Applying network coding on broadcasting service is known to reduce times of transmission in the process of recovering the loss packets. In previous design, coding coefficients are put in the packet headers so that MSs can decode the coded packets. However, it causes extra overhead. Moreover, since many mechanisms depend on feedback information to encode packets, they may cease operating once they are out of feedback support. To address these problems, we propose a codebookbased network coding scheme, Pseudo Random Network Coding (PRNC). The codebook defines the coding coefficients of the transmission packets; therefore, we only need to put the index of the codebook in the packet header to decode packets. The simulation result shows that PRNC decodes more packets and has higher all-perfect decoding ratio compared with other network coding schemes.

Since uplink (UL) and downlink (DL) traffic loads are time-variant in femtocells, it is essential to adapt dynamic-TDD (D-TDD) to effectively adjust the UL and DL transmission resources. However, due to the cross-link interference of D-TDD, the benefits of dynamically adjusting the UL and DL resources are diminished. The currently used interference mitigation scheme, the clustering scheme, achieves interference mitigation but reduces the ability of DL and UL resource adaptation. In this research, we propose "Soft Reconfiguration" to reduce interference while allowing femtocells to dynamically adjust their UL and DL resources. In Soft Reconfiguration, the femtocells which highly interfere with each other will be categorized in the same interference group, but unlike the clustering scheme, the femtocells in the same group are allowed to choose different subframe configurations. In the simulation, we compared Soft Reconfiguration with two schemes, D-TDD without interference mitigation and the clustering scheme. The results show that the soft reconfiguration scheme outperforms the other two schemes in better throughput and effective resource utilization.

Since uplink (UL) and downlink (DL) traffic loads are time-variant in femtocells, it is essential to adopt dynamic time-division duplexing (TDD) to effectively adjust the uplink and downlink transmission resources. However, the cross-link interference between dynamic TDD femtocells decreases the throughput gain of dynamic TDD. In this paper, we propose an evolutionary game-based distributed approach to choose the UL-DL configuration in order to minimize interference and maximize the system throughput in a large-scale femtocell network. A multiple populations evolutionary game is formulated to model femtocells with different traffic loads. We prove that the evolutionarily stable strategy (ESS) of the considered multiple populations evolutionary game is the optimal configuration which maximizes the system throughput. Simulation results confirm the effectiveness of the proposed evolutionary game-based approach for system throughput optimization in femtocells employing dynamic TDD.

No abstract

No abstract