• Journal of KIISE
• /
• v.44 no.5
• /
• pp.469-475
• /
• 2017

The emerging NVMe-based multi-queue SSDs provides a high bandwidth by parallel I/O, i.e., each core performs I/O through its dedicated queue in parallel with other cores. To provide a bandwidth share for each application with I/O, a fair-share scheduler that provides a bandwidth share to each core is required. In this study, we proposed a multi-core scalable fair-queuing algorithm for multi-queue SSDs. The algorithm adopts randomization to minimize the inter-core synchronization overheads and provides a weight-proportional bandwidth share to each core. The results of our experiments indicated that the proposed algorithm gives accurate bandwidth partitioning and outperforms the existing FlashFQ scheduler, regardless of the number of cores for a Linux kernel with block-mq.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.3B
• /
• pp.171-182
• /
• 2003

In this paper, we propose an efficient and simple fair queueing algorithm, called Minimum Possible Virtual Start Time Fair Queueing (MPSFQ), which has O(1) complexity for the virtual time computation while it has good delay and fairness properties. The key idea of MPSFQ is that it has an easy system virtual time recalibration method while it follows a rate-proportional property. MPSFQ algorithm recalibrates system virtual time to the minimum possible virtual start time of all backlogged sessions. We will show our algorithm has good delay and fairness properties by analysis.

• Journal of KIISE:Computer Systems and Theory
• /
• v.33 no.9
• /
• pp.585-591
• /
• 2006

Generalized Processor Sharing(GPS) model provides instantaneous fair services to currently backlogged sessions. Since this fair service distributes server capacity to backlogged sessions in proportion to their weights, the fairness is only valid between the sessions serviced at the same time. From the long time view, however, this fair service provides different server capacity with one session or another, even if these sessions have the same weights. This paper proposes a cumulative fair service(CFS) model to provide fair server capacity to all sessions in the long time view. This model provides fair service in session Viewpoint because it distributes server capacity in proportion to the weights of sessions. The model and an algorithm referencing that model are analyzed for their properties and performances. Performance evaluations verify that the proposed algorithm provides proportional service capacity to sessions in the long time view.

• IE interfaces
• /
• v.17 no.spc
• /
• pp.111-121
• /
• 2004

In this paper, we propose an urgency and efficiency based wireless packet scheduling (UEPS) algorithm that is able to schedule real time (RT) and non-real time (NRT) traffics at the same time. The proposed UEPS algorithm is designed to support wireless downlink packet scheduling in the OFDMA system which is a strong candidate wireless system for the next generation mobile communications. The UEPS algorithm uses the time-utility function as a scheduling urgency factor and the relative status of the current channel to the average one as an efficiency indicator of radio resource usage. The design goal of the UEPS algorithm is to maximize throughput of NRT traffics with satisfying QoS requirements of RT traffics. The simulation study shows that the proposed UEPS algorithm is able to give better throughput performance than existing wireless packet scheduling algorithms such as proportional fair (PF) and modified-largest weighted delay first (M-LWDF) while satisfying QoS requirements of RT traffics such as the average delay and the packet loss rate under various traffic loads.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.11A
• /
• pp.861-869
• /
• 2011

In this paper, we propose coordinated scheduling and beamforming based on limited feedback to overcome the effects of inter-cell interference in multi-cell environment. The goal of this study is to improve a performance of cell-edge area significantly affected by the inter-cell interference and also to guarantee a performance of the entire system. Existing coordinated scheduling and beamforming based on the perfect channel state information generates an overhead for the channel state information feedback and exchange at the users and base stations, respectively. In this study, first we propose a novel codebook to obtain a null space of an interference channel efficiently which can be used to reduce the inter-cell interference. Using the proposed codebook, we propose coordinated scheduling and beamforming to improve the performance of cell-edge area and the entire system. Through the proportional fair scheduling, the performance of the proposed technique is shown as improvement of channel capacity for users located cell-edge area and entire system.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.48 no.11
• /
• pp.60-66
• /
• 2011

