Research On Cross-Layer Key Technologies In Cognitive Wireless Network

Based on software defined radio (SDR) and cognitive radio (CR), cognitive wireless network (CWN) evolves into one of the frontier research fields of communication technologies. CR allows secondary users to opportunistically use the spectrum holes and can greatly enhance the utilization efficiency of scarce spectrum. CWN extends the concept of CR from network perspective, and enables intelligent cognition of the heterogeneous environment. CWN can monitor the status of surroundings, distinguish the changes, and obtain cognitive information from the environment. Afterwards, it conducts a comprehensive and objective analysis of cognitive information for unified planning and decision-making, and then responds to the changes. Under the complex environment with a variety of wireless access technologies overlapped, sharing and competing for wireless resources, CWN aims to achieve end-to-end optimization goals. However, the end-to-end goals need to obtain information from multiple layers in the protocol stack for analysis, learning and decision-making. Thus the information exchange is required among nonadjacent layers and the interface between adjacent layers needs to be expanded. Therefore, cross-layer technology is indispensable and necessary for achieving end-to-end goals of CWN.Cross-layer technology is a creative design of protocol stack. It violates the design concept of traditional layered architecture, permitting direct communication among nonadjacent layers and information sharing among multiple layers. Simultaneously, it actively explores the dependence and relationship among layers to achieve system performance gain and high quality transmission.This dissertation pays attention to the cross-layer key technologies in CWN, and mainly focuses on current blind spots of cross-layer designs. From the perspective of involving different protocol layers, the dissertation carries out in-depth studies on cross-layer architecture, lifetime and effective capacity issues, and jointly configures and adjusts multiple layers to optimize network performance. The contribution of the dissertation is summarized as follows:1) From the perspective of physical layer to application layer, the dissertation focuses on the compatibility issue of a variety of cross-layer designs. Based on the impedance matching idea, a novel cross-layer architecture is proposed. The architecture abstracts the scenarios of time-varying and error-prone surroundings as source impedances with specific characterization vectors, and regards the diverse cross-layer strategies as specific load impedances that match in accordance with the source impedances. Thus a cross-layer optimization refers to an impedance matching process based on certain criteria. Through high-degree abstraction, the architecture blurs the differences of various cross-layer schemes to some extent. Then a large number of cross-layer designs can be effectively implemented on a unified platform. Moreover, a matching algorithm with low complexity is proposed, in order to optimally or suboptimally match the dynamic changing environment with cross-layer schemes, and achieve performance gain.2) From the perspective of physical layer and network layer, the dissertation concentrates on the lifetime optimization issue of typical CWN, cognitive radio sensor network. To prolong the lifetime of cognitive radio sensor network, the dissertation proposes a creative cross-layer scheme in terms of energy efficiency and fairness based on a new factor, residual energy ratio (RER), from the view of energy efficiency of data transmissions. Firstly, based on power control in the physical layer, the preliminary candidate nodes for potential transmission are decided. Subsequently, jointly considering the access probability of each node, the per-hop transmission success probability, and energy efficiency of end-to-end transmission links, equilibrium routing of network layer is accomplished. Thus the end-to-end communication can be achieved in cognitive radio sensor network.3) From the perspective of physical layer and data link layer, the dissertation investigates the effective capacity based on statistical delay QoS constraint in underlay CWN. Under a novel cooperative communication scenario of secondary network, a creative cross-layer strategy for calculating effective capacity is illustrated in detail. The novel cooperative communication scenario takes both the direct link and relay link of secondary user into consideration for data transmission. Then the receiver of secondary user merges the signals from the two links to obtain diversity gain and improve system performance. The cross-layer strategy jointly optimizes power control at physical layer and effective capacity at data link layer. Under the statistical delay QoS constraint of secondary user, the dissertation obtains the maximum arrival-rate supported by the secondary user’s cooperative links, and derives the closed-form expression of effective capacity at last.