Analytical modelling and performance evaluation of realistic time-controlled M2M scheduling over LTE cellular networks

Supporting emerging machine-to-machine (M2M) communications over Long-term Evolution (LTE)/LTE Advanced cellular networks in an efficient way will be beneficial for both telecommunication communities. The first step to migrate to an M2M-enabled cellular standard is to provide these new services through the existing architectures and protocols, while maintaining seamless backward compatibility. To this end, we thoroughly examined a key LTE Medium Access Control entity, which is the packet scheduler, and proposed solutions based on the time-controlled M2M feature, to deal with the diverse M2M traffic characteristics and quality-of-service requirements. Starting from the single M2M class case, we extended our study to more realistic scenarios, involving more M2M classes with diverse quality-of-service requirements. We defined analytical models for predicting the system performance on the basis of queueing theory concepts and considered the interaction between classes with different priorities. The proposed analytical models are validated through extensive system-level simulations. On the basis of the insight obtained from our analytical approach, we modified an existing scheduling algorithm to improve the performance of low-priority M2M device groups, and we demonstrated its superior performance both experimentally and analytically. Copyright © 2013 John Wiley & Sons, Ltd. *Correspondence Antonis G. Gotsis, Department of Digital Systems, University of Piraeus, 126 Grigoriou Lampraki Street, Office 501, GR18532, Piraeus, Greece. E-mail: [email protected] Received 28 May 2012; Revised 30 November 2012; Accepted 2 February 2013