Leaky Bucket Research Articles

Efficient Transmission of MPEG-2 Coded Video Traffic over ATM Networks

Video traffic is expected to account for a significant share of the traffic volume in the future asynchronous transfer mode (ATM) networks. MPEG-2 proposed by Moving Picture Expert Group is one of the most promising compression techniques for such applications. One of the critical issues in MPEG-2 is to realize effective variable bit rate (VBR) video transfer thorough ATM networks. The Leaky Bucket (LB) scheme has been widely accepted as the usage parameter control (UPC) mechanism to police the VBR sources. We proposed a new Adaptive Dynamic Leaky Bucket (ADLB) congestion control mechanism, which is based on the LB scheme. Unlike the conventional LB, the leak rate of the ADLB is controlled using delayed feedback information of available bandwidth sent by the network. This scheme allows sources to get varying amounts of bandwidth over time, while reserving a minimum guaranteed bandwidth (MCR) for the entire duration of the connection. At the time of congestion, the leak rate of the ADLB is adjusted according to the feedback indicating the currently available bandwidth to the connection. The simulation results show that the end-to-end cell transfer delay and cell loss of each source has been improved significantly.

Maximal Effective Bandwidth of Constrained Traffic

We investigate the worst possible behavior of a stationary traffic source when the traffic emanating from it is required to meet certain constraints. Specifically, the peak rate of the source is required not to exceed a level ρ and realizations must obey a “leaky bucket” constraint with bucket size β and leak rate σ. The worst case source is considered to be the one with the largest effective bandwidth, a concept which arises in the large deviation theory of queueing networks and governs the asymptotic loss rate when a large number of sources send traffic to a single server queue. We conjecture the form of the worst case traffic in general and prove the conjecture for the special case when T, the time-scale parameter of the effective bandwidth, is less than both β/(ρ−σ) and β/σ, the times taken respectively to fill and empty the leaky bucket.

Scheduling in multimedia cdma wireless networks

Wireless systems in the future will have to provide multimedia services where different users have different physical-layer quality of service (QoS) requirements (e.g., bit energy per interference power density, E/sub b//N/sub 0/, or bit error rate and power constraints) and network-layer QoS requirements (e.g., delay bound, delay-jitter, throughput, and loss). We investigate the use of power control, processing gain and/or multiple codes, and scheduling in CDMA systems to accommodate these diverse service requirements. We first show that the instantaneous capacity region, given in terms of the set of user bit rates that can be supported simultaneously subject to peak power and E/sub b//N/sub 0/ constraints, is nonconvex. This suggests that by time-sharing the channel, one may be able to get better system throughput. We define the capacity region as the convex hull of the instantaneous capacity region and we show that it may be obtained by time sharing between operating points, where each user either uses full power or is silent (bang-bang control). We then consider the problem of scheduling so as to meet prespecified delay bounds or minimum service curve requirements for traffic streams, which are specified in terms of a traffic profile such as a sigma-rho constraint (enforced by a leaky bucket) and a guarantee that the system is stable.

Simple tariffs based on price multipliers for ATM VBR services

We present an approach for constructing simple time-based tariffs for variable bit-rate (VBR) connections from constant bit-rate (CBR) prices, using price multipliers that depend on the traffic parameters of the VBR connection and reflect the resource usage of VBR connections relative to CBR connections. Our approach employs an effective bandwidth bound as a proxy for the maximum amount of traffic that conforms to the connection's traffic contract, and thus for the maximum resource usage. We compare the price multipliers computed by our approach with the multipliers published by an actual ATM service provider, and with those computed using a proxy for resource usage that is based on ATM Forum's generic CAC algorithm. Although our approach is presented in the context of ATM VBR services, it can be applied for creating tariffs for service level agreements (SLAs) where the maximum amount of conforming traffic is given by leaky (or token) bucket constraints. Copyright © 2003 AEI.

