A bubble oscillation algorithm for distributed geographic load balancing in mobile networks

This paper investigates a new load balancing scheme for mobile networks that changes cellular coverage according to the geographic traffic distribution in real time. The performance of the whole cellular network is improved by contracting the antenna pattern around a traffic "hot spot" and expanding adjacent cells coverage to fill in the coverage loss. This is essentially a constrained multidimensional optimization problem. A novel bubble oscillation algorithm is proposed to address this problem. Any unserved traffic in the network is absorbed by the geographic load balancing in a similar way that a vacuum between bubbles is filled by bubble oscillations. Simulations have been performed to evaluate the system performance for different traffic scenarios, and the results are presented. Some discussions on this algorithm are also presented. The bubble oscillation algorithm described has the potential of being used in other similar multidimensional resource allocation problems.