Papers by Gagan Choudhury
Journal of Optical Communications and Networking, 2019
Recently, Internet service providers (ISPs) have gained increased flexibility in how they configu... more Recently, Internet service providers (ISPs) have gained increased flexibility in how they configure their in-ground optical fiber into an IP network. This greater control has been made possible by improvements in optical switching technology, along with advances in software control. Traditionally, at network design time, each IP link was assigned a fixed optical path and bandwidth. Now modern colorless and directionless reconfigurable optical add/drop multiplexers (CD ROADMs) allow a remote controller to remap the IP topology to the optical underlay on the fly. Consequently, ISPs face new opportunities and challenges in the design and operation of their backbone networks [IEEE Commun. Mag. 54, 129 (2016); presentation at the International Conference on Computing, Networking, and Communications, 2017; J. Opt. Commun. Netw. 10, D52 (2018); Optical Fiber Communication Conference and Exposition (2018), paper Tu3H.2]. Specifically, ISPs must determine how best to design their networks to take advantage of new capabilities; they need an automated way to generate the least expensive network design that still delivers all offered traffic, even in the presence of equipment failures. This problem is difficult because of the physical constraints governing the placement of optical regenerators, a piece of optical equipment necessary to maintain an optical signal over long stretches of fiber. As a solution, we present an integer linear program (ILP) that does three specific things: It solves the equipment placement problem in network design; determines the optimal mapping of IP links to the optical infrastructure for any given failure scenario; and determines how best to route the offered traffic over the IP topology. To scale to larger networks, we also describe an efficient heuristic that finds nearly optimal network designs in a fraction of the time. Further, in our experiments our ILP offers cost savings of up to 29% compared to traditional network design techniques.
Proceedings of INFOCOM'95
ABSTRACT We extend our recently developed algorithm for computing (exact) steady-state blocking p... more ABSTRACT We extend our recently developed algorithm for computing (exact) steady-state blocking probabilities for each class in product-form loss networks to cover general state-dependent arrival and service rates. This generalization allows us to consider, for the first time, a wide variety of buffered and unbuffered resource-sharing models with non-Poisson traffic as may arise with overflows in the context of alternative routing. As before, we consider non-complete-sharing policies involving upper-limit and guaranteed-minimum bounds for the different classes, but here we consider both bounds simultaneously. Major features of the algorithm are: dimension reduction by conditional decomposition based on special structure, an effective scaling algorithm to control errors in the inversion, the efficient treatments of multiple classes with identical parameters and the truncation of large sums
IEEE Transactions on Communications, 1985
... Some prioritized CSMA-CD schemes have been reported in the ... of packet length and also a ge... more ... Some prioritized CSMA-CD schemes have been reported in the ... of packet length and also a general packet length distribution within each priority class. As noted in [lo], the one-persistent mode is superior to the p-persistent mode in the presence of quick collision detection. ...
An approximate analysis of a flow control technique for improving stability and delay characteris... more An approximate analysis of a flow control technique for improving stability and delay characteristics of multiple access schemes employing collision type channels is presented. The technique allows users to enter two retransmission modes with differing retransmission rates. The condition under which such systems exhibit absolute stability and the minimum delay compatible with that condition have been determined. It is found that a significant improvement in delay is possible without having to sacrifice absolute stability.
Operations Research, 1993
The Fundamental Role of Teletraffic in the Evolution of Telecommunications Networks, 1994
Proceedings of GLOBECOM '93. IEEE Global Telecommunications Conference
We present the two-dimensional transforms of the transient workload and queue-length distribution... more We present the two-dimensional transforms of the transient workload and queue-length distributions in the single-server queue with general service times and a batch Markovian arrival process (BMAP). This arrival process includes the familiar phase-type renewal process and the Markov modulated Poisson process as special cases, as well as superpositions of these processes, and allows correlated interarrival times and batch sizes. Numerical results are obtained via two-dimensional transform inversion algorithms based on the Fourier-series method. From the numerical examples we see that predictions of system performance based on transient and stationary performance measures can be quite different.
We derive the two-dimensional transforms of the transient workload and queue-length distributions... more We derive the two-dimensional transforms of the transient workload and queue-length distributions in the single-server queue with general service times and a batch Markovian arrival process (BMAP). This arrival process includes the familiar phase-type renewal process and the Markov modulated Poisson process as special cases, and allows correlated interarrival times and batch sizes. Numerical results are obtained via two-dimensional transform inversion algorithms based on the Fourier-series method. From the numerical examples we see that predictions of system performance based on transient and stationary performance measures can be quite different.
