Inference and Evaluation of Split-Connection Approaches in Cellular Data Networks

Numerous mechanisms have been proposed for improving TCP performance over wireless links, including those in wireless cellular networks. In this paper, we infer the existence and investigate the performance of one class of these performance-enhancing approaches, split-connection approaches, in commercial cellular data networks. Special attention is given to the so called Split-TCP and TCP proxy approaches. We present inference techniques to identify whether a cellular provider implements Split-TCP or TCP proxy. Through end-to-end measurements over commercial cellular networks operated by three different providers (including two CDMA2000 networks and one GPRS network), we find that all three providers selectively implement Split-TCP or TCP proxy for certain applications (e.g., HTTP). We also find that the implementations differ from provider to provider. Experimental results demonstrate that the performance gains from Split-TCP depend on flow sizes and that TCP proxy using data compression outperforms Split-TCP.