Results 1  10
of
42
Efficient Fair Queuing using Deficit Round Robin
 SIGCOMM '95
, 1995
"... Fair queuing is a technique that allows each flow passing through a network device to have a fair share of network resources. Previous schemes for fair queuing that achieved nearly perfect fairness were expensive to implement: specifically, the work required to process a packet in these schemes was ..."
Abstract

Cited by 349 (3 self)
 Add to MetaCart
Fair queuing is a technique that allows each flow passing through a network device to have a fair share of network resources. Previous schemes for fair queuing that achieved nearly perfect fairness were expensive to implement: specifically, the work required to process a packet in these schemes was O(log(n)), where n is the number of active flows. This is expensive at high speeds. On the other hand, cheaper approximations of fair queuing that have been reported in the literature exhibit unfair behavior. In this paper, we describe a new approximation of fair queuing, that we call Deficit Round Robin. Our scheme achieves nearly perfect fairness in terms of throughput, requires only O(1) work to process a packet, and is simple enough to implement in hardware. Deficit Round Robin is also applicable to other scheduling problems where servicing cannot be broken up into smaller units, and to distributed queues.
Starttime Fair Queuing: A Scheduling Algorithm for Integrated Services Packet Switching Networks
 In Proceedings of the ACM SIGCOMM '96 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication
, 1996
"... We present Starttime Fair Queuing (SFQ) algorithm that is computationally efficient, achieves fairness regardless of variation in a server capacity, and has the smallest fairness measure among all known fair scheduling algorithms. We analyze its throughput, single server delay, and endtoend delay ..."
Abstract

Cited by 191 (12 self)
 Add to MetaCart
We present Starttime Fair Queuing (SFQ) algorithm that is computationally efficient, achieves fairness regardless of variation in a server capacity, and has the smallest fairness measure among all known fair scheduling algorithms. We analyze its throughput, single server delay, and endtoend delay guarantee for variable rate Fluctuation Constrained (FC) and Exponentially Bounded Fluctuation (EBF) servers. We show that SFQ is better suited than Weighted Fair Queuing for integrated services networks and it is strictly better than Self Clocked Fair Queuing. To support heterogeneous services and multiple protocol families in integrated services networks, we present a hierarchical SFQ scheduler and derive its performance bounds. Our analysis demonstrates that SFQ is suitable for integrated services networks since it: (1) achieves low average as well as maximum delay for lowthroughput applications (e.g., interactive audio, telnet, etc.); (2) provides fairness which is desirable for VBR video; (3) provides fairness, regardless of variation in server capacity, for throughputintensive, flowcontrolled data applications; (4) enables hierarchical link sharing which is desirable for managing heterogeneity; and (5) is computationally efficient.
Stratified Round Robin: A Low Complexity Packet Scheduler with Bandwidth Fairness and Bounded Delay
, 2003
"... Fair queuing is a wellstudied problem in modern computer networks. However, there remains a gap between scheduling algorithms that have provably good performance, and those that are feasible and practical to implement in highspeed routers. In this paper, we propose a novel packet scheduler called S ..."
Abstract

Cited by 52 (0 self)
 Add to MetaCart
Fair queuing is a wellstudied problem in modern computer networks. However, there remains a gap between scheduling algorithms that have provably good performance, and those that are feasible and practical to implement in highspeed routers. In this paper, we propose a novel packet scheduler called Stratified Round Robin, which has low complexity, and is amenable to a simple hardware implementation. Stratified Robin Robin exhibits good fairness and delay properties that are demonstrated through both analytical results and simulations. In particular, it provides a single packet delay bound that is independent of the number of flows. This property is unique to Stratified Round Robin among all other schedulers of comparable complexity.
TimeShift Scheduling – Fair Scheduling of Flows in HighSpeed Networks
 IEEE/ACM Transactions on Networking
, 1998
"... We present a scheduling protocol, called TimeShift scheduling, to forward data packets from multiple input flows to a single output channel. Each input flow is guaranteed a predetermined forwarding rate and an upper bound on packet delay. The protocol is an improvement over existing protocols becau ..."
Abstract

Cited by 25 (6 self)
 Add to MetaCart
(Show Context)
We present a scheduling protocol, called TimeShift scheduling, to forward data packets from multiple input flows to a single output channel. Each input flow is guaranteed a predetermined forwarding rate and an upper bound on packet delay. The protocol is an improvement over existing protocols because it satisfies the properties of low delay, fairness, and efficiency, while existing protocols fail to satisfy at least one of these properties. In TimeShift scheduling, each flow is assigned an increasing timestamp, and the packet chosen for transmission is taken from the flow with the least timestamp. The protocol features the novel technique of time shifting, in which the scheduler's realtime clock is adjusted to prevent flow timestamps from increasing faster than the realtime clock. This bounds the difference between any pair of flow timestamps, thus ensuring the fair scheduling of flows. 1.
Jitter Control in QoS Networks
, 2001
"... We study jitter control in networks with guaranteed quality of service (QoS) from the competitive analysis point of view: we propose online algorithms that control jitter and compare their performance to the best possible (by an offline algorithm) for any given arrival sequence. For delay jitter, ..."
Abstract

