Online Load Estimation for Efficient Random Access of Machine Type Communications in LTE Networks

Massive access of Machine to Machine (M2M) devices may lead to overload at the access of the LTE networks as the current most popular technology for this type of communications. Several approaches have been proposed to relieve the massive access by time barring or time distributing of access requesting. In virtual-frame technique the requests are spread in consecutive frames to increase the successful accesses. The length of virtualframe, the number of LTE uplink frames which are concatenated for managing the random access of M2M requests, should be selected appropriately according to the network load to compromise the successful access rate and the incurred delay of each access. In this paper, we propose an adaptive scheme by which the eNB is able to find the length of the virtual-frame dynamically and broadcasts this information at the end of each virtualframe. We use an online load estimation technique to adjust the length of virtual-frame. Simulation results show that using this scheme the length of virtual-frame is just increased when the access network load is high which prevents unnecessary delays for lightly loaded conditions.