looks like an interesting paper and the abstract seems to confirm my theory about latency spikes and packet loss. i'll have to give it a thorough read tonight. thanks!