Study On Topology Management In CR-MANETs

Mobile Ad hoc Wireless Networks (MANET) is a wireless communication system with non-fixed infrastructures, it has lots of advantages such as flexible deployment and rapid mobility and so on. It is widely used in military communication, emergency communication and other special areas. As an intelligent spectrum sharing technology, cognitive radio relies on the support of artificial intelligence to sense the wireless communication environment. Based on certain learning and decision-making algorithms, the working parameters of system are changed adaptively in real-time, and then the idle spectrum is tested dynamically and can be used effectively. Theoretically, cognitive radio technology allows the reusing and sharing of spectrum in time domain, frequency domain and space domain; it is considered as an effective way to alleviate the contradiction between the current increasing spectrum demand and the limited spectrum resources. By integrating cognitive radio into the MANET, through monitoring on the operating condition of the main user and adaptive resources adjustment, the sharing of spectrum resources in the network can be achieved to meet the users'QoS requirementsThe topology management is one of the most important parts in the research and application of cognitive MANET, aiming at optimizing the network topologies by cutting unnecessary links or redundant links. There are two main research directions in topology management: power control and clustering.This paper focuses on topology management on clustering in cognitive MANET. The main contents include:1) Considering network management requirements, the paper analyzes the status and importance of topology management in MANET, focusing on the analysis of clustering algorithms and the method to establish a virtual backbone in ad hoc network.2) By integrating the cognitive radio into the MANET, the network configuration and application model of cognitive MANET are established to elaborate the network architecture of the cognitive MANET, and to elaborate the identification and positioning of main users in cognitive MANET, focusing on analyzing the main causes affecting the performance of the cognitive MANET.3) Subject to cognitive MANET, by spectrum sensing and cognitive node positioning, a virtual backbone construction method is proposed, which is based on the space and time information of the primary user's working status. Virtual backbone construction includes cluster head election and gateway selection. Cluster heads are taken as anchor points, the connected dominating sets are obtained by selecting gateway nodes that are less affected by the primary user, and then the virtual backbone can be established. A MANET network clustering algorithm based on link availability is proposed by improving the classic Lin-Gerla clustering algorithm. Subject to the problem that the excessive clusters concentration might occur, a rapid cluster splitting method is presented to be applied in the phase of cluster maintenance. Based on the weight comparison, through the sensing of the location of the primary user, the spatial distribution area of the cognitive nodes are divided and assigned different weights, thus realizing partition processing in the network, maintaining relative stability of the sub-clusters in motion and reducing the probability of interference links.