Enhancing Mobile Data Offloading with Mobility Prediction and Prefetching

TL;DR

This paper proposes mobility prediction and prefetching techniques to improve traffic offloading from mobile networks to WiFi hotspots, enhancing efficiency for both delay-tolerant and delay-sensitive data transfer.

Contribution

It introduces new procedures that leverage mobility prediction and prefetching, evaluated through empirical measurements for their effectiveness and robustness.

Findings

Increased percentage of offloaded traffic

Reduced data transfer delay

Robustness to estimation errors

Abstract

We present procedures that exploit mobility prediction and prefetching to enhance offloading of traffic from mobile networks to WiFi hotspots, for both delay tolerant and delay sensitive traffic. We evaluate the procedures in terms of the percentage of offloaded traffic, the data transfer delay, and the cache size used for prefetching. The evaluation considers empirical measurements and shows how various parameters influence the performance of the procedures, and their robustness to time and throughput estimation errors.