Alleviating network load in dense urban multi-access Application-Layer Multicast

Application-Layer Multicast (ALM) is a promising approach to overcome the lack of global IP Multicast availability. Since ALM protocols relieve load from centralized content servers and offer flexible group communication without requiring dedicated support in the network, they can be considered a viable solution for future content distribution. Anyway, using ALM in capacity-constrained environments like dense urban cellular networks may worsen congestions due to higher bandwidth consumption inside the access networks. In this paper, we propose the integration of multi-access communication in ALM in order to unburden access networks from traffic load and avoid congestion. Doing so, we propose a two-step ALM building approach that first establishes a single-homed cellular overlay tree, before efficiently looking up and integrating alternative WiFi communication possibilities in the dissemination tree. Finding WiFi peers in range is not trivial in dense urban environments, since the number of concurrently visible public WiFi domains may be high and Ad-hoc scanning should be avoided if possible due to poorer properties concerning energy, stability, and range. Our proposal follows a space-efficient Bloom Filter-based technique, allowing to precisely join public Wifi domains with high chance of meeting foreign peers without having to try out several candidate domains first.