Research On Load Balancing Mechanism In Long Term Evolution System

As a critical radio resource management (RRM) functionality in next generation wireless communication system, Load Balancing plays a prominent role in avoiding cell congestions, ensuring overall high occupation of radio resources in the network, consistent user experience and low delays. This paper focuses on the research of inter-cell load balancing mechanism on the X2 interface.There are many weaknesses in current load balancing schemes: user equipment (UE) could only obtain information about its serving cell and is unaware of resource situation of neighbor cells, which may lead to"Ping-pong Effect"during handovers; Also, a single UE has no awareness of handover decisions of its peers in the same cell, which could result in handover competition; the load balancing procedure is initialized by less loaded cells, and information regarding UEs is seldom considered during the process, this may result in ineffective handover requests; the exchange of location information regardless handover happens or not lead to excessive network overhead; and finally, the LB procedure is triggered once load threshold is reached, which may cause too frequent LB process and thus a drop in operation efficiency.A deep investigation of load distribution in cell networks, major signaling and procedures,network set-up and principles for evaluation of LB algorithms was present, and based on these analysis, a novel LB approach featured in two-stage threshold to avoid"Ping-pong effect", additional cell periodic timers for greater convergence speed, narrowed search space and each cell is set to report only on the condition that it was overloaded to reduce unnecessary signaling and network overhead. In the end, performance of this algorithm is evaluated in a self-designed network simulator in a scenario specifically designed with positive results which showed that greater convergence speed and less fluctuation of UE quantity within a cell were obtained.