Scalable Probe Tool Design for Active Packet Loss Measurement using Geometric Distribution
Journal Title: International Journal of Research in Computer and Communication Technology - Year 2014, Vol 3, Issue 1
Abstract
Packet loss due to congestion is a fundamental problem in wide-area packet switched networks. Researchers have expanded much effort in characterizing this phenomenon and designing protocols that operate effectively in lossy environments, but the Internet’s dynamic nature requires continued empirical evaluation of packet loss behavior. The packet loss estimation is quite challenging due to the rare occurrence and typically short duration of the packet loss episodes. Network operators have the ability to passively monitor nodes within their network for packet loss on routers using Simple network Management Protocol (SNMP). Most of the existing network probing tools work based on Poisson Arrival See Time Averages (PASTA) approach. The limitation of this approach is that it requires higher moments of measurement to report accurate results. Increasing the probe rate leads to the situation that the probes themselves may skew the result. The proposed approach mainly focuses on the design and development of a one-way UDP based active packet loss measurement tool. Among all discrete probability distributions supported on a set of events the geometric distribution the one with the largest entropy where entropy is a measure of the uncertainty associated with a random variable. The probes are generated at the sender and are dispatched to the queue at geometrically distributed interval. The aim of each probe is to record the loss frequency and the delay that is experienced along the path. It is also aimed to get the network statistics at the time of probing. The First in First Out queuing mechanism is used for buffering the packets and queue is forced to introduce the packet loss as the packet loss is a rare event in the network.
Authors and Affiliations
Gayathri c, Suma B
Keywords
Related Articles
- EP ID EP27818
- DOI -
- Views 277
- Downloads 0
VHDL Implementation of High Speed Fault Injection Tool for testing FPGA based designs
Fault injection is mainly used to test and evaluate the fault-tolerance based designs. In current VLSI technology fault injection has become a popular technique for experimentally determining dependability parameters...
Augmentation of very fast decision tree algorithm aimed at data mining
The reason for information order is to build a grouping model. The choice tree calculation is a more broad information characterization capacity estimate calculation taking into account machine learning. The choice t...
Applications of Graph Labeling in Major Areas of Computer Science
The field of Graph Theory plays a vital role in various fields. In Graph theory main problem is graph labeling. Graph Labeling is the assignment of integer’s form 1 to n for vertex, edges and both of the graphs respec...
Sprinkling Data Prevents From Intruders
The habitual communication of delay-tolerantnetwork (DTN) model is having partial infrastructural model for prevailing mobile patronsfor instance Bluetooth and Wi-Fi Direct. Mobile ad hoc network is an anthology of mo...
Implementation and Analysis of Image Contrast Enhancement using Gaussian Mixture Model
In this paper, we propose a new global contrast enhancement algorithm using the histogram color and depth images. On the basis of the histogram-modification framework, the color and depth image histograms are first p...