• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 통신소사이어티 추계학술대회논문집
• /
• pp.168-171
• /
• 2003

IEEE 802.16 Wireless MAN standard specifies the air interface of fixed point-to-multipoint broadband wireless access systems providing multiple services. Among the service classes supported by the wireless MAN, the best effort service class is ranked on the lowest position in priority and is assisted by a MAC scheme based on reservation ALOHL Such MAC scheme must include a number of components, while many of them are not specified in the standard. In this paper, we thus reveal main components of a MAC scheme supporting the best effort service and present candidate schemes implementing such components. Combining schemes for implementing components, we then construct distinctive MAC schemes supporting best effort service. In designing a MAC scheme, the delay performance induced by the scheme should be considered since scarce resource may be available for the best effort service after the preemptive resource occupation by other service classes. In this paper, we focus on the delay fairness among the subscriber stations using the best effort service. For evaluating a MAC scheme in delay fairness, we present two definitions of delay fairness and provide a criterion for optimal MAC scheme according to each definition of delay fairness. Using a simulation method we investigate the mean delay performance exhibited by each MAC scheme arid find an optimal scheme in delay fairness. From numerical examples, we observe that SR/ED/PG＋P scheme has strong delay fairness compared with MR/ED/PG＋P and SR/ED/PG＋D schemes according to a definition of delay fairness. However, other schemes are rather shown to have better delay fairness when the other definition is adopted.

• Journal of electromagnetic engineering and science
• /
• 제9권3호
• /
• pp.124-129
• /
• 2009

Although the maximum sum-rate capacity of multiple-input multiple output(MIMO) broadcast channels(BCs) can be achieved by dirty-paper coding(DPC), the results were obtained without fairness considerations in uncorrelated MIMO channels. In this paper, we propose new multiuser scheduling algorithms, which find a best user set for approaching the maximum sum-rate capacity while maintaining fairness among users. We analyze the performance of the proposed algorithms using zero-forcing dirty paper coding(ZF-DPC) in the correlated MIMO BCs for throughput and delay fairness, respectively. Numerical results demonstrate that a large time window can reduce the average throughput difference between users, but it increases head-of-line(HOL) delay jitters in the case of delay fairness.

• ETRI Journal
• /
• 제40권3호
• /
• pp.347-354
• /
• 2018

Using various offset times to separate differential services is the most common form of service differentiation in optical burst switching networks. In this approach, a larger offset time is given to a higher priority burst, but it causes this burst to have a longer delay. One solution to this problem is to adjust the burst assembly time so that the buffering delay of the higher priority burst is always shorter than that of the lower priority burst. However, this adjustment causes another problem, called delay unfairness, for bursts with differential priorities that share the same path to their destination. This article proposes a new solution for delay fairness using the burst assembly.

• 한국통신학회논문지
• /
• 제22권1호
• /
• pp.139-150
• /
• 1997

For high-speed networks, a new ring protocal is proposed in this paper. A ring network combined with destination removal can achieve much higher network throughput than the channel transmission rate. However, such a network exhibits fairness problems. Over a past few years, global fairness algorithms such as ATMR and Metaring have been proposed to solve such problems. But the ring access time delay and fairness in such networks are sensitive to the network parameters such as network size and traffic distribution. In addition to guaranteeing fair ring access to all nodes, there are several other important performance aspects in such networks. The one is that fairness is enforced while node throughputs are kept as high as possible. And another performance measure is access delay and more specifically Head-Of-Line(HOL) delay, i.e., the amount of time the first cell in the transmission buffer of a node has to wait before it accesses the ring. HOL delay is a mijor component in the transmission jitter of the synchronous traffic transmission. A key idea of the proposed ring protocol is to find the nodes that have much more chances to access the ring than any other nodes in the independently distributed node architecture. Since destined by many cells need to share a part of the bandwidths with the next node for the fairness in as much as performance degradation does not become critical. To investigate the performance behavior of the proposed ring protocol for various network condition,s several performance parameters wuch as ring access time delay, and throughput are compared with those of the ATMR and Metaring protocols using simulation package, SIMAN.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.181-181
• /
• 2000

We identified the performance problems of scheduling algorithms such as FCFS, and demonstrated the superiority of WFQ in terms of realtime performance measures. For this purpose, we presented the service scenario and performed the analysis for the delay bound and fairness which are required to support the realtime applications in the Internet.

