Energy-efficient scheduling and grouping for machine-type communications over cellular networks

Abstract

In this paper, energy-efficient scheduling for grouped machine-type devices deployed in cellular networks is investigated. We introduce a scheduling-based cooperation incentive scheme which enables machine nodes to organize themselves locally, create machine groups, and communicate through group representatives to the base station. This scheme benefits from a novel scheduler design which takes into account the cooperation level of each node, reimburses the extra energy consumptions of group representatives, and maximizes the network lifetime. As reusing cellular uplink resources for communications inside the groups degrades the Quality of Service (QoS) of the primary users, analytical results are provided which present a tradeoff between maximum allowable number of simultaneously active machine groups in a given cell and QoS of the primary users. Furthermore, we extend our derived solutions for the existing cellular networks, propose a cooperation-incentive LTE scheduler, and present our simulation results in the context of LTE. The simulation results show that the proposed solutions significantly prolong the network lifetime.Also, it is shown that under certain circumstances, reusing uplink resource by machine devices can degrade the outage performance of the primary users significantly, and hence, coexistence management of machine devices and cellular users is of paramount importance for next generations of cellular networks in order to enable group-based machine-type communications while guaranteeing QoS for the primary users.

Aggregation and Resource Scheduling in Machine-Type Communication Networks: A Stochastic Geometry Approach

“…Each scheme with L = 1 performs better than for L = 2. Also, as expected and in general, the CRS setups outperforms the RRS scheme, except when choosing a i , b i according to (33) and (34), for which a greater fairness is attained instead. Notice that when N is small, CRS outperforms RRS; while as N increases, their curves tend to overlap.…”

“…Of course, a given system is projected to work given one value of N, and by properly choosing some fixed a i and b i we can reduce the performance gap between the MTDs sharing the channel, which is not possible for the RRS scheme since weighting the transmit powers is not available. Notice also that by choosing a i and b i according to (33) and (34) both MTDs reach similar performance, hence the fairest scheme. 9 Notice the gap also relies strongly on the value of µ.…”

“…It is expected that a smaller a i , hence larger b i , work better as µ increases (see (22) and (23)). The setup with a i and b i given by (33) and (34), respectively, adapts better to the different µs since this is a parameter taken into account when performing their calculations. This is, the values of a i and b i that allow a similar reliability between both MTDs sharing the same channel in a given cluster, depend on µ, thus no additional adjustment on them is required.…”

Self-Organized Low-Power IoT Networks: A Distributed Learning Approach

“…In [13], network-side reinforcement learning has been proposed for overload control in LTE systems serving massive IoT traffic. In [14], self-organized clustering and clustered-access for massive IoT deployments have been investigated. In [15], use of multi-arm bandit (MAB) for IoT networks has been proposed, where devices learn how to avoid sub-channels suffering from a high level of static interference.…”

Network Lifetime Maximization for Cellular-Based M2M Networks

“…To the best of our knowledge, accurate modeling of energy consumption in machine-type communications, individual and network battery lifetime models, and corresponding scheduling algorithms are absent in literature. As an extension of [5], which investigates clustered-access for massive M2M, in [20] joint energy efficient clustering and scheduling has been investigated, i.e. the cluster-size, selection of clusterheads, and the amount of scheduled resources to clusterheads have been optimized to prolong the battery lifetime.…”