Performance analysis of a machine-to-machine friendly MAC algorithm in LTE-advanced

One of the main challenges in Machine Type Communication (MTC) services in the 3GPP LTE networks is to improve device power efficiency for Machine to Machine (M2M) communication. We consider MTC application scenarios which consist of a number of devices connecting to the network concurrently. The methods considered in TR 37.868 involve Disjoint Allocation (DA) and Joint Allocation (JA). The DA scheme is based on separating the available RACH (random access channel) preambles into two disjoint sets for Human-to-Human (H2H) and M2M calls. In the joint allocation scheme the H2H calls have access to the preamble set of M2M consumers. Performance analyses of the schemes assume that both H2H and M2M calls follow Poisson traffic models. In this paper we consider the joint allocation scheme in which M2M calls have access to H2H preambles as we expect more M2M calls in the future. We derive and present analytical expressions for the throughput, the collision probability, the success probability, and the idle probability for DA and JA methods based on the assumption that RA (random access) arrivals of M2M and H2H devices follow Beta and Poisson distributions respectively. The energy consumption performances of the two methods are compared. Results show that with the new JA method we can reduce the number of collisions and save the power consumption of M2M devices up to 4%.