Performance analysis of an enhanced delay sensitive LTE uplink scheduler for M2M traffic

The Long Term Evolution (LTE) standard is one of the most promising wireless access technologies for Machine to Machine (M2M) communications because of its high data rates, low latency and economies of scale. M2M communications typically involves a large number of autonomous devices sending traffic in a coordinated manner (and possibly even simultaneously), therefore creating an uplink-heavy trend which needs an efficient radio resource management scheme. The conventional scheduling algorithms and performance metrics are not suitable for M2M systems because of the different characteristics and service requirements of M2M traffic. In this paper, we analyze the performance of an enhanced delay sensitive uplink scheduler in context of LTE TDD configurations 0 and 1 for delay sensitive event based M2M traffic. We show that unlike an ordinary equal capacity fair scheduler, our proposed delay sensitive scheduler can make utmost use of the maximally uplink-biased TDD configuration 0, attaining higher capacity and maximizing the chance of satisfying packet delay budget of M2M traffic. We also introduce a new performance metric called 'Effective Allocated Bits/RB pair' to measure the allocation efficiency of a scheduler, evaluate the performance of the proposed scheduler in terms of this metric and identify the scope of possible improvements.