Multi-hop WiFi Direct Implementation via Efficient Communication Backbone Construction.

The ever increasing availability and popularity of mobile devices has led to new designs to provide better communication in different environments. Accordingly, the WiFi Direct standard has been introduced to address ad hoc communication demands, as needed in cyber-physical systems, providing long range and fast communication at the quality of service level of traditional WiFi. WiFi Direct is a standard for single group networks, but due to its promising features there have been studies to utilize this standard in multi-group networks. However, previous multi-group WiFi Direct efforts are either theoretical or implementations in small networks. In this work, we implement and evaluate communication backbone forming algorithms using WiFi Direct networks that will take a step towards dynamic multi-group WiFi Direct network implementation in real devices at large scale. Specifically, we consider our network as a graph and construct the communication backbone using graph theory principles, and we implement these backbone algorithms using MATLAB and the mininet-wifi emulator, taking into account the constraints of real WiFi Direct networks. Using both centralized and distributed algorithms, we evaluate the effects of network size, graph density, and randomization of the algorithms on the communication backbone.