Research On Backbone Assisted Efficient Transmission Technology In Cognitive Networks

Cognitive network has provided an effective solution to the problem of increasingly scarce spectrum resources. So far, researches about cognitive networks focus on dynamic spectrum access technologies, including dynamic spectrum sensing, allocation and sharing.Regarding to traditional multi-hop wireless networks, cognitive multi-hop networks are much more complex. Due to the "dynamic spectrum" characteristic of cognitive networks, it’s hard to achieve multi-hop transmission. Secondary users along a path may reside in the inteference ranges of different primary users, they may have a different set of available channels due to the interference imposed by different primary users. If two adjacent secondary users don’t have an overlapping set of available channels, they could not communicate directly via cognitive spectrum, and the network may be divided into mutiple partitions that are not interconnected.In traditional cognitive network technology, secondary users must use cognitive spectrum for data transmission, and they may invest in a common control channel to exchange control messages. When the cognitive network is partitioned, data transmission across different partitions can only be carried out with the assistance of the common channel. However, the common channel is a precious resource in cognitive network that transmitting data via the common channel is quite expensive. Therefore, the secondary usersare not allowed to access the common channel for abitary data transmission. To improve the efficiency of cross-partition data transmission via the common channel, virtual backbones are selected at the edge of each partition which are granted access rights to the common channel for data transmission across neighboring partitions. This paper focuses upon the algorithms to select virtual backbones for efficient cross-partition data transmission.Since the respossibility of data transmission across partitions is put upon the virtual backbones, the rules to select virtual backbones will directly affect the transmission efficiency of the common channel, connectivity of partitions and stability of the network. For this purpose, chapter three proposes a stability based algorithm, and a network coding based algorithm is presented in chapter four.Since the cognitive spectrum is dynamic, the network partitions may vary according to the change of cognitive spectrum, thus the virtual backbones may have to be re-selected in order to keep the partitions interconnected. To reduce the cost of maintaining virtual backbones, chapter three proposes a stabiltiy based algorithm to select virtual backbones. The algorithm consists of four steps, including partition cognition, selection of candidate backbones, selection of backbones andconnection of partitions. Since the selected virtual backbones are more stable during the change of partitions, the maintenance cost can be reduced, and the data transmission between adjacent partitions is more stable. The simulation results show that the algorithm is able to obtain a set of virtual backbones with good stability.In the fourth chapter, in order to improve transmission efficiency between areas, a backbone selection algorithm based on network coding is proposed. Network coding is adopted to improve the efficiency of cross-partition data transmission. An utility function is designed, and a mathematical optimization model is established. Since the proposed model is a nonlinear integer programming problem, an improved binary particle swarm optimization algorithm is presented to solve it. Simulation results show that the algorithm is able to obtain a set of virtual backbones that maximizes the utility function.