Wireless Mesh Network is a special network infrastructure which uses only wireless connections. The network is wireless multi-hop, meaning that the connections possibly include multiple intermediate stations. The lack of a wired backbone promises more flexibility, compared to classic infrastructure networks.
When wireless mesh networks are used in dynamic propagation environments (e.g. manufacturing, logistics), the changes in the environment (e.g. new obstacles) can disturb the wireless communication. This affects both the radio coverage and the connectivity of the network. The radio coverage ensures that the mobile stations can connect to the network while they are within a service area. The connectivity ensures that the network topology is connected.
This dissertation contributes a fault-tolerance method for guaranteeing the availability of radio coverage and connectivity of wireless mesh networks in dynamic propagation environments. The services in normal state have a redundancy, tolerating the fault environmental dynamics. The occurrence of faults lead to error state of the services. In this state the service is still correct, because of the initial redundancy, but the redundancy is lost. Our method avoids the failures by detecting the error states and performing system recovery before an error leads to failure. The system recovery restores the original redundancy of the services.
We have developed new methods for error detection and system recovery which are required for radio coverage and connectivity of wireless mesh networks. The error detection and system recovery are especially challenging in dynamic propagation environments. For this purpose we have developed a new method for automatic radio model calibration. This method uses measurements from the network to adapt a radio propagation model to the real environment. The measurements are obtained in an automatic way from the infrastructure and from a new localization service, developed specifically for this purpose. Based on the calibrated model our error detection method detects the dynamics in the propagation environment. Based on the model and a new automatic base station planning algorithm, our system recovery method restores the normal state of the services.
The evaluation in different office and industrial environments has shown that the error detection method successfully detects the errors and the system recovery method successfully restores the normal state of the service. This guarantees the availability of radio coverage and connectivity of wireless mesh networks in dynamic propagation environments.
Wireless Network, Wireless Mesh Network, Fault-tolerance, Radio Coverage, Connectivity, Dynamic Propagation Environment, Radio M