S. Q. Li and J. D. Pruneski, "The linearity of low frequency traffic flow: An intrinsic I/O property in queueing system," IEEE/ACM Trans. Networking, vol. 5, pp. 429--443, June 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... is packetized into ATM cells introducing AAL5 overhead (one AAL5 PDU made of 8 ATM cells) The cells are then evenly spaced over a time unit period; this, however, is not critical since it is well known that the high frequencies in the power spectrum are negligible in terms of queueing behaviour [5]. The number of generated ATM cells for each trace is approximately 10 8 , with a mean of 595 cells time unit. Cells are then processed into the queue with deterministic service time, FIFO mode. The outgoing data frequency is varied so as to consider a load ranging from 0.1 to 0.95, with step ....

S. Li and J. D. Pruneski, "The linearity of low frequency traffic flow: an intrinsic property in queueing system", IEEE/ACM Trans. Networking, vol. 5, n. 3, pp. 429-443, 1997

.... to match the video traces [3] 17] 21] Recent traffic measurement [11] also identifies the significance of long range dependencies, which are described by the correlation behavior in large time scales (or, equivalently, by the dominant power spectrum in low frequency band) In fact, 12] [16], and [7] indicate that a finite buffer queueing system is noneffective for the transport of low frequency traffic subject to negligible loss rate. That is, the link bandwidth should be at least equal to the peak rate of low frequency traffic whose flow stays intact through the queueing system. We ....

S. Q. Li and J. D. Pruneski, "The linearity of low frequency traffic flow: An intrinsic I/O property in queueing system," IEEE/ACM Trans. Networking, vol. 5, pp. 429--443, June 1997.

....analysis. The traffic prediction issue is illustrated in Fig. 1. The figure depicts the measurement based control procedure for a network node with a finite buffer capacity. The queueing process of the network node can be characterized by the low pass filter (LPF) with cutoff frequency 1=T c [14]. Only the high frequency dynamics of the input traffic are affected by the queueing process, whereas the low frequency dynamics remain unchanged at the output. Such a T c can be used as the sampling smoothing interval for online traffic measurement. We assume that it is specified by the region ....

J. D. Pruneski and S. Q. Li, The Linearity of Low Frequency Traffic Flow: An Intrinsic I/O Property in Queueing Systems, Proc. IEEE Infocom'95 Conference, April 1995, pp.613-623.

....varying scale, high correlation or long range dependence. Intuitively, one may characterize a queueing system by a nonlinear low pass filter, where the high frequency variation of the input traffic can be well absorbed into the buffering but its low frequency variation remains largely unchanged [16, 17]. For instance, the lowest frequency component of the traffic, i.e. the DC term, is its average arrival rate, which always stays intact through a queueing system unless with traffic loss by buffer blocking. Moreover, such low frequency behavior of the traffic will remain largely unchanged through ....

S. Q. Li and J. D. Pruneski, "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," IEEE/ACM Trans. Networking, Vol. 5, No. 3, July 1997, pp. 429-443.

....varying scale, high correlation or long range dependence. Intuitively, one may characterize a queueing system by a nonlinear low pass filter, where the high frequency variation of the input traffic can be well absorbed into the buffering, but its low frequency variation remains largely unchanged [12, 13]. For instance, the lowest frequency component of the traffic, i.e. the DC term, is its average arrival rate, which always stays intact through a queueing system unless with traffic loss by buffer blocking. Moreover, such low frequency behavior of the traffic will remain largely unchanged through ....

S. Q. Li and J. D. Pruneski, "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," IEEE/ACM Trans. Networking, Vol. 5, No. 3, July 1997, pp. 429-443.

....for the general construction of a(t) and s(t) In consequence, two state Markov chains are frequently used to construct different processes with limited statistic properties. Our work in the past several years focused on the development of fast algorithms for both modeling and queueing analysis [1, 2, 3, 4, 5, 6, 7, 8]. SMAQ tool (Statistical Match And Queueing tool) naturally grew out of this development for the integration of traffic service modeling and queueing analysis. A fundamental distinction of our work from others is that SMAQ tool is measurement based. In our view, both a(t) and s(t) can be composed ....