Rate controller for high-speed networks

In the paper, the statistical performance of a leaky bucket rate controller is analysed. A new operating mechanism is proposed for the leaky-bucket rate controller. The proposed mechanism has the major advantage of allowing a simple embedded Markov chain analysis for the rate controller, when the input traffic is modelled using a generally distributed interarrival process. Analysis is developed to study the statistical performance of the rate controller, including the cell delay distribution and the cell loss ratio. Numerical experiments are carried out to illustrate the analysis results and to provide insight into the system design.

On the performance of generalized processor sharing servers under long-range dependent traffic

This paper introduces the computation of delay and backlog bounds for a generalized processor sharing (GPS) server under long-range dependent (LRD) traffic. The traffic is supposed to be regulated by the Fractal Leaky Bucket policing mechanism, which is an appropriate regulator for LRD traffic. Results are extended to a network of GPS servers with arbitrary topology, and the stability of those networks is analysed.

Worst case burstiness increase due to FIFO multiplexing

We consider a FIFO multiplexer fed by flows that are individually constrained by arrival curves, and look for the best possible arrival curve for every output flow. This problem arises in scenarios where aggregate multiplexing is performed, such as differentiated services or front ends to optical switches. We obtain an exact result for a fluid model and for piecewise linear concave arrival curves, which are common in practice and correspond to combinations of leaky buckets.

Some properties of variable length packet shapers

The min-plus theory of greedy shapers has been developed from R.L. Cruz's results (1991) on the calculus of network delays. The theory of greedy shapers establishes a number of properties such as the series decomposition of shapers or the conservation of arrival constraints by reshaping. It applies either to fluid systems or to packets of constant size such as ATM. For variable length packets, due to the distortion introduced by packetization, the theory is no longer valid. We elucidate the relationship between shaping and packetization effects. We show a central result, the min-plus representation of a packetized greedy shaper. We find a sufficient condition under which series decomposition of shapers and conservation of arrival constraints still holds in the presence of packetization effects. This allows us to demonstrate the equivalence of implementing a buffered leaky bucket controller based on either virtual finish times or on bucket replenishment. However, in some examples, if the condition is not satisfied, then the property may no longer hold. Thus, for variable size packets, there is a fundamental difference between constraints based on leaky buckets and constraints based on general arrival curves, such as spacing constraints. The latter are used in the context of ATM to obtain tight end-to-end delay bounds. We use a min-plus theory and obtain results on greedy shapers for variable length packets which are not readily explained with the max-plus theory of C.S. Chang (see "Performance Guarantees in Communication Networks", Springer-Verlag, 2000).

Can equity and efficiency complement each other?

Economists tend to assume that redistributive transfers increase equity but cause a loss in efficiency, the so-called “leaky bucket” effect. This paper explores situations where efficiency losses are small or where equity and efficiency might even complement each other. A simple model identifies key parameters that cause leaky buckets and which policy can affect. Three situations are discussed where the equity/efficiency tradeoff may be low: when transfers go to populations with no capacity to change their behavior; when transfers go to programs that limit efficiency losses through behavioral requirements; and when commodities are subsidized that function as long-term investments and create future income gains.

Procedures and tools for analysis of network traffic measurements

We present procedures and tools for the analysis of network traffic measurements. The tools consist of stand-alone 1 The stand-alone software modules can be downloaded from http://www.ics.forth.gr/netgroup/msa; the web interface is available at http://trace.ucnet.uoc.gr 1 modules that implement advanced statistical analysis procedures and a flexible web-based interface through which a user can create, modify, save, and execute experiments using the statistical analysis modules. The tools do not assume a specific source traffic model, but rather process actual measurements of network traffic. Indeed, the theory that the tools are based on can identify the time-scales that affect a link’s performance, hence suggest the appropriate time granularity of traffic measurements. We present and discuss case studies that demonstrate the application of the tools for answering questions related to network management and dimensioning, such as the maximum link utilization when some quality of service is guaranteed, how this utilization is affected by the link buffer and traffic shaping, the acceptance region and the effects of the scheduling discipline, and the token (or leaky) bucket parameters of a traffic stream. The case studies involve actual traffic measurements obtained by a high performance measurement platform that we have deployed at the University of Crete network.

