Samuli Aalto

Beyond processor sharing

While the (Egalitarian) Processor-Sharing (PS) discipline offers crucial insights in the performance of fair resource allocation mechanisms, it is inherently limited in analyzing and designing differentiated scheduling algorithms such as Weighted Fair Queueing and Weighted Round-Robin. The Discriminatory Processor-Sharing (DPS) and Generalized Processor-Sharing (GPS) disciplines have emerged as natural generalizations for modeling the performance of such service differentiation mechanisms. A further extension of the ordinary PS policy is the Multilevel Processor-Sharing (MLPS) discipline, which has captured a pivotal role in the analysis, design and implementation of size-based scheduling strategies. We review various key results for DPS, GPS and MLPS models, highlighting to what extent these disciplines inherit desirable properties from ordinary PS or are capable of delivering service differentiation.

Analysis of PDCCH performance for M2M traffic in LTE

As Long-Term Evolution (LTE) is starting to be widely deployed, the volume of machine-to-machine (M2M) traffic is increasing very rapidly. From the M2M traffic point of view, one of the issues to be addressed is the overload of the random access channel. The limitation in the physical downlink control channel (PDCCH) resources may severely constrain the number of devices that an LTE Evolved Node B (eNB) can serve. We develop a Markov model that describes the evolution of the Message 4 queue in the eNB formed by several users performing the random access procedure simultaneously, and then, we study its stability and performance. Our model explicitly takes into account the four initial steps in the random access procedure. By utilizing the model, we are able to determine the stability limit of the system, which defines the maximum throughput and the probability of failure of the random access procedure due to different causes. We observe that the sharing of the PDCCH resources between Messages 2 and 4 with different priorities makes the performance of the whole random access procedure deteriorate very rapidly near the stability limit. However, we can extend the maximum throughput and improve the overall performance by increasing the PDCCH resource size. Furthermore, we estimate the upper limit of the number of devices that can be served by an LTE eNB and determine the minimum PDCCH resource size needed to satisfy a given traffic demand.

Size- and State-Aware Dispatching Problem with Queue-Specific Job Sizes

We consider the dispatching problem in a size- and state-aware multi-queue system with Poisson arrivals and queue-specific job sizes. By size- and state-awareness, we mean that the dispatcher knows the size of an arriving job and the remaining service times of the jobs in each queue. By queue-specific job sizes, we mean that the time to process a job may depend on the chosen server. We focus on minimizing the mean sojourn time (i.e., response time) by an MDP approach. First we derive the so-called size-aware relative values of states with respect to the sojourn time in an M/G/1 queue operating under FIFO, LIFO, SPT or SRPT disciplines. For FIFO and LIFO, the size-aware relative values turn out to be insensitive to the form of the job size distribution. The relative values are then exploited in developing efficient dispatching rules in the spirit of the first policy iteration.

Two-level processor-sharing scheduling disciplines: mean delay analysis

Inspired by several recent papers that focus on scheduling disciplines for network flows, we present a mean delay analysis of Multilevel Processor Sharing (MLPS) scheduling disciplines in the context of M/G/1 queues. Such disciplines have been proposed to model the effect of the differentiation between short and long TCP flows in the Internet. Under MLPS, jobs are classified into classes depending on their attained service. We consider scheduling disciplines where jobs within the same class are served either with Processor Sharing (PS) or Foreground Background (FB) policy, and the class that contains jobs with the smallest attained service is served first. It is known that the FB policy minimizes (maximizes) the mean delay when the hazard rate of the job size distribution is decreasing (increasing). Our analysis, based on pathwise and meanwise arguments of the unfinished truncated work, shows that Two-Level Processor Sharing (TLPS) disciplines, e.g., FB+PS and PS+PS, are better than PS scheduling when the hazard rate of the job size distribution is decreasing. If the hazard rate is increasing and bounded, we show that PS outperforms PS+PS and FB+PS. We further extend our analysis to study local optimality within a level of an MLPS scheduling discipline.