Cited by 21 (0 self)
 Add to MetaCart
(Show Context)
We study jitter control in networks with guaranteed quality of service (QoS) from the competitive analysis point of view: we propose online algorithms that control jitter and compare their performance to the best possible (by an offline algorithm) for any given arrival sequence. For delay jitter, where the goal is to minimize the difference between delay times of different packets, we show that a simple online algorithm using a buffer of slots guarantees the same delay jitter as the best offline algorithm using buffer space P. We prove that the guarantees made by our online algorithm hold, even for simple distributed implementations, where the total buffer space is distributed along the path of the connection, provided that the input stream satisfies a certain simple property. For rate jitter, where the goal is to minimize the difference between interarrival times, we develop an online algorithm using a buffer of size P C for any I, and compare its jitter to the jitter of an optimal offline algorithm using buffer size. We prove that our algorithm guarantees that the difference is bounded by a term proportional to B/h.
Preserving Quality of Service Guarantees in spite of Flow Aggregation
 In Proceedings of the 6th International Conference on Network Protocols (ICNP’98
, 1998
"... Abstract — We investigate how quality of service may be guaranteed to a flow of packets in the presence of flow aggregation. For efficiency, multiple flows, known as the constituent flows, are merged together resulting in a single aggregate flow. After the network node where the aggregation occurs, ..."
Abstract

Cited by 15 (0 self)
 Add to MetaCart
(Show Context)
Abstract — We investigate how quality of service may be guaranteed to a flow of packets in the presence of flow aggregation. For efficiency, multiple flows, known as the constituent flows, are merged together resulting in a single aggregate flow. After the network node where the aggregation occurs, packet schedulers are aware of the aggregate flow, but are unaware of its constituent flows. In spite of this, we show that quality of service may be guaranteed to the constituent flows, provided the aggregation is performed fairly. When the delay bound of a flow is decoupled from the flow’s reserved rate, flow aggregation preserves the delay bound. When the delay bound of a flow is coupled to the flow’s reserved rate, flow aggregation preserves, and in some cases improves, the delay bound.
Leap Forward Virtual Clock: An O(log log N) Fair Queuing Scheme with Guaranteed Delays and Throughput Fairness
, 1996
"... We describe an efficient fair queuing scheme, Leap Forward Virtual Clock, that provides endtoend delay bounds almost identical to that of PGPS fair queuing, along with throughput fairness. Our scheme can be implemented with a worstcase time O(log log N) per packet (inclusive of sorting costs), wh ..."
Abstract

Cited by 12 (0 self)
 Add to MetaCart
We describe an efficient fair queuing scheme, Leap Forward Virtual Clock, that provides endtoend delay bounds almost identical to that of PGPS fair queuing, along with throughput fairness. Our scheme can be implemented with a worstcase time O(log log N) per packet (inclusive of sorting costs), which improves upon all previously known schemes that achieve guaranteed delay and throughput fairness. As its name suggests, our scheme is based on Zhang's virtual clock. While the original virtual clock scheme does not achieve throughput fairness, we can modify it with a simple leap forward mechanism that keeps the server clock from lagging too far behind the packet tags. We prove that our scheme guarantees a fair share of the available bandwidth to each of the backlogged users, while precisely matching the delay bounds of PGPS schemes. In order to improve computational efficiency, we introduce a "coarsened" version of our scheme in which all tags assume values from a set of O(N) integers. We then use "approximate sorting" and a finiteuniverse priority queue to achieve O(log log N) processing time per packet. We can show that the coarsening of tags increases the delay bound by a very small additive constant. Finally, our proofs are based on a dual version of the algorithm called Leap Backward, whose behavior is identical to the Leap Forward but that admits a simpler analysis.
The stratified round robin scheduler: design, analysis and implementation
 IEEE/ACM Trans. Netw
"... ..."
(Show Context)
A theory of multichannel schedulers for quality of service
 J. High Speed Netw
, 2002
"... A computer network consists of a set of computing nodes interconnected via communication channels. It is commonly assumed that, for each pair of network nodes u and v, there is at most one channel from u to v. However, it is often desirable to have multiple channels between nodes. That is, for every ..."
Abstract

Cited by 7 (0 self)
 Add to MetaCart
(Show Context)
A computer network consists of a set of computing nodes interconnected via communication channels. It is commonly assumed that, for each pair of network nodes u and v, there is at most one channel from u to v. However, it is often desirable to have multiple channels between nodes. That is, for every pair of network nodes u and v, there may be multiple channels from u to v. In this paper, we consider the problem of providing deterministic quality of service guarantees when there are multiple channels between nodes. We show that any packet scheduling protocol that operates over a single channel can be modified to operate over multiple channels. In addition, this transformation increases the packet delay through the node by only a small amount. However, having multiple channels between nodes may cause packet reorder. This reorder significantly increases the upper bound on endtoend delay. We show how this increase in delay is avoided through the use of efficient sorting techniques. 1