Buffered and Unbuffered Leaky Bucket Policing: Guaranteeing QoS, Design and Admission Control

Traffic shaping and smoothing using buffers or leaky buckets does not necessarily improve Quality of Service (QoS). In fact there is a trade-off between controlling user traffic and guaranteeing QoS to the users. We consider the first two stages (source node and border node before entering a network cloud) of an end-to-end QoS problem and assume that the QoS requirements across each of the two stages are given. We formulate and solve a mathematical programming problem to select optimal leaky bucket parameters that would enable high-speed telecommunication network providers to optimize traffic policing subject to guaranteeing a negotiated Quality of Service requirement across the first stage namely the source end. We address both the buffered and unbuffered leaky bucket cases where using fluid models we characterize the output process from the leaky buckets for general traffic sources. Using the optimal leaky bucket parameters and output characteristics (effective bandwidths in particular), we solve design and connection admission control problems given QoS requirements at the second stage, namely the border node.

Buffered Leaky Bucket Algorithm for Congestion Control in ATM Networks

Buffered Leaky Bucket algorithm, a simple and efficient scheme to regulate the information sources to the agreed rate at the admission time in ATM networks, allowing a certain degree of source burstiness is presented in this paper. Discrete-time simulation study is presented to evaluate the performance of the Buffered Leaky Bucket algorithm for the critical ON/OFF bursty traffic. The flow enforcement action with traffic smoothing effects is evaluated with the help of ATM traffic modeling and the simulation of traffic scenario at congestion control device. The simulation results for waiting time in finite input queue case are presented. Assuming Poisson cell arrival process, Cell Loss Probability evaluation in the context of bit rate variation and mean burst duration variation is carried out. The results are also compared with Leaky Bucket algorithm. The influence of traffic burstiness on the congestion control device selection is analysed and the numerical results are presented.

A framework for guaranteeing statistical QoS

Continuous-media traffic (i.e., audio and video) can tolerate some loss but have rigid delay constraints. A natural QoS requirement for a continuous-media connection is a prescribed limit on the fraction of traffic that exceeds an end-to-end delay constraint. We propose and analyze a framework that provides such a statistical QoS guarantee to traffic in a packet-switched network. Providing statistical guarantees in a network is a notoriously difficult problem because traffic flows lose their original statistical characterizations at the outputs of queues. Our scheme uses bufferless statistical multiplexing combined with cascaded leaky buckets for smoothing and traffic contracting. This scheme along with a novel method for bounding the loss probability gives a tractable framework for providing end-to-end statistical QoS. Using MPEG video traces, we present numerical results that compare the connection-carrying capacity of our scheme with that of guaranteed service schemes (i.e., no loss) using GPS and RCS. Our numerical work indicates that our scheme can support significantly more connections without introducing significant traffic loss.

Packet multiplexers with adversarial regulated traffic

We consider a finite-buffer packet multiplexer to which traffic arrives from several independent sources. The traffic from each of the sources is regulated, i.e. the amount of traffic that can enter the multiplexer is constrained by known regulator constraints. The regulator constraints depend on the source and are more general than those resulting from cascaded leaky buckets. We assume that the traffic is adversarial to the extent permitted by the regulators. For lossless multiplexing, we show that if the original multiplexer is lossless it is possible to allocate bandwidth and buffer to the sources so that the resulting segregated systems are lossless. For lossy multiplexing, we use our results for lossless multiplexing to estimate the loss probability of the multiplexer. Our estimate involves transforming the original system into two independent resource systems, and using adversarial sources for the two independent resources to obtain a bound on the loss probability of the transformed system. We show that the adversarial sources are not extremal on–off sources, even when the regulator consists of a peak rate controller in series with a leaky bucket. We explicitly characterize the form of the adversarial source for the transformed problem. We also provide numerical results for the case of the simple regulator.

