Mobile Communications Engineering
Mobile Communications Engineering
Algorithmic aspects of capacity in wireless networks
SIGMETRICS '05 Proceedings of the 2005 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Impact of interference on multi-hop wireless network performance
Wireless Networks - Special issue: Selected papers from ACM MobiCom 2003
Capacity regions for wireless ad hoc networks
IEEE Transactions on Wireless Communications
The capacity of wireless networks
IEEE Transactions on Information Theory
Capacity bounds of deployment concepts for Wireless Mesh Networks
Performance Evaluation
Interference aware resource allocation for hybrid hierarchical wireless networks
Computer Networks: The International Journal of Computer and Telecommunications Networking
Distributed scheduling schemes for wireless mesh networks: A survey
ACM Computing Surveys (CSUR)
| Hi-index | 0.00 |
Local area wireless networks are based on the cell topology: Clients associate to one of several access points, which are connected using a wired backbone. As high data rates are only available close to the access points, a dense infrastructure is needed. This results in high costs, especially for the installation of the wired backbone. A wireless mesh network can be used to reduce the deployment costs by connecting only few access points to the backbone; mesh nodes extend the coverage by forwarding data over wireless hops. Since the wireless medium has to be shared by the nodes, multi-hop traffic requires a high capacity. Hence, mechanisms which increase the system capacity in wireless mesh networks are needed. In this paper, we rate how much capacity can be gained by the introduction of spatial reuse. First, a system model of the wireless network is presented. This model includes a stochastic channel behavior and the signal strength/SINR requirements of a link. Additionally, the possibility of link adaption is incorporated. As the exact calculation of the system capacity using this model is NP-hard, we develop and survey heuristics that reduce the complexity. Then, we apply the developed algorithms to evaluate different spatial reuse strategies. An upper bound is given by a network controlled by a omniscient scheduling entity; a lower bound is provided by refraining from spatial reuse. The results show that under the assumptions of the models at least a capacity increase by a factor of two is feasible; under optimal conditions a 12-fold increase is possible.