Research On Cross Layer Multicast Algorithm In Ad Hoc Networks Based On Cognitive Radio

The increasing use of collaboration applications and wireless devices may further add to the need and usage of mobile Ad Hoc networks(MANETs). MANET is widely used in various areas such as wireless sensor networks, military battlefields, wireless personal area networks and disaster reliefs for its special features, for example, self-organized, highly independent and distributed. However, restraint in power and link bandwidth as well as dynamic topology lead to different communication requirements. Compared to unicast or broadcast, multicast technology is an essential way to distribute packets and effectively use the limited bandwidth.Cognitive Radio technology, as an effective solution to scarce spectrum resources, was proposed to enable unlicensed users to sense and intelligently access the unoccupied spectrum. In Cognitive Radio networks, a CU(cognitive user) firstly senses and gets available spectrum named Spectrum Opportunity (SOP), then divides the SOPs into multiple channels and selects one channel which is out of interference to PUs(primary user) to transmit data. Pyo et.al introduced Common Link Control Radio(CLCR) as a common active wireless system for cognitive terminals and developed a Minimum Weight Routing Protocol, which selects a minimum communication overhead route.The spectrum-aware routing protocol(SPEAR) integrates the end-to-end optimization of flow-based approaches with the flexibility of link-based approaches. Even through many Cognitive Radio researches have been conducted, they mainly focus on spectrum sensing/detecting, media access control(MAC),or network layer unicast routing protocols design, multicast communication using Cognitive Radio is still an unexplored area. We proposed a novel multicast protocol for Ad Hoc Networks based on Cognitive Radio, in which the physical layer gets available spectrum information by Cognitive Radio technology.A lot of multicast routing protocols have been proposed for Ad Hoc networks over the last few years. Based on the way that data packets are disseminated by, existing Ad Hoc multicast routing protocols can be classified into Tree-based, Mesh-based and hybrid structure. Tree-based multicast protocols establish and maintain a shared or a source tree for data transmitting, some examples are as PUMA(Protocol for Unified Multicasting through Announcement), MAODV(Multicast Ad hoc On-demand Distance Vector), which is a multicast modification to AODV, AMRIS(Ad hoc Multicast Routing Protocol utilizing Increasing id-numberS).Tree-based multicast protocols attain lower delay but suffer from a larger packet loss ratio. PUMA attracts much attention for its better PDR and control overhead, and its uniqueness originates from use of a single control packet MA(multicast announcement) to accomplish all the functions needed in the creation and maintenance of a multicast tree. Mesh-based multicast routing protocols like ODMRP(On Demand Multicast Routing Protocol),CAMP(Core-Assisted Mesh Protocol), which try to establish and maintain a shared mesh to forward data packets. Mesh-based protocols achieve better robustness while have bad efficiency performance. Hybrid structure routing protocols tend to establish mesh before a shared multicasting tree, as a combination of the above two mechanisms. AMRoute is a typical hybrid structure multicast protocol in Ad Hoc networks. Hybrid structure routing protocols bring more overhead although overcome shortcomings of the other two styles.In CR-PUMA protocol, the network layer implements multicast communication based on PUMA protocol.We proposed a novel multicast protocol based on Cognitive Radio and a cross-layer approach, which is named CR-PUMA. The physical layer obtains available spectrum information by cognitive radio technology, then the available probability of each channel is calculated in the network layer, which is used to decide the channel’s priority to be scheduled in the MAC layer. Meanwhile, PUMA algorithm is running in the network layer to accomplish the whole multicast process. In simulation experiments under NS2, our algorithm, which is based on cognitive radio and cross-layer approach, demonstrates a superior packet delivery ratio and throughput performance. During multicast communication in MANETs,hidden terminal problem may occur because we assume nodes equipped with single transceiver. To solve this problem, we introduce the JOIN/LEAVE mechanism. When nodes want to switch to another channel, they need broadcasting a LEAVE message to all its neighbors to indicate the leaving,and in the same way broadcasting a JOIN message when they join a new channel...