Extreme Value High Quantile Estimation in Wireless Broadcasting - 02/06/2009

Weiyao Xiao

The advent of practical rateless codes enables implementation of highly efficient packet-level forward error correction (FEC) strategies for reliable data broadcasting in loss-prone wireless networks. Yet, the critical question of accurately quantifying the proper amount of redundancy has remained largely unsolved. In this talk, I will first briefly discuss our results on quantifying FEC redundancy with the knowledge of the number of receivers and their homogeneous packet loss rates. I will next talk about the problem of FEC redundancy estimation when the packet loss rates among receivers are heterogenous and unknown. I will introduce the high quantile estimation technique based on Extreme Value Theory. We will then discuss about the application of extreme quantile estimation to wireless broadcasting problem and our preliminary simulation result.