THESIS
2007
xv, 75 leaves : ill. ; 30 cm
Abstract
Both hidden terminals, intrinsic in a multi-hop wireless network, and unfairness, due to location-dependent contention and binary exponential backoff, have significant impact on the performance of the network from MAC layer perspective, and their destructive effect is intensified and worsened by the upper layer protocols, like routing protocol and TCP congestion control. These two problems, existing in MAC layer, are open problems that have been haunting in the research community for more than a decade. How to solve the hidden terminals and unfairness efficiently is a very important issue in ad hoc networks and WLAN....[
Read more ]
Both hidden terminals, intrinsic in a multi-hop wireless network, and unfairness, due to location-dependent contention and binary exponential backoff, have significant impact on the performance of the network from MAC layer perspective, and their destructive effect is intensified and worsened by the upper layer protocols, like routing protocol and TCP congestion control. These two problems, existing in MAC layer, are open problems that have been haunting in the research community for more than a decade. How to solve the hidden terminals and unfairness efficiently is a very important issue in ad hoc networks and WLAN.
Prior research works on the hidden terminal problem in ad hoc networks are based on the assumption that the SNR requirement is rate independent. But for a given digital modulation scheme, it is E
b/N
o (where E
b is the energy per bit, and N
o is the noise power spectral density) that should be kept relatively a constant, and the SNR requirement must change with rates. We show that because of this assumption, all the derived results and conclusions will be put in doubt. We correct this problem and present a new analytical model for the hidden terminal problem. Based on the insights provided by the new analysis, we propose a simple, yet efficient, scheme to eliminate hidden terminals in ad hoc networks. The new method requires no protocol changes from the 802.11 standard. The NS2-based simulations proof our analysis and the effectiveness of the scheme proposed.
In Chapter 3, unfairness is addressed from transmission power control perspective. Differ from previous research to leverage the potential of power control to increase a network's traffic carrying capacity or reduce the energy consumption, we show that power control can also significantly improve the fairness performance at MAC layer with help from receiver side. A new receiver-assisted power control MAC protocol (RAPC) is presented to exploit this gain. The NS-2 based simulations and performance comparisons show that our approach significantly improves the fairness performance while maintain comparable throughput performance as that in original 802.11 MAC.
Post a Comment