The End-to-end Random Performance Evaluation Of Lte System

In order to satisfy the mobile communication demands of terminal users, theThird Generation Partnership Project (3GPP) proposed the Long Term Evolution(LTE) project, which aims at realizing smooth evolution from3G to4G. This paperstudies the performance evaluation of a LTE system, specifically its stochasticEnd-to-End (E2E) performance.First, we analyze the system architecture and the Quality of Service (QoS) guar-antee mechanisms of the LTE system. Based on the characteristics of data flows, weapply stochastic network calculus, a queuing system performance evaluation theory,to construct a generalized Stochastically Bounded Burstiness (gSBB) traffic model.Furthermore, we model the service provided by the Multiple Input Multiple Output(MIMO) system in the Radio Access Network (RAN) and the Core Network (CN)using a two-state Markov Chain and a weak stochastic service curve model, respec-tively.Next, we propose an E2E performance evaluation framework for the LTE systemgiven aforementioned traffic model and service model. The proposed framework ena-bles us to obtain certain system metrics (e.g., the stochastic upper bounds on the E2Edelay and backlog). To verify the efficacy of the framework, we conduct simulationstudies using NS2. Simulation Results demonstrate that the proposed framework cansuccessfully distinguish real-time traffic and non-real-time traffic, and provide sto-chastic service guarantees.Finally, we apply the proposed QoS evaluation framework to conduct applicationstudies. Specifically, the efficacy of the framework is verified through varying a set ofparameters in the stochastic traffic and service models.The results presented in this paper can be adopted as a theoretical benchmark forparameter settings in real applications (e.g., traffic arrival rate, traffic burstiness, buff-er size), thus simplify the performance evaluation of LTE systems and improve QoS.In addition, the proposed framework is applicable in wired-cum-wireless networksand a wide variety of mobile networks. It is of both theoretical and application values.