| Cognitive Wireless Networks (CWNs) are the networks with cognitive process, which analyze currents status of network according to the channel feature in wireless environment, topology of wireless networks and service characteristic of wireless users. CWNs make plan, decision and operate in line with such status. In the meantime, the networks are able to learn in self-adaption process and can be used to future decision for cognitive circulation in order to realize end-to-end optimization. Cognition technologies are a series of the most basic and important technologies in the field of CWN, among which the detection on spectrum holes is a vital part that serves as the premise to realize dynamic spectrum sharing technologies. The dissertation studies the interactive instructing function of knowledge to cognition in cognitive circulation. That is to make decision on parameter configuration by learning historical cognitive information in order to promote efficiency of successive cognition, relying on which several intellectual optimized algorithm based on cooperative spectrum sensing is to be proposed. The main items are listed as follows:First, the dissertation puts forwards a solution from the aspect of system level to wideband spectrum sensing problems by integrated sensing capacity of the set of all cognitive terminals, namely coordinated spectrum sensing. The optimization of coordinated spectrum sensing concentrates on the assignment of sensing task based on discrepancy of sensing capacity of different cognitive terminals to different channels. A combination of cooperative and coordinated spectrum sensing is available at the time that the number of cognitive terminals is more than that of channels. The overall sensing accuracy of the system can be optimized by assigning several terminals to sense one channel with cooperation. Chapter 3 analyzes the problem and proposes iterative Hungarian based and greedy based algorithms to solve this problem.Second, in Chapter 4 the dissertation proposes a coordinated spectrum sensing solution with QoS-oriented optimization of cognitive users, focusing on the maximization of end-to-end efficiency in CWN. Chapter 4 at first discusses QoS-oriented optimization in local spectrum sensing scenario by defining a "quality of transmission" factor as the evaluation of the QoS gain of the cognitive user from data transmission in each frame, deduces the relationship of the "quality of transmission" factor and the observation time, then proposes the algorithm optimizing the factor by choosing the best observation time. Afterward, the QoS-oriented optimization in coordinated spectrum sensing scenario is discussed. With the analysis of different status of terminals and channels in CWN, a new sensing sequence proposal that sensing of unoccupied channel with idle terminals puts to communication slot is put forward. The proposal makes it possible that all cognitive terminals perform coordinated spectrum sensing only on a small amount of channels occupied by cognitive users in the sensing slots, leading to the increment of the number of cooperative sensing terminals. The proposed solution can get more reliable sensing results in relatively short period of time and promote QoS of cognitive users.Third, this dissertation studies the localization of authorized system transmitter, which is a key problem of cognitive technology, proposing a localization algorithm by virtue of traditional,1-bit-report cooperative spectrum sensing routine in Chapter 5. The study fully considers the fact that the signal transmitting power of authorized systems cannot be acknowledged to CWN. So the proposed algorithm creatively regard the signal transmitting power of the authorized systems as another parameter to be estimated, then eliminated by the transformation of the equations. At last, a promotive algorithm with sensing data selection and weighting is discussed, which could further improve the performance of localization. |
PDF Full Text Request