24-06-2013, 04:54 PM
An Application-Oriented Routing Protocol for Multi-hop Cognitive Radio Networks
An Application-Oriented.docx (Size: 474.28 KB / Downloads: 15)
Abstract
A challenge for routing in cognitive radio networks is the intermittent connection due to the occupying and releasing licensed channels of primary users. To solve this problem, the routing protocol will choose the most stable channel (the channel with the longest average length of idle period) to build up the stable path. This approach, however, can lead to the increase of channel competition between cognitive users since the most stable channels are preferred to be utilized by all cognitive users. In this paper, an application-oriented routing protocol is proposed. The main motivation is to reduce the channel competition between cognitive users by finding the appropriate path according to the application of cognitive users. Simulation results show that our proposed routing protocol reduces the loss ratio and increases the throughput significantly.
INTRODUCTION
In wireless communication, the most valuable resource is the radio spectrum. However, according to recent studies[1][2], while the assigned spectrum (licensed spectrum) is under-utilized in various geographical locations and time, the un-licensed spectrum is always overloaded because of the growth of wireless services. To reduce the waste of licensed spectrum usage and provide more available spectrum resource for un-licensed users, an efficient way is allowing unlicensed users to opportunistically access the licensed spectrum without pe-nalizing Quality of Service (QoS) of licensed users. Cognitive Radio (CR) [3] is envisaged as the sufficient technology that aims to flexibly use wireless radio spectrum.
RELATED WORK
Some stability routing techniques in cognitive radio network already exist. In [8], the authors first define the maximum link hold-time as the maximum length of available period before a cognitive link is considered to be failed. After that, the authors expressed the link hold-time as a function of the maximum link hold-time and the primary user usage pattern. The route hold-time then will be derived as the accumulated hold-time of links which compose the route. Finally, the routing metric based on the route hold-time and end-to-end throughput is proposed. Sharma et al. propose in [7] a new routing metric for multihop cognitive networks. This new metric includes the stability factor, also known as the average channel availability time which is calculated as weighted moving average of previously measured availability times and the time measured in the current measurement duration. Same remarks apply on [9][10], where the proposed routing metric is based on the channel utilization which depends on the average duration of available period of the channel. [6] proposes a collaborative strategy for route and spectrum selection in cognitive radio networks.
Application-Oriented Routing
In this section, we introduce our protocol using on-demand routing approach. When an application of CR node demands to transmit data, CR source node will initiate route discovery by broadcasting route request packet on CCC. After that, each intermediate node along the routes to destination will add the AOS information of channels into the route request packet in sequence. Whenever route request packets reach the destination node, destination node will use AOS information of channels to calculate the metric of each path and choose the path with minimum metric as the route of source node to the destination node.
SIMULATION RESULTS
We study the proposed metric Application-Oriented Sta-bility (AOS) through a simulation using ns2. The ON-OFF model and the AOS metric will be implanted in the MAC layer and Routing layer respectively. A MAC (named AOS-MAC) is modified from MAC IEEE 802.11 supported by ns2. AOS-MAC is implemented to allow secondary users (SUs) to transfer data on the channel transparently to primary users (PUs). In order to do this, SUs only send traffic in OFF periods and do not access the licensed channel in ON periods when PUs are present.
Two test cases are designed to validate the proposed ON-OFF model and the AOS metric. Throughout the simula-tions, we consider all topologies created in a flat area of 1000m 1000m. In each test case, a number of intermediate nodes are placed between source node and destination node. Each intermediate nodes has its own accessible channel list. In addition, two connections TCP/FTP and UDP/VoIP are kept to maintain data transfer during the running time of the scenario.
CONCLUSION AND FUTURE WORK
In this paper, we have proposed an application-oriented routing protocol in cognitive radio network. We studied the transmission behaviour of VoIP and FTP applications to figure out the appropriate channel for each type of application. Based on this study, the application-oriented stability metric and application-oriented routing protocol are proposed to find an appropriate path according to the application running in cognitive nodes. Through simulations, the application-oriented routing protocol was shown to outperform S-AODV. In the future, we envisage to investigate more applications’ transmis-sionbehaviour to provide more efficient channel allocation and route composition schemes.