Capacity estimation for growth planning of cellular networks in the presence of temporal and spatial traffic fluctuations

Mobile traffic patterns are subject to rapid fluctuations over time, hence making network growth planning a cumbersome task for cellular-network providers. We introduce a network capacity metric that captures both temporal and spatial traffic fluctuations, and that can be derived from hourly service measurements as used for network growth planning. In a simulation trial, we compare this new metric to traditional methods that derive network capacity from performance evaluations of daily busy-hour traffic or time-averaged cell-load levels. We show that in the presence of temporal and spatial traffic fluctuations, the traditional methods overestimate network capacity and incorrectly project the relative gains achieved through capacity enhancement measures such as cell-hardware upgrade or dynamic load-balancing features.