Throughput-Maximum Resource Provision in the OFDMA-Based Wireless Virtual Network

In this paper, we propose a throughput-maximum resource provision scheme in the OFDMA-based Wireless Virtual Network (WVN). The proposed scheme takes the dynamics of both traffic arrivals and wireless channels into consideration. Furthermore, we consider a more flexible service contract where average resource provision is guaranteed for each slice. As it is practically impossible to know future traffic arrival and channel information, we employ the Lyapunov Optimization to develop an online resource provision algorithm. The proposed online algorithm performs joint subcarrier and power allocation to each slice based on the traffic information, wireless channel state as well as historical resource provision. Theoretic analysis implies that the proposed algorithm can arbitrarily get close to the theoretic optimal throughput with degradations in delay, and the two performances can be balanced on demand. Extensive simulation results show that our proposed scheme can significantly improve the average throughput, while reducing the average delay and achieving an excellent isolation performance.