablast, Mamma and Excite. We reflected their performance in terms of response time, load, health status, down time, uptime, availability and reliability. These new results would provide a deep insight to every researcher regarding to searching usages and related technological enhancement in future.

Comprehensive Exploration for Proposing Hybrid Scheme upon Congestion Avoidance Algorithms

free services are ultimate preference of every network consumer and service providers. Variety of parameters like packet dropping rate, latency, jitter, throughput, bandwidth, fair response of resources, link utilization and queue length are responsible to fabricate or reduce congestion. Current TCP model for high speed networks is unstable and ineffectual due to slow response, large window size and fairness issues. The ideal and positive utilization of indicated factors can reduce congestion up to ideal strength with enhanced fairness. These entire factors cannot be handled with single congestion handling technique but a joint committee of congestion techniques can manage all these constraints. We considered packet loss as a primary congestion and fairness metric that differs with already conveyed hybrid congestion techniques that utilize delay as primary metric. We reviewed several congestion algorithms to find out most essential parameters to negate congestion in packet switched networks among the above mentioned parameters. We proposed a hybrid congestion handling technique after performing sufficient comparison with already conveyed hybrid congestion management techniques. Our propose hybrid congestion management technique (ECN + IFRC) is empirically superior to exiting hybrid congestion management techniques in some extents.

Prominent Causes of Poor Quality of Service in Network Communication

Capacity overflows, heterogeneous delay and packet losses are the resultant factors of network congestion and overloading. Such sort of dominant factors degrade the Quality of Service (QoS) and finally the network enters in a state of communication cut off. In this article, firstly we have discussed salient foundations of poor QoS such as network congestion, link overloading, routing metrics, design conflicts, server-side-scripting, denial of service attacks, RFC congestion schemes and weakness of existing TCP protocols with comparative analysis. Secondly, we have contributed some experimental observations against social, non-social and multimedia networks to realize QoS through comparison of response latencies and transmission errors. The scope of this study is to identify maximum causes of poor QoS in order to judge the effect of loaded and normal network conditions upon end-to-end service quality in both experimental and theoretical prospective. Such sort of comprehensive effort has not been attempted yet in such capacity, as we have conducted in this article. Presented analysis and experimental observations are significantly noticeable to evolve forthcoming congestion control policies.