| S. Giordano and J.-Y. Le Boudec, On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service, Technical Report SSC/1998. |
No context found.
J. Y. Le Boudec S. Giordano. On a class of time varying shapers with application to the renegotiable variable bit rate service. Journal on High Speed Networks, 9(2):101--138, June 2000.
No context found.
J. Y. Le Boudec S. Giordano. On a class of time varying shapers with application to the renegotiable variable bit rate service. Journal on High Speed Networks, 9(2):101--138, June 2000.
No context found.
S. Giordano and J.-Y. Le Boudec, "On a class of time varying shapers with application to the renegotiable variable bit rate service," Journal on High Speed Networks, vol. 9, no. 2, pp. 101--138, June 2000.
....the status of the system at t i and the input function R(t) in I i . The result is a sequence of local optimal oe i . Alternately,we can minimise the cost of the global sequence of oe i given the complete input function R(t) The result is the optimal sequence of oe i . The latter, studied in [28], can be seen as a theoretical limit to the previous one and is not presented here. In the next section, as our first finding, wecharacterise a leaky bucket shaper system with non zero initial conditions in terms of input output functions. Second, we define the bucket level and the backlog for ....
....R (t) min[oe (t)# (oe Omega R) t) 8t 0 (5) where oe is the shaping function oe(u) min (u)g = min is definedas 0 g The proof, which is not given here, comes by applying to Corollary 2 the same min plus result as in Proposition 1. The formal proof is given in [28]. Finally we derive the characterisation of a leaky bucket shaper that starts with non zero initial conditions. Theorem 1 (Leaky Bucket Shaper with non zero initial conditions) with j = 1# 2 : J (leaky bucket shaper) Assume that the initial buffer level of the shaping buffer is given ....
[Article contains additional citation context not shown here]
S. Giordano, J.-Y. Le Boudec, "On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service," Technical Report SSC/1998.
....All computations so far were done with the assumption that the systems are empty at time 0. This is valid for static reservations, but not for dynamic reservations, which are supported by IntServ and ATM ABR. The modifications to the calculus presented above were found by Giordano et al. in [33]. Delay and Delay Jitter. For playback operations, only the variable part of delay, called delay jitter, is important (Section II D) In contrast, for interactive services, the total delay is also of importance. Thus, both delays must be accounted for; this can be done as follows. If the latency ....
S. Giordano and J.-Y. Le Boudec, "On a class of time varying shapers with application to the renegotiable variable bit rate service," Journal on High Speed Networks, vol. 9, no. 2, pp. 101--138, June 2000.
....possible only if requests for reservation fit as much as possible the effective resource occupation. It follows that applications must be enabled to directly manage the QoS in order to limit the resource lost. The introduction of the renegotiable variable bit rate (RVBR) service [Giordano, 1999] [Giordano,2000] at application layer is assumed to simplify and generalise this task. Whenever re negotiation is underway, the RVBR scheme generates a traffic specification conforming to the real demand to renegotiate the network resources in an optimal way while guaranteeing QoS to the traffic flows. The RVBR ....
....based on RSVP that integrates RVBR services. Armida provides MPEG4 streaming video over an IP network in Microsoft environment. The rest of the chapter is organised as follows. In the next section we provide a short overview of the RSVP protocol. Then we describe the RVBR mechanism as defined in [Giordano,2000]. In the fourth section we introduce the Armida application pointing out the component implementing the signalling protocol. Finally we provide a set of results related to a real case in which we compared the required bandwidth (derived from generated traffic) and the reserved QoS, varying the ....
[Article contains additional citation context not shown here]
S. Giordano, J.-Y. Le Boudec: "On a Class of Time Va .rying Shapers with Application to the Renegotiable Variable Bit Rate Service", Journal on High Speed Networks, 2000.
....optimisation is possible only if requests for reservation fit as much as possible the effective resource occupation. It follows that applications must be enabled to directly manage the QoS in order to limit the resource lost. The introduction of the renegotiable variable bit rate (RVBR) service [1], 2] at application layer is assumed to simplify and generalise this task. Whenever re negotiation is underway, the RVBR scheme generates a traffic specification conforming to the real demand to reallocate the network resources in an optimal way while guaranteeing QoS to the traffic flows. The ....
....perfectly the dynamics of the traffic generated by multimedia applications. Moreover it naturally integrates with the soft state mechanism of RSVP, which allows for resources renegotiating. We first recall the characterisation of the RVBR service in terms of input and output functions as given in [1]. Figure 1 RVBR reference configuration There is a renegotiable leaky bucket specification (with rate r and depth b) plus a fixed size buffer X drained at maximum at renegotiable peak rate p. The elements of a RVBR source, Figure 1, are a renegotiable leaky bucket specification (with rate r and ....
