On the Design of Robustness-Oriented Load Balanced Routing and Interference-Minimized Channel Assignment with Utilization Consideration in Multi-hop Cognitive Radio Networks

• Main Authors: Shih-PingLiang, 梁世平
• Other Authors: Ching-Fang Hsu
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/94574136121469076953

碩士 === 國立成功大學 === 資訊工程學系 === 103 === In recent years, cognitive radio technique significantly improves the spectrum utilization efficiency. However, due to the privilege of primary users (PUs,) the network performance of secondary users (SUs) would shrink for returning the occupied spectrum whenever the PU becomes active. Therefore, channel assignment strategies, spectrum handoff methods and routing should be considered thoroughly to avoid the shrinkage of network performance. Besides, the feature of co-channel interference also influences the channel assignment for each node, which is solved by choosing highly robust channel in the existing approach to reduce the occurrence of interference and the probability of network corruption. However, concentrating network load on the highly robust route or the shortest path would cause network congestion. Based on the issues described above, two-stage robust and minimum interference topology algorithm (TSRMITA) is proposed in the thesis to minimize the co-channel interference between PUs and SUs and to improve the stability of data transmission in multi-hop cognitive radio networks (CRNs). Moreover, we also propose robustness-oriented and load balanced routing algorithm (RLBRA) which distributes maximum load of the network to provide more stable and low total delay route for data transmission. From simulation results, we can see that TSRMITA outperforms several existing scheme and improves network performance. The average throughput of TSRMITA is superior to RCA and CRTCA by 10-35.2% and 15.7-23.3% improvement, respectively. Additionally, we compare the performance among the proposed routing RLBRA, robustness-oriented routing and shortest path routing. The simulation results show that under the situation of high traffic load, our proposed routing RLBRA has better performance in end-to-end delay than SFA and DAU routing schemes by 4.4-16.7% and 2.9-13.5% improvement, respectively.