An optimal rate control algorithm for guaranteed services in broadband networks

Rate control is a known technique for preventing congestion in multi-service networks. Among the proposed control schemes, the connection bit rate can be modified in order to conform with a given “Leaky Bucket”. In this paper, we present a new algorithm for rate control that satisfies the Leaky Bucket constraints. In the proposed algorithm, the bit rate is controlled in order to minimize a cost function defined as a linear combination of two terms related to the difference between the original and the controlled bit rate and to the content of the token buffer, respectively. The obtained solution has some interesting mathematical properties. Moreover, a sub-optimal solution is proposed that is very close to the optimal one and is interesting for practical purposes. In fact it turns out to be a linear combination, with constant coefficients, of few samples of the original sequence. Numerical results confirm the effectiveness of the proposed method.

An Open Briefing for President Bush

Bad Presidents do nothing, mediocre Presidents do only what they're forced to do, good Presidents do what's possible, and great Presidents do what's necessary, Mr. Sutton avers. But we've not had any great Presidents for some time. Mr. Sutton's suggestions for George Bush are intended to help the new President break that pattern. MR. PRESIDENT, great opportunity awaits you. Decades ago, it took a Republican to reach out to China, and maybe today only a Republican can reach out to education. It's worth trying and worth doing. Your easiest task will be to avoid the mistakes of your predecessors. President Clinton loaded his education brain trust with university people - with predictable results. I prefer K-12 experts who have practical experience in schools. I'm funny that way. Like most of President Clinton's mistakes, this wasn't entirely his fault. Having university folk determine K-12 policy has been the rage since the 1930s, when the Carnegie Corporation imposed its will on schools. University folk structured teacher licensure to correspond exactly with colleges of liberal arts. It's still that way in 50 states. Such a practice maximizes credit hours in liberal arts, and university presidents are funny about the economics of credit hours. If you want to change things, try putting educators in charge of education policy; it's about the only thing that hasn't been tried. You've made a good start with your choice for secretary of education. Yes, Roderick Paige has a background in higher education, but he did go back to K-12. Don't waste him on vouchers. You and I both know that this issue serves mostly to keep ideologues happy - until you've tamed them. He's too good to waste carrying water in a leaky bucket. Give him something real; you won't regret it. The problem has never been managing professionals, but unleashing them. Here's another Clinton mistake you can avoid: he relied on Arkansas to set national policy. It didn't occur to him that what worked in his home state wouldn't necessarily work in states with high student achievement. These states already have adequacy and are poised to achieve excellence. Exporting solutions from low-achieving states lowers performance in high-achieving states. Forget what you learned in Texas. It won't transfer to Iowa, Wisconsin, Minnesota, Illinois, or Connecticut. Avoid bullying teachers. This has been going on since the publication of A Nation at Risk in 1983. Political fund raisers needed an enemy, so they created As experts in marketing, they knew that the only way to bring down one was to convince people that something was wrong with one. No sense trying to convince people that two is as good. But just as Hertz is still number one, people still believe that public education is best. The bullies' only achievement has been the Lake Wobegon Effect: Everyone is fine in my school, thank you, but the school down the road is a basket case. You can confirm this by reading the annual Phi Delta Kappa/Gallup polls. What's more, beating up teachers hasn't lowered public esteem for teachers. People still rate teaching at the top of the list of respected careers. Besides, the line between conservatism and fascism is a fine one: a conservative who bullies or scapegoats becomes a fascist. Here's something to bear in mind: teachers are the only major political group that goes both ways - one-third Democrat, one-third Republican, one-third Independent. It's pretty dumb politically to alienate this swing vote; you may need it next time. The political dynamic is simple: teachers consistently endorse the more moderate candidate because support for education increases as candidates approach the center of the political spectrum. (You can't expect teachers to support those who don't support them.) Indeed, moderation is a professional habit for teachers. They work to get students to see both sides of an issue and to help them make up their own minds. …

