Research On LTE Uplink Resource Scheduling Of Heterogeneous Services

In the future, LTE will become a complex network carrying both customer services (H2H)and Internet of Things services (M2M). LTE follows the3GPP36series of standard, andprovides QoS guarantees for different performance of services by dividing nine kinds of QoSstandard and establishing dedicated EPS bearer. However, since M2M services are based onusers demand which leads to various types, unique activities, different needs of service, andthe number of M2M terminals is tens of times of H2H terminals in one LTE cell. When thenetwork serves the two kind of terminals simultaneously, it will lead to the increase of thecollision probability of LTE, also the spectrum utilization rate will decline. It is not onlyunable to meet QoS requirements of M2M services, but also affect the quality of H2H serviceresulting in system performance and user experience degradation.Recently, wireless optical communication (OW) have attracted much attention because ofits characteristics of high bandwidth, anti-interference, low-power, free-spectral and low cost.On the other hand, some nature features, like huge number of M2M terminal, energyconstrained, are matched to the characteristics of wireless optical communication.In order to make better support for M2M service in LTE networks, this paper introducesoptical access unit based on the behavior and characteristics of M2M services, and proposesnew LTE uplink schedule mechanisms in H2H only scenario, M2M only scenario and H2Hand M2M mixed scenario. QoS of H2H and M2M services are guaranteed while systemresource utilization is maximized.First, based on the analysis of the characteristics of M2M services, this paper introducesthe coordinator in LTE access network, realizing the group management of M2M terminals.M2M terminals will be grouped by location, service characteristics and QoS requirements.There is one coordinator in each group, M2M terminals access to the coordinator using IEEE802.15.7protocol, and coordinator interact with LTE as a hyperterminal which use3GPP36series protocol.Next, this paper proposes new LTE uplink schedule mechanisms in H2H service onlyscenario, M2M service only scenario and H2H and M2M mixed services scenario. In H2Hservice scenario, this paper proposes time domain-frequency domain resource allocationmethod named Im-alg. In time domain, the proportional Fair (PF) algorithm based on Analytic Hierarchy Process is improved considering QoS indicators as LTE system servicesclassification, delay, packet loss rate, service rate etc. In frequency domain, we consider twoconstraint conditions that the spectrum resources must be continuous and the samemodulation coding mechanism should be used in LTE uplink, and improve recursivemaximum expansion method. In M2M service only scenario, we propose a semi-staticscheduling mechanism named Im-alg1based on M2M services periodic arrival, small andfixed data and group manageable characteristics. In H2H and M2M mixed scenario, in orderto ensure the quality of service of M2M services, minimize M2M services’ impact on H2Hservices, we set a sending window for M2M services according to its needs of delay andsending rate. The window include a upper bounds and a lower bounds, The lower bounds canensure the quality of service of M2M services, and the upper bounds is the amount of datethat the user want to send or reach the system threshold. On the basic of it, we put forward aLTE uplink mechanism named Im-alg2.Last, this paper compares the system collision probability when M2M services aremanaged as a group or as a unit based on Matlab simulation. The results show that whenmanaged as a group the system collision probability significantly decreased. Moreover, thenew schedule mechanism simulation comparison of three scenarios is also realized on theVienna LTE Simulation Platform. Simulation results show that the proposed schedulingmechanism has some advantages compared to ordinary proportional fairness algorithm.