ACM SIGCOMM Computer Communication Review, 1981
Demand Assigned Multiple Access (DAMA) schemes in which requests for communication channels are t... more Demand Assigned Multiple Access (DAMA) schemes in which requests for communication channels are transmitted over collision type channels to a master controller and in which messages of different priorities are present is analyzed. Priority use of the request channels is achieved by having users select among those channels according to a priority dependent probability distribution. Two important schemes of this class are considered. Additionally, three schemes for priority assignment of message channels are treated; two are non-preemptive and one is preemptive. Trade offs among bandwidth utilization and delay are considered as well as optimization of system parameters.
2009 7th International Workshop on Design of Reliable Communication Networks, 2009
The design of an IP link topology (i.e., deciding which set of router-to-router communication lin... more The design of an IP link topology (i.e., deciding which set of router-to-router communication links to establish) is an important factor in determining both the cost and performance of an IP network. In an IP-over-WDM architecture, these routerto-router links consist of wavelengths carried over an optical network. Each physical link in the optical network can carry multiple router-to-router wavelengths. The IP network must be designed for survivability in the event of an optical network failure (i.e., the loss of any given physical link or optical switch). Therefore, if there are particular physical links that carry an inordinate number of router-to-router wavelengths, a substantial amount of capacity may be required elsewhere in the IP network in order to assure survivability. In this paper, we describe a heuristic for designing an IP link topology in an IP-over-WDM architecture. In choosing which IP links to establish, the heuristic explicitly considers which routerto-router wavelengths that would be carried over each physical link and the various failure scenarios that could result. It is particularly targeted toward large, sparse networks. Some computational comparisons are included.
We show that probability distributions of performance measures in a large class of polling models... more We show that probability distributions of performance measures in a large class of polling models can be effectively computed by numerically inverting transforms (generating functions and Laplace transforms). We develop new efficient iterative algorithms for computing the transform values and then use the Fourier-series method to perform the inversion. We also use this approach to compute moments. The algorithms apply to the transient behavior of stationary or nonstationary models as well as to the steady-state behavior of stationary models. Our main focus is on computing exact tail probabilities, but even for mean waiting times, our algorithm is faster than previous algorithms for large models. The computational complexity of our algorithm is O(Nα) for computing performance measures at one queue and O(N1+α) for computing performance measures at all queues, where N is the number of queues and α is typically between 0.6 and 0.8. We demonstrate effectiveness by computing performance m...
We propose a new algorithm for closed queueing networks and related product-form models based on ... more We propose a new algorithm for closed queueing networks and related product-form models based on numerical inversion of the generating function of the normalization constant (or partition function). It is known that the generating function of the normalization constant often has a remarkably simple form, but numerical inversion evidently has not been considered before. We also show that moments of steady-state distributions can be calculated directly by only performing two inversions. For closed queueing networks with p closed chains, the generating function is p dimensional. For these generating functions, the algorithm recursively performs p one-dimensional inversions. The required computation grows exponentially in the dimension, but we show that the dimension can often be reduced dramatically by exploiting special structure. Other key ingredients in the algorithm are scaling and the computation of large sums efficiently by Euler summation. Numerical examples indicate that this n...
Archiv für Elektronik und …, 1993
Communications in Statistics. Stochastic Models, 1994
In great generality, the basic steady-state distributions in the BMAP / G /1 queue have asymptoti... more In great generality, the basic steady-state distributions in the BMAP / G /1 queue have asymptotically exponential tails. Here we develop asymptotic expansions for the asymptotic decay rates of these tail probabilities in powers of one minus the traffic intensity. The first term coincides with the decay rate of the exponential distribution arising in the standard heavy-traffic limit. The coefficients of these heavy-traffic expansions depend on the moments of the service-time distribution and the derivatives of the Perron-Frobenius eigenvalue δ(z) of the BMAP matrix generating function D(z) at z = 1. We give recursive formulas for the derivatives δ (k) (1). The asymptotic expansions provide the basis for efficiently computing the asymptotic decay rates as functions of the traffic intensity, i.e., the caudal characteristic curves. The asymptotic expansions also reveal what features of the model the asymptotic decay rates primarily depend upon. In particular, δ(z) coincides with the limiting time-average of the factorial cumulant generating function (the logarithm of the generating function) of the arrival counting process, and the derivatives δ (k) (1) coincide with the asymptotic factorial cumulants of the arrival counting process. This insight is important for admission control schemes in multi-service networks based in part on asymptotic decay rates. The interpretation helps identify appropriate statistics to compute from network traffic data in order to predict performance.
Communications in Statistics. Stochastic Models, 1994
Uploads
Papers by Gagan Choudhury