Performance evaluation and viability of IFOM in heterogeneous LTE — WLAN network

Abstract

As the Internet usage on mobile networks is increasing at an alarming rate, operators need techniques to provide more capacity and bandwidth. One solution proposed by the 3rd Generation Partnership Project (3GPP) is to offload data to other access networks. The 3GPP architecture is designed to work along with non 3GPP access networks like Wireless Local Area Networks (WLANs). IP Flow Mobility (IFOM) is one method standardized, that enables offloading of data to these alternate accesses. IFOM enabled User Equipment (UEs) have multiple interfaces which allows them to connect to different access networks such as Long Term Evolution (LTE) and WLANs simultaneously. Thus concurrent flows with different traffic types can exist over these networks, and selective and seamless switching of an IP flow from one access network to the other is feasible. Our objective in this paper, is to investigate the benefits of IFOM on UE throughput in Heterogeneous Networks (HetNets) for different traffic types and latencies under diverse network environments, loads and fading conditions. We have carried out extensive ns-3 simulations. Our simulation results show that the IFOM enabled UEs experience high throughput at latencies below 18 ms for Voice over Internet Protocol (VoIP) and 50 ms for video and higher throughput with lower latencies in the proximity of WLANs. We further propose an algorithm to switch traffic flows for IFOM enabled UEs so that, the average per UE throughput is maximized, to provide a better Quality of service (QoS). We carry out simulations with the same and achieve a significant 14% to 30% increase in the average per UE throughput under varying load conditions.