A Dual-NIC AP Handoff Solution Based On Mobility Prediction In Software-defined Wireless Networks

The software defined wireless network framework is one of the candidate solutions for5 G network mobility management.The characteristics of SDWN framework,i.e.,the control and forwarding separation,greatly makes up for the shortcomings of traditional IP networks in the mobility management of wireless terminals.However,the handoff scheme of the SDWN framework is limited by the link layer handoff mechanism,which will cause the wireless terminal to generate network interruption during the AP handoff.In addition,the single evaluation indicators for APs of existing solutions are too simple,resulting in unsatisfied APs selection,and poor user's experience of real-time application service after handoff.Aiming at the problems of network interruption during handoff and ineffective selection of APs in existing research,this thesis proposes a dual-NIC AP handoff solution based on a technique for order preference by similarity to an ideal solution,named TOPSIS,in the SDWN environment.In the dual-NIC AP handoff solution based on TOPSIS,called DNIC-T,the wireless terminal is equipped with dual-NIC so that it can better adapt to the handoff characteristics of the SDWN,so as to ensure the network connectivity of the wireless terminal during the handoff process.Through simulation experiments,this thesis analyze how the four parameters(i.e.received signal strength indication,AP inherent bandwidth,current load,and current number of connections)affect the user experience of real-time application services,and evaluate the AP handoff based on TOPSIS in terms of these four parameters.However,the DNIC-T solution does not take into account the user's mobility characteristics,therefore,unnecessary handoff will occur to a certain extent,which will reduce the overall network performance.In order to reduce the unnecessary handoff times in the DNIC-T solution,this thesis proposes a dual-NIC AP handoff solution based on mobility prediction.The DNIC-M solution improves the TOPSIS-based AP selection algorithm in the DNIC-T solution,and the mobile prediction strategy based on Markov chain is introduced with the consideration of the user's mobile characteristics.The Markov-chain based mobility prediction strategy will calculate the user's one-step state transition matrix according to the user's historical movement trajectory,firstly.And then,according to the user's current location and the probability of one-step state transition,the probability of appearing at each location at the next moment is calculated through the Markov process.Finally,use the calculated probability of occurrence is used together with the parameters on the network side as the evaluation index for AP selection.The simulation results show that the DNIC-M solution has significantly reduced the number of handoff times compared to the DNIC-T solution,while the transmission delay,throughput and delay jitter after the handoff are slightly reduced,and the packet loss rate is slightly increased.That is,the DNIC-M solution can reduce unnecessary handoff times on the premise of ensuring the user experience of wireless terminals and save network resources.Compared with the dual-NIC AP handoff Solution based on RSSI(DNIC-R),the DNIC-M solution achives better handoff performance and user experience of real-time applications.