Remote Path Identification using Packet Pair Technique to Strengthen the Security for Online Applications

Abstract

As the technology grows in size, complexity connections, new applications and service requirements necessitate the development of suitable end-to-end tools that monitor and measure the properties of the paths become essential. In particular, the effective evaluation and measurement of the capacities along paths are of realistic interest. In fact, the end-to-end path capacity estimation in networks is much more important. Since the measurements are performed in any environment, end-hosts might try to change these measurements to increase their gain in the network, which leads to a wide range of bandwidth manipulation attacks such as packet loss, delay in packets, impersonation attacks and cross traffic, etc. This project aims to detect these security threats against bottleneck bandwidth estimation using the packet-pair technique. The security levels are adaptively enhanced for video files in packet pair technique. It further implements an algorithm for remote path identification using the distribution of packet-pair dispersions and evaluates its accuracy, robustness and potential use of online applications. This is achieved by the verifier checks the prover's claimed or assumed identity by measuring the path without active cooperation from the prover.