Some properties of variable length packet shapers

The min-plus theory of greedy shapers has been developed after Cruz's results on the calculus of network delays. An example of greedy shaper is the buffered leaky bucket controller. The theory of greedy shapers establishes a number of properties; for example, re-shaping keeps original arrival constraints. The existing theory applies in all rigor either to fluid systems, or to packets of constant size such as ATM. For variable length packets, the distortion introduced by packetization affects the theory, which is no longer valid. Chang has introduced the concept of packetizer, which models the effect of variable length packets, and has also developed a max-plus theory of shapers. In this paper, we start with the min-plus theory, and obtain results on greedy shapers for variable length packets which are not readily explained with the max-plus theory of Chang. We show a fundamental result, namely, the min-plus representation of a packetized greedy shaper. This allows us to prove that, under some assumptions, re-shaping a flow of variable length packets does keep original arrival constraints. However, we show on some examples that if the assumptions are not satisfied, then the property may not hold any more. We also demonstrate the equivalence of implementing a buffered leaky bucket controller based on either virtual finish times or on bucket replenishment.

Broadcast quality video over IP

We consider the problem of designing systems for the transmission of video signals of the quality found in current television broadcasts, over high-speed segments of the public IP network. Our most important contribution is the definition of a network/coder interface for IP networks which gathers channel state information, and then sets parameters of the video coder to maximize the quality of the signal delivered to the receiver, while remaining fair to other data or video connections. This interface plays a role analogous to that of a Leaky Bucket controller, in that it specifies traffic shaping parameters which result in simultaneous good Quality-of-Service (QoS) for the source and good network performance. Since the network is not assumed to provide any form of QoS guarantee, fundamental to our construction is a hidden Markov model for the channel, based on which the interface solves a problem of optimal stochastic control, to decide how to configure the encoder. Other contributions are: a) modifications to the standard Internet transport protocol, to make it suitable for the transport of delay-constrained traffic and to gather channel state information, and b) the design of an error-resilient video coder. Experimental studies reveal that the proposed system is able to stream video signals of the quality of current TV-broadcasts, among hosts in wide-area networks connected to the experimental vBNS backbone.

The transmission of MPEG‐2 VBR video under usage parameter control

AbstractThis paper studies the transmission of MPEG‐2 VBR video over ATM network under usage parameter control. The idea is to seek a compromise between the network utilization and the quality of video service by applying UPC‐based rate control strategies to the video source. A modified leaky bucket algorithm is proposed to calculate the constraints on the bit‐rate guaranteeing conformance to peak cell rate, sustainable cell rate and burst tolerance usage parameters. Two rate control strategies, one for real‐time generated video coding and the other for pre‐compressed video, are proposed for MPEG‐2 VBR video. The rate control strategies control the video source to generate traffic conforming to the constraints on the bit rate. The experimental results show that both the UPC‐based rate control strategies can provide lossless transmission from the source perspective as well as to reduce the burstiness of the traffic. To keep within the bit‐rate allowed, the control method uses coarser quantization to maintain better picture quality than that by removing the number of AC transformed coefficients. The slight degradation of picture quality caused by the source rate control is preferable than the severe drop of picture quality caused by the cell loss at UPC. Copyright © 2001 John Wiley & Sons, Ltd.

Rate control for VBR video over ATM: simplification and implementation

Rate control is essential to ensure high-quality video transmission over asynchronous transfer mode (ATM) networks; both source and channel rates must not violate buffer or network traffic constraints in order to guarantee decoded video quality. To date, source and channel rate selection has been performed jointly because the rates are related through buffer and network constraints and therefore appear to be interdependent. It is shown that selection of these rates can be separated and the rate control problem is therefore simplified, allowing sequential source and channel rate selection without the need for iteration. A rate control algorithm is proposed which implements noniterative, separate sequential selection of source and channel rates in order to minimize the distortion variation between frames, subject to all buffer and network constraints. In particular, the leaky bucket is used as the traffic policing mechanism. Achieving a target rate for individual frames is facilitated by using a rate-distortion optimized video coder which allows coding to an exactly specified rate or distortion. Channel rate selection is performed following source rate selection with the goal of providing smooth traffic to the network, thereby allowing a high statistical multiplexing gain. Experiments demonstrate that the proposed noniterative rate control algorithm, when implemented using the optimized coder, achieves consistent video quality while providing smooth traffic to the network.