• 한국통신학회논문지
• /
• 제27권6B호
• /
• pp.586-592
• /
• 2002

본 논문에서는 이동 접속망에서 발생하는 지연시간을 줄이기 위해 AMC/TDM 무선 데이터 통신에 적합한 새로운 스케줄러 알고리즘을 제안하였다. 새로이 제안된 알고리즘은 비례공정(Proportional fairness) 알고리즘과 M-LWDF(Modified Largest Weighted Delay First) 알고리즘에 기초를 하면서, Xin Liu에 의해 제안된 최소성능보장 알고리즘의 개념을 응용하여 제안되었다. 제안된 알고리즘은 best-effort 서비스뿐만 아니라, QoS 보장서비스에도 적용 가능한 알고리즘으로, 모의실험결과 본 논문에 의해 새로이 제안된 알고리즘을 적용하였을 경우 기존의 비례공정알고리즘이나, M-LWDF알고리즘에 비해 각각 최대 11.5%와 9%의 전송 지연시간을 줄일 수 있었다.

• 한국통신학회논문지
• /
• 제31권7A호
• /
• pp.717-724
• /
• 2006

본 논문에서는 MIMO-OFDM 시스템에서 물리 계층으로부터의 궤환(Feedback) 정보를 이용하여 MAC 계층에서의 전송 용량 극대화 및 공평성을 제공하는 FATM(Fairness Aware Throughput Maximizing) 스케줄링 방식 기반의 QoS 지원을 위한 우선순위 큐 기반의 스케줄링 기법을 제안한다. 제안된 우선순위 큐 스케줄링 기법은 서 비스 클래스별 지연 요구 조건을 최대한 보장함과 동시에, 강한 버스트 특정을 갖는 서비스 클래스의 성능 저하를 효과적으로 방지할 수 있다. 제안된 스케줄링 기법의 QoS 지원을 위한 큐 스케줄링 기법은 SPQ(Strict Priority Queueing), DCBQ(Delay Constraint Based Queuing), HDCBQ(Hybrid Delay Constraint Based Queuing)로 나누어지고, 세 가지 방식에 대한 모의실험을 통해 성능을 비교 평가하였다.

• 정보과학회 논문지
• /
• 제42권9호
• /
• pp.1162-1174
• /
• 2015

오늘날 네트워크는 높은 대역폭과 높은 지연을 갖는 HBDP (High Bandwidth Delay Product) 네트워크의 특징을 보인다. 기존 TCP는 혼잡윈도우 크기의 느린 증가와 급격한 감소로 인하여 HBDP 네트워크에 부적절하다. 기존 TCP의 문제점을 해결하기 위해 연구된 TCP들은 손실기반 TCP와 지연기반 TCP로 구분한다. 대다수의 TCP는 기존 Slow Start 동작을 사용하며 오버슈트로 인한 대량의 패킷 손실을 초래한다. Congestion Avoidance 동작의 경우 손실기반 TCP는 대역폭 낭비와 RTT (Round Trip Time) 공정성 문제가 있으며 지연기반 TCP는 낮고 느린 대역폭 점유 문제가 있다. 제안하는 기법은 병목구간의 버퍼상태를 통해 혼잡제어를 함으로써 Slow Start와 Congestion Avoidance의 문제를 개선한다. 성능평가를 통해 HBDP 네트워크에서 제안하는 기법이 기존 TCP보다 향상된 성능을 보임을 확인하였다.

• Journal of Communications and Networks
• /
• 제11권5호
• /
• pp.464-472
• /
• 2009

In this paper, we propose a packet scheduling algorithm for cellular relay networks by considering relay selection, variation of channel quality, and packet delay. In the networks, mobile users are equipped with not only cellular but also user relaying radio interfaces, where base station exploits adaptive high speed downlink channel. Our proposed algorithm selects a user with good cellular channel condition as a relay station for other users with bad cellular channel condition but can get access to relay link with good quality. This can achieve flexible packet scheduling by adjusting transmission rates of cellular link. Packets are scheduled for transmission depending on scheduling indexes which are calculated based on user's achieved transmission rate, packet utility, and proportional fairness of their throughput. The performance results obtained by using computer simulation show that the proposed scheduling algorithm is able to achieve high network capacity, low packet loss, and good fairness in terms of received throughput of mobile users.

• 한국ITS학회 논문지
• /
• 제14권3호
• /
• pp.24-30
• /
• 2015

Wi-Fi 표준 규격에 기술되어 있는 전력절감방식으로 동작하는 Wi-Fi 단말이 Wi-Fi AP로부터 데이터 프레임을 수신하는 경우에 시스템 성능(데이터 처리량, 시간지연)이 저하되는 문제점이 발생할 수 있다. 무선채널 상태로 인한 데이터프레임 수신 실패와 이에 따른 재전송 횟수 증가가 이러한 문제점의 원인이 된다. 본 논문에서는 이러한 문제점을 개선하기 위하여 공정성을 고려한 수정된 전력절감방식을 제안하였다. 모의 실험을 수행하여 제안한 패킷 전송 시도 횟수제한 방법과 HARQ를 사용할 경우 전송 시간지연의 변화없이 총 데이터 처리량을 증가를 얻을 수 있음을 보였다.