Abstract

Future wireless networks like 5G will carry an increasingly wide variety of data traffic, with different QoS requirements. In addition to conventional data traffic generated from HTTP, FTP and video streaming applications by mobile broadband users [human-type communication (HTC)], traffic from machine-to-machine (M2M) and Internet-of-Things (IoT) applications [machine-type communication (MTC)] has to be supported by 5G networks. Time-of-day variation in arrival rate of connection-level requests and randomness in length of data sessions in HTC result in randomness in aggregate traffic. In MTC, randomness in traffic arises from random number of devices trying to connect to the base station at any given time. Traffic generated by MTC devices may be either periodic or event-triggered. Nevertheless, it is difficult to model aggregate traffic due to non-stationary nature of traffic generated by each type of service. In this paper, special correlation functions of stochastic point processes called Product Densities (PDs) are used for estimating aggregate traffic under non-stationary arrival rates. For HTC, PDs are defined for estimating time-dependent offered load of connection-level service requests and expected number of ON periods in an interval of time \((0,T)\). The aggregate traffic is evaluated for light-tail (exponential) and heavy-tail (hyper exponential) servicing times. For MTC, PDs are defined for estimating the random number of devices connected to the base station at any time. Another QoS parameter of interest in high speed networks is the expected number of service requests/devices delayed beyond a critical value of delay. Bi-variate PD is defined to estimate the number of service requests/devices delayed beyond a given critical threshold. The results from PD model show close agreement with simulation results. The proposed PD technique proves effective in performance analysis under time-dependent traffic conditions, and is versatile for application to several studies in wireless networks including power consumption, interference and handover performance.

中文翻译：

---

在5G网络中建模与时间相关的聚合流量

摘要

诸如5G之类的未来无线网络将承载越来越多的数据流量，并具有不同的QoS要求。除了移动宽带用户从HTTP，FTP和视频流应用程序生成的常规数据流量[人类式通信（HTC）]，机器对机器（M2M）和物联网（IoT）应用程序的流量[ 5G网络必须支持[机器类型通信（MTC）]。连接级别请求的到达时间的一天中的时间变化以及HTC中数据会话长度的随机性导致总流量的随机性。在MTC中，流量的随机性是由在任意给定时间尝试连接到基站的随机设备数量引起的。MTC设备生成的流量可能是定期的，也可能是事件触发的。不过，由于每种服务类型产生的流量的非平稳性质，很难对总流量建模。在本文中，随机点过程的特殊相关函数称为产品密度（PD），用于估计非平稳到达率下的总流量。对于HTC，定义了PD，以估计与时间有关的连接级服务请求的负载以及在一定时间间隔内预期的ON周期数\（（（0，T）\）。评估总流量的轻尾（指数）和重尾（超指数）服务时间。对于MTC，定义了PD，以估计随时连接到基站的设备的随机数。高速网络中另一个令人关注的QoS参数是延迟的服务请求/设备的预期数量超过延迟的临界值。定义了双变量PD，以估计延迟超过给定关键阈值的服务请求/设备的数量。PD模型的结果与仿真结果非常吻合。所提出的PD技术在时变流量条件下的性能分析中被证明是有效的，并且对于在无线网络中进行多项研究（包括功耗，干扰和切换性能）具有广泛的应用。