Research On Key Technologies Of Virtual Access Network For Mobile Users

Due to high bit-rate and low cost,IEEE 802.11-based WiFi networks have been widely deployed as an important way to access the Internet in recent years.In practical application scenarios,some users or devices require guaranteed access bandwidths when they are moving.However,a competitive MAC protocol is used in traditional WiFi networks,which cannot guarantee the quality of service(QoS).Moreover,the mobility of the users make the problem more challenging.In order to provide guaranteed access bandwidth to mobile users,virtual access network(VAN)is proposed in this thesis based on the idea of network virtualization,and its key technologies are studied.By establishing isolated VANs for every mobile users,the users obtain guaranteed access bandwidths,so that their customized requirements on QoS are satisfied.The specific research work in this thesis includes:(1)A VAN construction framework is designed for downlink traffic dominated WiFi networks.Based on software-defined wireless network architecture and wireless parameter configuration,multiple physical access points(APs)are abstracted into a virtual AP.The virtual AP provides bandwidth-guaranteed service to a mobile user.The VANs are separated from each other in the dimension of time,which makes sure that any VAN does not have mutual interference.(2)In order to improve the network resource utilization under the premise that bandwidth requirements of primary users are guaranteed,two scheduling algorithms for VAN are designed.In the first algorithm,the maximum independent set in the VANs are selected to assign time-slot,which has an exponential computing complexity.The second algorithm adopts greedy strategy,and selects the VAN with the longest time-slot,which has a linear computing complexity.The experimental results indicate that both algorithms can improve the network utilization,but the second one has a much higher efficiency.Hence,the second algorithm is more suitable for practical scenario.(3)In order to reduce resource waste caused by reservation for mobile users,a network resource optimization scheme is designed.Thanks to the regularity of human activities,the moving trajectory of mobile users can be predicted by a Markov model.According to the prediction,the network resource reservation is optimized.Experimental results indicate that the proposed scheme can reduce network resource waste and operation cost in practical applications.In this thesis,the VAN construction method,the scheduling algorithms,and the resource optimization scheme are proposed.These research work can support guaranteed access bandwidths in WiFi networks,which can promote further development and application for WiFi networks.