Proximity-based peer selection for service lookup in areas of sudden dense population

Providing location-based services to mobile users in a cellular network has long been studied and many services already exist in the market. However, a sudden density increase in a certain geographical area can cause a remarkable quality of service degradation and an increase in connections failure. In a previously published work by us, we proposed a hybrid system that allows user, upon density increase, to still obtain their location-based data without any disruption. This is accomplished by forming, from the existing nodes, upon the occurrence of sudden density, built in logical entities to cache service data. Users' queries, then, are directed to those logical entities, called autonomous systems, for replies. In this paper, we propose a proximity-based peer selection for the formation of autonomous systems. We demonstrate that when the formation and assignment of autonomous systems are made highly sensitive to peers location, the overall query latency is improved. Simulation and experimental results demonstrate that this approach decreased the latency associated with service lookups as well as increasing the ability to accommodate more mobile users at times of disrupted services. (7 pages)