QoS-aware and Policy Based Mobile Data O oading

摘要

The rapid growth in the number of 3G/4G enabled devices such as smart-phones and tablets has created exceptional demand for ubiquitous con-nectivity and high quality applications. As a result, cellular networks arestruggling to keep up with this explosive demand for data trac. Theemergence to LTE has boosted cellular network throughput; however theseimprovements are not sucient given the limited availability of licensedspectrum. To meet the requirements of capacity-hungry applications, Wi-Fi ooading has been intensively researched as an essential approach toalleviate the mobile data trac load on cellular network by providing ex-tra capacity and improving overall performance. The ooading algorithmsshould be evaluated and compared to steer Wi-Fi ooading to increase thecombined throughput and network performance of LTE and Wi-Fi accesstechnologies connected to the evolved packet core (EPC) with at least thebaseline case of having all the data trac in LTE. In this thesis, novel of-oading algorithms are proposed and implemented to address challenges inWi-Fi ooading to LTE networks and provide solutions when performanceneeds exceed the capability of the LTE access network. In the design ofsuch smart ooading techniques, important issues such as scalability andstability are being considered. Through an extensive set of simulations, theperformance of the proposed techniques is thoroughly investigated focusingon the gure of merits that aects user experience. The end-to-end through-put that a flow can accomplish, ooading eciency and packet droppingrate are examined. Furthermore, these evaluations have demonstrated thatooading users from LTE to Wi-Fi reduces burden on the LTE networkwithout aecting user experience. Also it is shown that the mobile com-munication architecture can be improved further by applying the principlesof Software-Dened Networking (SDN) with providing logically centralizedcontrol of the overall infrastructure, and enabling programmability.