Reducing Handoff Delay in SDN-based 5G Networks Using AP Clustering

Abstract

Densification is one of the 5G networks solutions to increase data rate and network capacity. This technology aims for constructing a large number of small cells in a specific area which may include different wireless technologies (LTE, IEEE 802.11, WiMAX, etc.). Small cells provide shorter distance between the cell access point and the mobile node which will reduce the signal attenuation and noise. This could generate a better connection, which goes along with the expectations of high data rate requests for the year 2020. The downside of the densification is that frequent handoffs will be a challenge for the network nodes. A software-defined network SDN, on the other hand, is another promising solution for 5G networks which controls the network through centralization. Using these two solutions, the handoffs in the network are expected to be frequent and hence a more delay will be added. To reduce such delay, a distributed-mobility-management technique is used in SDN-based networks, where the mobile node takes the handoff decision. This paper has focused on the issue of the distributed-handoff delay and handover failure of mobile nodes in SDN-based IEEE 802.11 network. It proposes a hybrid clustering technique using K-mean and genetic algorithm to cluster the network. The clustering aims to reduce the scanning phase of the mobile node by minimizing the number of access points that need to be scanned for the handoff. The results show lower delays and a reduced number of handoff failure compared to the un-clustered network.