Queues and network computers.

Bibliography: p. 68-71.

Loss fluctuations and temporal correlations in network queues

We consider data losses in a single node of a packet- switched Internet-like network. We employ two distinct models, one with discrete and the other with continuous one-dimensional random walks, representing the state of a queue in a router. Both models have a built-in critical behavior with a sharp transition from exponentially small to finite losses. It turns out that the finite capacity of a buffer and the packet-dropping procedure give rise to specific boundary conditions which lead to strong loss rate fluctuations at the critical point even in the absence of such fluctuations in the data arrival process.

Algorithmic Methods in Queues and in the Exploration of Point Processes

This is a review of methodology for the algorithmic study of some useful models in point process and queueing theory, as discussed in three lectures at the Summer Institute at Sozopol, Bulgaria. We provide references to sources where the extensive details of this work are found. For future investigation, some open problems and new methodological approaches are proposed.

On Multi- Server Queues With Consultation With Main Server

Queueing Theory is the mathematical study of queues or waiting lines. Queues abound in every day life - in computer networks, in tra c islands, in communication of electro-magnetic signals, in telephone exchange, in bank counters, in super market checkouts, in doctor's clinics, in petrol pumps, in o ces where paper works to be processed and many other places. Originated with the published work of A. K. Erlang in 1909 [16] on congestion in telephone tra c, Queueing Theory has grown tremendously in a century. Its wide range applications includes Operations Research, Computer Science, Telecommunications, Tra c Engineering, Reliability Theory, etc.

Some heuristic algorithms for total tardiness minimization in a flowshop with blocking

The flowshop scheduling problem with blocking in-process is addressed in this paper. In this environment, there are no buffers between successive machines: therefore intermediate queues of jobs waiting in the system for their next operations are not allowed. Heuristic approaches are proposed to minimize the total tardiness criterion. A constructive heuristic that explores specific characteristics of the problem is presented. Moreover, a GRASP-based heuristic is proposed and Coupled with a path relinking strategy to search for better outcomes. Computational tests are presented and the comparisons made with an adaptation of the NEH algorithm and with a branch-and-bound algorithm indicate that the new approaches are promising. (c) 2007 Elsevier Ltd. All rights reserved.

Spectral properties of the tandem Jackson network, seen as a quasi-birth-and-death process

Quasi-birth-and-death (QBD) processes with infinite “phase spaces” can exhibit unusual and interesting behavior. One of the simplest examples of such a process is the two-node tandem Jackson network, with the “phase” giving the state of the first queue and the “level” giving the state of the second queue. In this paper, we undertake an extensive analysis of the properties of this QBD. In particular, we investigate the spectral properties of Neuts’s R-matrix and show that the decay rate of the stationary distribution of the “level” process is not always equal to the convergence norm of R. In fact, we show that we can obtain any decay rate from a certain range by controlling only the transition structure at level zero, which is independent of R. We also consider the sequence of tandem queues that is constructed by restricting the waiting room of the first queue to some finite capacity, and then allowing this capacity to increase to infinity. We show that the decay rates for the finite truncations converge to a value, which is not necessarily the decay rate in the infinite waiting room case. Finally, we show that the probability that the process hits level n before level 0 given that it starts in level 1 decays at a rate which is not necessarily the same as the decay rate for the stationary distribution.

On the importance function in splitting simulation

The splitting method is a simulation technique for the estimation of very small probabilities. In this technique, the sample paths are split into multiple copies, at various stages in the simulation. Of vital importance to the efficiency of the method is the Importance Function (IF). This function governs the placement of the thresholds or surfaces at which the paths are split. We derive a characterisation of the optimal IF and show that for multi-dimensional models the natural choice for the IF is usually not optimal. We also show how nearly optimal splitting surfaces can be derived or simulated using reverse time analysis. Our numerical experiments illustrate that by using the optimal IF, one can obtain a significant improvement in simulation efficiency.

Solving a signalized traffic intersection problem with an hyperbolic penalty function

Mathematical Program with Complementarity Constraints (MPCC) finds many applications in fields such as engineering design, economic equilibrium and mathematical programming theory itself. A queueing system model resulting from a single signalized intersection regulated by pre-timed control in traffic network is considered. The model is formulated as an MPCC problem. A MATLAB implementation based on an hyperbolic penalty function is used to solve this practical problem, computing the total average waiting time of the vehicles in all queues and the green split allocation. The problem was codified in AMPL.

Geração de simuladores de filas para diferentes contextos com estudo de casos para redes de computadores

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Capacity sharing and stealing in serverbased real-time systems

A dynamic scheduler that supports the coexistence of guaranteed and non-guaranteed bandwidth servers is proposed. Overloads are handled by an efficient reclaiming of residual capacities originated by early completions as well as by allowing reserved capacity stealing of non-guaranteed bandwidth servers. The proposed dynamic budget accounting mechanism ensures that at a particular time the currently executing server is using a residual capacity, its own capacity or is stealing some reserved capacity, eliminating the need of additional server states or unbounded queues. The server to which the budget accounting is going to be performed is dynamically determined at the time instant when a capacity is needed. This paper describes and evaluates the proposed scheduling algorithm, showing that it can efficiently reduce the mean tardiness of periodic jobs. The achieved results become even more significant when tasks’ computation times have a large variance.

On the adaptation of broadcast transactions in token-passing fieldbus networks with heterogeneous transmission media

Broadcast networks that are characterised by having different physical layers (PhL) demand some kind of traffic adaptation between segments, in order to avoid traffic congestion in linking devices. In many LANs, this problem is solved by the actual linking devices, which use some kind of flow control mechanism that either tell transmitting stations to pause (the transmission) or just discard frames. In this paper, we address the case of token-passing fieldbus networks operating in a broadcast fashion and involving message transactions over heterogeneous (wired or wireless) physical layers. For the addressed case, real-time and reliability requirements demand a different solution to the traffic adaptation problem. Our approach relies on the insertion of an appropriate idle time before a station issuing a request frame. In this way, we guarantee that the linking devices’ queues do not increase in a way that the timeliness properties of the overall system turn out to be unsuitable for the targeted applications.

Supporting real-time parallel task models with work-stealing

Dynamic parallel scheduling using work-stealing has gained popularity in academia and industry for its good performance, ease of implementation and theoretical bounds on space and time. Cores treat their own double-ended queues (deques) as a stack, pushing and popping threads from the bottom, but treat the deque of another randomly selected busy core as a queue, stealing threads only from the top, whenever they are idle. However, this standard approach cannot be directly applied to real-time systems, where the importance of parallelising tasks is increasing due to the limitations of multiprocessor scheduling theory regarding parallelism. Using one deque per core is obviously a source of priority inversion since high priority tasks may eventually be enqueued after lower priority tasks, possibly leading to deadline misses as in this case the lower priority tasks are the candidates when a stealing operation occurs. Our proposal is to replace the single non-priority deque of work-stealing with ordered per-processor priority deques of ready threads. The scheduling algorithm starts with a single deque per-core, but unlike traditional work-stealing, the total number of deques in the system may now exceed the number of processors. Instead of stealing randomly, cores steal from the highest priority deque.

Schedulability analysis of generalized multiframe traffic on multihop-networks comprising software-implemented ethernet switches

Consider a multihop network comprising Ethernet switches. The traffic is described with flows and each flow is characterized by its source node, its destination node, its route and parameters in the generalized multiframe model. Output queues on Ethernet switches are scheduled by static-priority scheduling and tasks executing on the processor in an Ethernet switch are scheduled by stride scheduling. We present schedulability analysis for this setting.