Efficient resource scheduling for eMTC/NB-IoT communications in LTE Rel. 13

Next generation cellular networks (5G) have to deal with deployment of massive machine-type-communication (MTC) devices as the internet of things (IoT) becomes pervasive. To cater to such demand, 3GPP has finalized specifications for enhanced machine type communication (eMTC) and narrowband IoT (NB-IoT) as part of LTE Release 13
¹
. To avoid resource congestion to legacy users, eMTC communication can be limited to predefined eMTC narrowband(s) with receiver bandwidth of 1.08MHz, whereas in NB-IoT all communications are limited to a single resource block
²
. Both eMTC and NB-IoT are designed to cater for bandwidth limited low complexity/coverage enhancement (BL/CE) devices, with MTC messages repeated across several BL/CE subframes lasting up to several seconds, to accomodate for large maximum coupling losses (MCL). To avoid MTC users from hogging, collide with legacy resources during lengthy repetitions, 3GPP defines BL/CE subframes and has made provisions for valid and invalid BL/CE subframes to be configured. With large number of MTC devices sharing limited resources and message repetitions lasting several SFs, the LTE medium access control (MAC) scheduler processing burden will be significantly increased. This is because the scheduler operates on a subframe (SF) basis and has to find the MTC users who can be scheduled either in the current SF, or at a future SF after ongoing repetitions finish, or within a gap between previously scheduled repetitions. Another complexity is that the control messages(e.g., MPDCCH) can only be scheduled at certain start intervals and within predefined boundaries. In this paper we address these problems with a low complexity constant time scheduling scheme for MTC users
³
when invalid BL/CE SFs are configured. The proposed algorithms are generic and applicable to both eNB and UE implementation. With detailed system level simulations we demonstrate the usefulness of the proposed scheme when large number of MTC users are active in the system.