Multiobjective Network Opportunistic Access for Group Mobility in Mobile Internet

The integration of existing and emerging heterogeneous wireless networks in mobile Internet is a combination of diverse but complementary wireless access technologies. Satisfying a set of imperative constrains and optimization objectives, access network selection (ANS) for mobile node (MN) is an inherent procedure in mobility management that needs to be solved in a reasonable manner for the whole system to operate in an optimal fashion. However, ANS remains a significant challenge. Because many MNs with distinctive call characteristics are likely to have correlated mobility and may need to perform mobility management at the same time, this paper, with the goal of investigating group mobility solutions, proposes a network opportunistic access for group mobility (NOA-GM) scheme. By analyzing the directional patterns of moving MNs and introducing the idea of opportunistic access, this scheme first identifies underloaded access networks as candidates. Then, the candidates are evaluated using normalized models of objective and subjective metrics. On this basis, the ANS problem for group mobility can be conducted as a multiobjective combination optimization and then transferred to a signal-objective model by considering the optimization of the performance of the whole system as a global goal while still achieving each MN's performance request. Using an improved genetic algorithm with newly designed evolutionary operators to solve the signal-objective model, an optimal result option for ANS for group mobility is achieved. Simulations conducted on the NS-2 platform show that NOA-GM outperforms the compared schemes in several critical performance metrics.