LTE (Long Term Evolution) supports QoS (Quality of Service) with several service classes. For each class of traffic, a big difference exists on BER (Bit Error Rate) requirement. This leads to a considerable difference in transmission power for various classes of traffic. In this paper, a novel scheduler is designed and proposed for LTE which supports CoS (Class of Service) with the consideration of priority as well as target BER. By the CQI (Channel Quality Indicator) and QCI (QoS Class Identifier), a minimum transmission power is assigned from the target BER for each class of traffic per each user. Hence, with the other information such as user's used rate in the past and the priority of traffic, the probability of occupying channels is determined. The simulation results of Service Class scheduling are compared with that of Maximum Rate and Proportional Fair. The results show that the service class-aided scheduling can improve the throughput of whole system significantly.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.265-267
• /
• 2013

본 논문에서는 멀티 셀 셀룰러 네트워크에서 채널 환경에 따른 스케줄링 알고리듬의 성능을 평가한다. 무선 통신 환경에서 대표적인 채널모델인 가우시안 채널, 레일리 페이딩 채널, SCME(Spatial Channel Model Extended) 채널 모델에 PF(Proportional Fair)와 MR(Maximum Rate) 스케줄링 기법을 적용했을 때, 시스템 성능을 평가하고 비교한다. 기지국에서 송신된 신호는 채널을 통과하며 변화하고 이는 채널 모델에 따라 달라진다. 채널 모델에 따른 스케줄링 알고리듬 성능 평가 결과 가우시안 채널모델과 레일리 페이딩 채널모델 SCME보다 좋은 성능을 보인다.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.43 no.11 s.353
• /
• pp.85-94
• /
• 2006

Simple-relay aided resource allocation (SRARA) schemes are incorporated with throughput guarantee scheduling (TGS) in IEEE 802.16 type TDD-OFDMA downlink to enhance service coverage, where the amount of resources at each relay is limited due to either its available power which is much smaller than base station (BS) power or the required overhead The performance of SRARA schemes is evaluated with both proportional fair (PF) and TGS schedulers at 64 kbps and 128 kbps user throughput requirements. For SRARA with RSs of relatively lower power, a scheme putting total power into only one subchannel shows larger coverage than when both subchnnels are used in a manner of equal power allocation, while the RS with evenly power-allocated two subchannels could provide larger coverage gain for a relatively higher power. In a lower target (64kbps), more improvement comes from relay scheme rather than scheduler design. For a relatively higher level (128 kbps), it comes from scheduler design rather than relay.

• Journal of Advanced Navigation Technology
• /
• v.10 no.3
• /
• pp.275-281
• /
• 2006

Simple-relay aided resource allocation (SRARA) schemes are incorporated with throughput guarantee scheduling (TGS) in IEEE 802.16 type time division duplex - orthogonal frequency division multiple access (TDD-OFDMA) downlink in order to enhance service coverage, where the amount of resources for relaying at each relay is limited due to either its available power which is much smaller than base station (BS) power or the overhead required for exchanging feedback information. The performance of SRARA schemes is evaluated with schedulers such as proportional fair (PF) and TGS at 64kbps and 128kbps user throughput requirements when total MS power is set to 500mW or 1 W. For 64kbps throughput requirement level, more improvement comes from relay than scheduler design. For 128kbps case, it comse from scheduler design than relay due to the fact that simple relay can't help using strictly limited amount of resources for relaying function.

• Journal of Korea Multimedia Society
• /
• v.9 no.3
• /
• pp.323-332
• /
• 2006

Fair packet scheduling algorithms supporting quality-of-services of real-time multimedia applications can be classified into the following two schemes in terms of the reference time used in calculating the timestamp of arriving packet; the Finish-Time (FT) and Start-Time (ST) schemes. The FT scheme, used in most schedulers, that has the property of an inversely rate-proportional latency is suitable to support various delay bounds because it can adjust the latency of a flow with raising the flow's reserved rate. However, the scheme may incur some bandwidth loss due to excess rate reservation. Meanwhile, although the ST scheme does not suffer from the bandwidth loss, it is hard to support multiple delay bounds because of its latency property relying on the number of flows. This paper is devoted to propose a ST scheme based scheduler to effectively support multiple delay bounds and analyze its performance comparing to the FT scheme based scheduler. The comparison results show that the proposed scheduler gives better utilization by up to 50%.