Research On Performance Optimization Of Hybrid Traffic Based On Scalable Transmission Time Interval

The fifth generation of mobile network(5G)technologies can support multiple traffic types including ultra-reliable low-latency communication(URLLC)and enhanced mobile broadband(eMBB),which are the two main usage scenarios in 5G.URLLC refers to achieving strict requirements of a 99.999%reliability within 1ms,and eMBB refers to a very high demand for data transmission rate.In order to meet the quality of service requirements of the hybrid traffic,there is a trade-off between latency and spectrum efficiency when the above two traffic types coexist in the system.Therefore,how to effectively multiplex URLLC and eMBB traffic on the same frequency band is a main challenge for 5G wireless resource management.In view of the above problems,two user-centric scheduling strategies are proposed to optimize the system performance in hybrid traffic scenarios by exploiting the scalable transmission time interval(TTI)in 5G mobile edge computing architecture and flexible frame structure.The main innovations and contributions are summarized as follows:Firstly,a semi-shared bandwidth strategy based on the scalable TTI is proposed.The main idea of this strategy is to sacrifice part of the capacity of eMBB to meet URLLC latency requirements.When the URLLC traffic is scheduled,the available resource blocks are all allocated for it,and the scheduling process is accelerated by using a shorter TTI of 0.143 ms.When the eMBB traffic is scheduled,the available resource blocks are divided into two parts:one is provided for eMBB which is scheduled with a longer TTI of 1 ms to reduce the overhead of the control layer,and the other is reserved for the subsequent possible burst URLLC traffic.When the reserved resources are idle,the eMBB can share these resources with the URLLC according to a certain borrowing mechanism.Secondly,in order to compensate for eMBB throughput,a novel TCP slow start optimization strategy based on scalable TTI is proposed.The main idea of this strategy is to use a shorter TTI for scheduling during the slow start phase of the eMBB traffic to speed up the TCP slow start process.When time threshold is reached and the throughput approaches saturation,eMBB traffic switches to its own band with long TTI for lower transmission overhead.The overall scheduling of the two schemes are both completed in three phases:the bandwidth initialization phase,the time domain user scheduling phase,and the frequency domain resource allocation phase.Thirdly,a large number of dynamic system level simulations are carried out to evaluate the performance of the proposed schemes and to verify the potential of the proposal.For the semi-shared bandwidth strategy,when the system operates on three different levels of URLLC load,the simulation results show that the proposed scheme can effectively enable the URLLC to meet the latency requirement of 1ms under the low-load and medium-load conditions.Although the standard within 1ms is not reached under high-load condition,a larger URLLC latency gain can be achieved compared to the traditional scheduling scheme.The cost is to bring a negative gain to the eMBB throughput,but no more than 10%.For the TCP slow start optimization strategy,the effectiveness of the scheme is verified under various application scenarios and different system bandwidth conditions.The simulation results show that the proposed scheme can significantly improve the performance of eMBB with almost no impact on throughput and transmission latency for URLLC users.In addition,the TCP slow start optimization strategy can achieve the greatest potential gain when eMBB traffic load is adapted to the available bandwidth.