S. Giordano, J.-Y. Le Boudec: "On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service" , SSC Technical Report no. 98/035, 1998. 10
....traffic profile negotiation can lead to an excessive usage of resources and unacceptable performance. In these cases a straightforward solution is to renegotiate the traffic profile during the lifetime of the connection. The introduction of the the renegotiable variable bit rate (RVBR) service [1], 2] at application layer is assumed to simplify and generalise this task. Whenever renegotiation is taking place, the RVBR scheme generates the traffic specification that conforms to the real demand, in order to reallocate the network resources in an optimal way while guaranteeing QoS to the ....
.... ) the backlog in the shaping buffer at time t , the function b i is given by b i (s) max ( a i (s) a 0 i(s) w(t i ) q(t i ) and the minimum p by =max ( sup i (s ) X) s , sup (s ) X w(t ) we can use for our prototype the algorthm for RVBR as defined in [1 ] and [2] a i and a can be used in a real implementation, because computed with only four parameters: p , r , b , p ) These parameters can be easily stored and passed from the application level to the RVBR module 2. Prototype description In this section we present the design ....
S. Giordano, J.-Y. Le Boudec: "On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service" , SSC Technical Report no. 98/035, 1998
....All computations so far were done with the assumption that the systems are empty at time #. This is valid for static reservations, but not for dynamic reservations, which are supported by IntServ and ATM ABR. The modifications to the calculus presented above were found by Giordano et al. in [30]. Delay and Delay Jitter. For playback operations, only the variable part of delay, called delay jitter, is important (Section II D) In contrast, for interactive services, the total delay is also of importance. Thus, both delays must be accounted for; this can be done as follows. If the latency ....
S. Giordano and J.-Y. Le Boudec, "On a class of time varying shapers with application to the renegotiable variable bit rate service," Journal on High Speed Networks, vol. 9, no. 2, pp. 101--138, June 2000.
....able to achieve a statistical multiplexing gain on many such input flows [3] In our model scenario, the RVBR parameters are renegotiated periodically; at every renegotiation, there is a tradeoff to be made between the various parameters, which define the two leaky buckets in the next interval. In [4] we analyse this tradeoff and propose an algorithm (localOptimum1) to select, for the next interval, the parameters that minimise a given linear cost function. Our main goal in this paper is to validate our service by means of simulations and to prove its applicability to real scenarios through ....
.... leaky bucket is constant, the system is identical to the ordinary, time invariant, leaky bucket shaper [9] 8] Moreover, it is in line with the Dynamic Generic Cell Rate Algorithm (DGCRA) used to specify conformance at the UNI for the available bit rate (ABR) service of ATM [18] 19] In [4] the practical implications of the no reset approach are studied in terms of losses. The input output characterisation of the time varying leaky bucket shapers in the interval I i is given in Theorem 2 of [4] R (t) min oe 0 i (t Gamma t i ) R (t i ) inf t i st foe i (t ....
[Article contains additional citation context not shown here]
S. Giordano, J.-Y. Le Boudec, "On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service," Technical Report SSC/1998/035, DI-EPFL, CH-1015 Lausanne, Switzerland, 1998. http://lrcwww.epfl.ch/ ~giordano/tvShaperv33.ps.
....the status of the system at t i and the input function R(t) in I i . The result is a sequence of local optimal oe i . Alternately, we can minimise the cost of the global sequence of oe i given the complete input function R(t) The result is the optimal sequence of oe i . The latter, studied in [28], can be seen as a theoretical limit to the previous one and is not presented here. In the next section, as our first finding, we characterise a leaky bucket shaper system with non zero initial conditions in terms of input output functions. Second, we define the bucket level and the backlog for ....
....foe j (u)g = min 1jJ fr j Delta u b j g and oe 0 is defined as oe 0 (u) min 1jJ fr j Delta u b j Gamma q j 0 g Proof: The proof, which is not given here, comes by applying to Corollary 2 the same min plus result as in Proposition 1. The formal proof is given in [28]. 2 Finally we derive the characterisation of a leaky bucket shaper that starts with non zero initial conditions. Theorem 1 (Leaky Bucket Shaper with non zero initial conditions) Consider a shaper system defined by J leaky buckets (r j ; b j ) with j = 1; 2 : J (leaky bucket shaper) ....
[Article contains additional citation context not shown here]
S. Giordano, J.-Y. Le Boudec, "On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service," Technical Report SSC/1998/035, DI-EPFL, CH-1015 Lausanne, Switzerland, 1998. http://lrcwww.epfl.ch/~giordano/publications.html.
No context found.
S. Giordano and J.-Y. Le Boudec, On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service, Technical Report SSC/1998.
No context found.
S. Giordano and J.-Y. Le Boudec, On a Class of Time Varying Shapers with Application to the Renegotiable Variable Bit Rate Service, Technical Report SSC/1998.
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