Channel-width Adaptation And Physical-layer Secrecy In Wireless Networks

As wireless networks rapidly evolve into the next generation, many challenging issues need to be resolved. Among them, throughput and security are the most critical ones. With high throughput, a wireless network saves precious resource of spectrum and delivers quality of service for wireless applications. With sufficient security, a wireless network wins trust of customers. To this end, this thesis targets at novel mechanisms for enhancing throughput of wireless mesh networks and for ensuring security of wireless networks.Throughput research in this thesis is focused on wireless mesh networks, which is an in-dispensable part of the next generation wireless networks. An orthogonal frequency division multiple access (OFDMA) based channel-width adaptation scheme is first proposed to improve the throughput of wireless mesh networks. This scheme is then analyzed under the framework of network optimization. Its computational complexity is proved to be NP-complete. Thus, a greedy algorithm is derived to obtain a suboptimal solution. Based on the greedy algorithm, a distributed medium access control (MAC) protocol is designed to achieve OFDMA-based channel-width adaptation in wireless mesh networks. Performance results show that the new protocol dramatically improves the throughput of wireless mesh networks.Security research in this thesis is focused on physical layer secrecy, as it is the ultimate path to wireless network security. Two topics are studied in this thesis. In the first topic, a fast secret key generation scheme is developed for wireless networks with long coherence time. This scheme combats the shortage of time diversity via an innovative virtual channel approach. Opportunistic beamforming is adopted to increase the key generation rate, and frequency diversity is exploited to ensure that the key secrecy grows with the key size. This new secret key generation scheme can be easily adopted in both narrowband and wideband systems. In the second topic, impact of artificial noise to the secrecy of a large scale wireless network is analyzed based on the notion of secrecy transmission capacity. Analytical results reveal the relationship between the secrecy transmission capacity and system parameters, and also provide insights on how system parameters can be designed to improve secrecy transmission capacity of a wireless network.