Article Preview
TopIntroduction
The enormous growth of mobile data traffic has surprised mobile operators and vendors by choking existing communication networks a lot earlier than expected: bottlenecks of anchor nodes, sub-optimal routes and hierarchical/centralized architectures became hot research and development topics. As a result of the progressive innovation, novel technologies have emerged quickly to manage routing optimization of data flows, divert user packets around network bottlenecks and critical entities of the architecture, and to redirect traffic directly to/from the Internet. All the techniques around this problem space are called data offloading and used for any kind of complementary network technologies for optimized delivery of data originally targeted for cellular networks. The set of rules aiming at triggering the offloading action can either be set by the mobile operator or the mobile subscriber and the code operating on the rules can run in the user terminal, in a core network entity or can be divided between different elements of the architecture. For the end users the gain of applying offloading schemes relies on extended data service cost control and the availability of higher bandwidth. From the operators’ perspective the main purpose for introducing offloading in the network is to avoid congestion of the cellular architecture due to mobile data traffic evolution. Key technologies include pico- and femtocells, Wi-Fi, Content Delivery Networks (CDNs), routing and media optimization, Traffic Engineered Handover (TEHO), and Core Network Offloading as the most important schemes to reduce the congestion and load on the operator’s network.
TEHO techniques focus on decision mechanisms to be involved in radio access changes with a goal not restricted to cope with degradation of signal conditions but also to cover the improvement of traffic conditions in the IP Connectivity Access Network (IP-CAN) together with the improvement and maintenance of the user QoS/QoE. The two main offloading techniques which support TEHO in reaching the above goals are the IP Flow Mobility (IFOM) (3GPP TS 23.261, September 2010), (3GPP TR 23.829, Sept. 2010) and Multi-Access PDN Connectivity (MAPCON) (3GPP TR 23.861, February 2010). MAPCON refers to the technique of simultaneously using two or more APNs (Access Point Names) which enables use cases such as using LTE/LTE-A for QoS demanding applications and Wi-Fi for best effort traffic. IFOM refers to the technique of using the same APN across two or more wireless access networks (e.g., LTE/LTE-A and Wi-Fi) and aims to enable seamless roaming of applications across different radio access technologies. Currently the key tool to handle IP handoffs between 3GPP and non-3GPP technologies is the IEEE 802.21 Media Independent Handover (MIH) standard (IEEE, Jan. 2009) which provides information and handover assistance services for heterogeneous mobility scenarios. Another tool is the Access Network Discovery and Selection Function (ANDSF) specified in (3GPP TS 23.261, September 2010), (3GPP TS 23.402, June, 2011) for EPS. The ANDSF transfers the mobile network operator policy rules to connect through non-3GPP access technologies to the UE and thus enables an advanced traffic steering that adapts to the QoS/QoE and the actual traffic conditions of the controlled 3GPP network.
The currently standardized IFOM solution in 3GPP is strictly UE centric as the operator must first deliver the flow routing policies to the UE, and then the UE must provide these policies to the PDN Gateway. Also the ANDSF has no interface to the Policy and Charging Control (PCC) system, therefore requires other ways to get informed about the updated flow routing policy for a particular UE.