Research On Routing And Transport Protocol For Performance Optimization In Multi-Hop Cognitive Radio Networks

The Cognitive Radio (CR) has introduced the idea to exploit spectrum holes.(i.e., bands). Powered by CR, the newly emerged cognitive radio networks (CRNs) are proposed. Secondary user (SU) in CRNs, accesses the temporarily idle channels in a secondary and opportunistic fashion while preventing harmful interference to primary user (PU). CRN is a prominent solution to improve the efficiency of spectrum utilization. In recent years, CRNs have attracted the attention of many researchers in the wireless networks domain. Routing and transmission reliability are key technologies in CRNs, which are hot research directions for multi-hop CRNs in recent years. How to transmit data efficiently in multi-hop CRNs is a fundamental research problem. Facing this challenge, this thesis considers data transmission of CRNs from two aspects, routing and transport protocol, and proposes different strategies and algorithms. The main contents of this thesis can be summarized as follows.First, this thesis proposes a strategy for routing optimization for network capacity analysis, which is based on small world model. Capacity analysis is very useful in investigating the ultimate performance limits for wireless networks. However, due to spectrum sharing in CRNs, the network capacity analysis of Secondary Network (SRN) is much more difficult to analyze than that of traditional wireless networks. Although few works about capacity analysis from a routing perspective, they did not improve the network capacity through reducing average path length. Motivated by these observations, this thesis introduces a novel strategy based on small world model, whose average path length is shorter, for routing optimization for network capacity analysis in multi-hop CRNs. The spectrum opportunity of licensed channels is modeled by preemptive resume priority queue model in this strategy and the connection ratio of SUs can be obtained. And then, a novel scheme to create shortcuts based on connectivity ratio and a new channel assignment algorithm, which jointly considers the available time and transmission time of the channel, are proposed in multi-hop CRNs. Moreover, the network capacity of multi-radio multi-channel SRN based small world model is derived. The effectiveness of the proposed strategy is evaluated through the simulations.Second, considering channel assignment, a novel opportunistic routing (OR) based on intra-session network coding, is proposed. Due to the intermittency of the spectrum in space and time domain, the available spectra of SU are dynamically change and heterogeneous. Opportunistic routing has to face new challenges in multi-hop CRNs, e.g., the impact of PU behavior and spectrum sensing on forwarding candidates selection. Although a few opportunistic routing protocols are proposed for CRNs, the routing metric of these OR is mainly based on spectrum information and Expected Transmission Count (ETX), ignoring the buffer of the node and the relationship between these nodes in a candidate set. To overcome this difficulty, this thesis proposes a novel routing metric, which is based on buffer of the node, ETX and spectrum availability, and gives an optimization model for joint channel assignment and OR based on intra-session network coding. In order to reduce the computational complexity of the optimization problem, a heuristic algorithm to select forwarding candidates and assign channels is presented for multi-hop CRNs, including candidate selection algorithm considering the queue state of a node and ETX, and channel assignment algorithm taking into account the transmission time and the available time of a given channel. The simulation results show that the proposed scheme performs significantly better than traditional routing and classical opportunistic routing in which channel assignment strategy is employed.Third, in inter-session network coding, the packets from different sessions are coded together, which can significantly improve throughput by means of increasing information in a transmission. Integrating inter-session network coding and OR is a promising solution to improve throughput. But existing OR protocols mainly focus on traditional wireless networks, ignoring the features of CRNs. Moreover, there are very few works about OR in CRNs. Motivated by these observations, a novel coding aware OR, cognitive radio opportunistic routing (CROR) is proposed. It exploits a novel routing metric, successful delivery ratio (SuDR), which is based on packet loss rate and successful spectrum utilization. And then, a novel candidate select algorithm and coding algorithm are proposed for multi-channel CRNs. Simulation results demonstrate that the proposed routing CROR achieves better performance in increasing the throughput and bandwidth utilization efficiency and reducing the probability of PU-SU packet collision compared with traditional opportunistic routing.Finally, TCP is a key technology to guarantee reliable data transmission. Due to the uncertainty of the channel availability, TCP performance may significantly degrade in CRNs. To address the challenges, some transport protocols have been proposed for reliable transmission in CRNs. However, they mainly exploited a single metric for TCP performance evaluation and did not comprehensively consider the features of CRNs, failing to dealing with loss type. Aiming at solving aforementioned problems, a novel transmission control protocol called TCPJGNC (TCP Joint Generation Network Coding, JGNC) based on network coding is proposed. And considering the delay metric, an improved strategy, TCPNC-DGSA (TCP Network Coding Dynamic Generation Size Adjustment) is presented. In TCPJGNC, the coding scheme is adaptive, which is set according to the changing ratio of ACK and the rank of the matrix consisting of received packets’coding coefficient. TCPJGNC can significantly reduce the retransmissions and provide a higher decoding probability, and then enhance the TCP performance in multi-hop CRNs. In order to apply TCP to multi-hop CRNs efficiently, the TCP mechanism is modified, by considering slow start, spectrum sensing state, spectrum changing state and presence of PUs. The effectiveness of the proposed strategy is verified through the simulations. In order to improve delay of TCPJGNC in multi-hop CRNs, a novel transmission control protocol called TCPNC-DGSA is proposed. In TCPNC-DGSA, the network coding condition is set according to the Generation Round Trip Time (GRTT). The simulation results indicate that TCPNC-DGSA can significantly improve end-to-end delay.