....by filtering in the frequency domain. y The time dynamic system performance, such as measured by the second order statistics of queue length and loss rate and the transient time statistics, can also be derived by the SMAQ tool, but it is beyond the scope of this article. One may refer to [7, 11, 13, 17] for further information. 4 To answer the first question, let us consider a finite buffer system in real networks. For buffer capacity K with constant bandwidth S, the maximum allowable queueing delay z is given by dmax = K=S. Such a dmax constrained queueing system is naturally described as ....

[Article contains additional citation context not shown here]

S. Q. Li and J. D. Pruneski, "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," IEEE/ACM Transactions on Networking, Vol. 5, No. 3, June, 1997, pp. 429-443.

No context found.

J. D. Pruneski and S. Q. Li "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," Proc. IEEE Infocom'95 Conference, April 1995, pp. 613-623.

....subject to zero loss, where conjecturs are made for the best LF sample path and the worst HF sample path. It is then verified in stochastic analysis with random traffic subject to a negligible cell loss rate, where we use a programming approach [17] for construction of circulant 1 The study in [10, 19, 23, 13, 25] also explored the concept of buffer noneffectiveness for transport of LF (large timescale) traffic, but none of them were able to quantitatively develop guidelines for transport of generic traffic subject to a maximum queueing delay. modulated Poisson process (CMPP) to match a wide range of P ....

....traffic, but none of them were able to quantitatively develop guidelines for transport of generic traffic subject to a maximum queueing delay. modulated Poisson process (CMPP) to match a wide range of P ( and f(x) in different frequrency regions. In relation to author s previous works [6, 7, 10, 15, 16, 17, 19], this paper focuses on the timescale decomposition in traffic measurement for link bandwidth allocation design. The paper is organized as follows. Section 2 provides the sample path deterministic analysis for the solution of ( L ; H ) along with the discussion of their significant implications ....

J. D. Pruneski and S. Q. Li, "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," Proc. IEEE Infocom'95, April 1995, pp. 613-623, also accepted by IEEE/ACM Trans. Networking.

No context found.

J. D. Pruneski and S. Q. Li "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," Proc. IEEE Infocom'95 Conference, April 1995, pp. 613-623.

....one important step to replace r(n) by r L (n) in the dynamic equation (1) x L (n 1) x L (n) r L (n) Gamma r L (n 1) N X k=1 b k (u k (n Gamma n k Gamma 1) Gamma u k (n Gamma n k ) 5) where r L (n) is the filtered r(n) in a properly selected low frequency band. Refer to [7, 10] for detail discussion of the cut off frequency selection for filtering. x L (n) therefore corresponds to the unused low frequency link capacity, defined by x L (n) C Gamma N X k=1 b k u k (n Gamma n k Gamma 1) Gamma fl L (n) where C represents the link capacity at node L. Similar to ....

J. D. Pruneski and S. Q. Li "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," Proc. IEEE Infocom'95 Conference, April 1995, pp. 613-623.

.... has been successfully applied to traffic rate control [21] to link capacity allocation and network control [15, 22] to design and analysis of buffer congestion control [19] to delay jitter correlation analysis of individual streams [23] and to input output low frequency linearity analysis [24]. Appendix Proof of Eq. 1) Define R a ( Efa(t)a(t )g and R( Effl(t)fl(t )g, where a(t) is an MMPP and fl(t) is the corresponding input rate process described in Section 2. a(t) is a Poisson point process with its rate driven by fl(t) When = 0, R a (0) Efa(t) 2 g = ....

J. D. Pruneski and S. Q. Li "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," Proc. IEEE Infocom'95 Conference, April 1995, pp. 613-623.

No context found.

J. D. Pruneski and S. Q. Li, "The Linearity of Low Frequency Traffic Flow: an Intrinsic I/O Property in Queueing System," Proc. IEEE Infocom'95, April 1995, pp. 613-623.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC