Scheduling Policies in Time and Frequency Domains for LTE Downlink Channel: A Performance Comparison

Abstract

A key feature of the Long-Term Evolution (LTE) system is that the packet scheduler can make use of the channel quality information (CQI), which is periodically reported by user equipment either in an aggregate form for the whole downlink channel or distinguished for each available subchannel. This mechanism allows for wide discretion in resource allocation, thus promoting the flourishing of several scheduling algorithms, with different purposes. It is therefore of great interest to compare the performance of such algorithms under different scenarios. Here, we carry out a thorough performance analysis of different scheduling algorithms for saturated User Datagram Protocol (UDP) and Transmission Control Protocol (TCP) traffic sources, as well as consider both the time- and frequency-domain versions of the schedulers and for both flat and frequency-selective channels. The analysis makes it possible to appreciate the difference among the scheduling algorithms and to assess the performance gain, in terms of cell capacity, users' fairness, and packet service time, obtained by exploiting the richer, but heavier, information carried by subchannel CQI. An important part of this analysis is a throughput guarantee scheduler, which we propose in this paper. The analysis reveals that the proposed scheduler provides a good tradeoff between cell capacity and fairness both for TCP and UDP traffic sources

5MART: A 5G SMART Scheduling Framework for Optimizing QoS Through Reinforcement Learning

“…In (15), each normalized componentĉ k,p,u can be considered as a product of random variables due to the uncertainty given by (3), in which the evolution of controllable state at TTI t+1 depends on s[t] ∈ S that includes both c[t] ∈ S C and v[t] ∈ S U . According to [17], each vectorĉ k,p = [ĉ k,p,u1 ,ĉ k,p,u2 , ...,ĉ k,p,uU p ], k = {1, 2, ..., 7}, p = {1, 2, ..., P }, can be modeled as normally distributed variables.…”

“…Due to application diversity and the heterogeneity of QoS requirements, most of these state-of-the-art schedulers are rather static, being unable to meet the QoS requirements under dynamic networking conditions [15]. By performing time domain prioritization under the same static metric, some traffic classes will be over-provisioned while others will be starved, resulting in poor QoS provisioning.…”

Proposta De Um Novo Algoritmo Qos-Aware Para O Escalonamento Downlink Lte-a Em Cenários De Tráfego Misto: Uma Comparação De Desempenho

“…Apresenta-se, a seguir, os algoritmos de escalonamento downlink comparados neste trabalho e a Tabela I, adaptada de Grondalen et al (2017), resume as principais características de cada algoritmo.…”

Optimal Features-based Channel Selection and Neural Network Learning for LTE applications

“…The comparative results of improved fairness and high-quality cell throughput was obtained. [11] Suggested the LTE scheduling downlink channel policies with time and frequency domain. They promoted several flourishing scheduling algorithms with the different purpose and allowed wide discretion of resource allocation.…”