Abstract
Machine Type Communications (MTC) is a new type of communication in 3GPP LTE-A networks. The number of MTC devices will be significantly larger than human to human devices in the coming decade due to the proliferation of Internet of Things applications. It may cause the signalling network overload problem in an LTE-A network when a large number of MTC devices try to access network within a short period of time. In this paper, we propose a prediction-based scheme with access barring to solve this problem in LTE-A networks. In the proposed scheme, the Mobility Management Entity (MME) of the core network records the scheduled transmission time of the MTC devices which connect to network periodically. MME then predicts the total signalling traffic load in order to prevent unnecessary accesses from MTC devices while the network has no resources for serving them. With the help of traffic prediction, total signaling traffic of the core network can be dispersed resulting in increase of access success probability of MTC devices. Moreover, access delay of MTC devices can be decreased by using our dynamic reservation with access barring scheme. Furthermore, our proposed scheme provides priority for MTC devices and alleviates the starvation problem of them. Simulation results show that the proposed scheme significantly decreases the access delay and improves the access success probability of MTC devices in comparison with the extended access barring scheme adopted by 3GPP.