Blocking of dynamic multicast connections in a single link

In this paper, a method for calculating blocking experienced by dynamic multicast connections in a single link is presented. A service center at the root of a tree-type network provides a number of channels distributed to the users by multicast trees which evolve dynamically as users join and leave the channels. We reduce this problem to a generalized Engset system with nonidentical users and generally distributed holding times, and derive the call and channel blocking probabilities as well as the link occupancy distribution.

Flow-level stability and performance of channel-aware priority-based schedulers

Channel-aware scheduling in modern wireless networks enables the system to exploit the random rate variations across different users to increase the performance of the system. We analyze channel-aware priority-based downlink scheduling policies at the so-called flow level with a stochastically varying number of users. The priority can be any monotonously increasing function of the instantaneous rate of the user, which generalizes the well-known linear weight-based policies. Also, ties are allowed within a user class, as well as between user classes. As the main result, we characterize when these priority-based policies are stable under an intuitive necessary condition, which holds for arbitrary tie breaking rules and is independent of the flow size distribution. Additionally, for the policies for which the necessary condition is not sufficient, a more stringent condition is derived in the case of two traffic classes. Finally, extensive simulations have been performed to compare the performance of different priority-based and utility-based policies.

Task Assignment in a Heterogeneous Server Farm with Switching Delays and General Energy-Aware Cost Structure

Abstract We consider the task assignment problem to heterogeneous parallel servers with switching delay, where servers can be switched off to save energy. However, switching a server back on involves a constant server-specific delay. We will use one step of policy iteration from a starting policy such as Bernoulli splitting, in order to derive efficient task assignment (dispatching) policies that minimize the long-run average cost. To evaluate our starting policy, we first analyze a single work-conserving M / G / 1 queue with a switching delay and derive a value function with respect to a general cost structure. Our costs include energy related switching and processing costs, as well as general performance-related costs, such as costs associated with both means and variability of waiting time and sojourn time. The efficiency of our dispatching policies is illustrated with numerical examples.

M/M/1-PS queue and size-aware task assignment

We consider a distributed server system in which heterogeneous servers operate under the processor sharing (PS) discipline. Exponentially distributed jobs arrive to a dispatcher, which assigns each task to one of the servers. In the so-called size-aware system, the dispatcher is assumed to know the remaining service requirements of some or all of the existing jobs in each server. The aim is to minimize the mean sojourn time, i.e., the mean response time. To this end, we first analyze an M/M/1-PS queue in the framework of Markov decision processes, and derive the so-called size-aware relative value of state, which sums up the deviation from the average rate at which sojourn times are accumulated in the infinite time horizon. This task turns out to be non-trivial. The exact analysis yields an infinite system of first order differential equations, for which an explicit solution is derived. The relative values are then utilized to develop efficient dispatching policies by means of the first policy iteration (FPI). Numerically, we show that for the exponentially distributed job sizes the myopic approach, ignoring the future arrivals, yields an efficient and robust policy when compared to other heuristics. However, in the case of highly asymmetric service rates, an FPI based policy outperforms it. Additionally, the size-aware relative value of an M/G/1-PS queue is shown to be sensitive with respect to the form of job size distribution, and indeed, the numerical experiments with constant job sizes confirm that the optimal decision depends on the job size distribution.

Analyzing the dynamics and resource usage of p2p file sharing by a spatio-temporal model

In this paper we study the population dynamics and resource usage optimization of a P2P file sharing system, where the availability of the requested file is not guaranteed. We study the system first by a deterministic fluid model and then by a more detailed Markov chain analysis that allows estimating the life time of the system. In addition, the underlying topology of the network is modelled by a simple geometry. Using the resulting spatio-temporal model we assess how much the resource usage of the network can be reduced, e.g., by selecting the nearest seed for download instead of a random one.

On the Gittins index in the M/G/1 queue

For an M/G/1 queue with the objective of minimizing the mean number of jobs in the system, the Gittins index rule is known to be optimal among the set of non-anticipating policies. We develop properties of the Gittins index. For a single-class queue it is known that when the service time distribution is of type Decreasing Hazard Rate (New Better than Used in Expectation), the Foreground–Background (First-Come-First-Served) discipline is optimal. By utilizing the Gittins index approach, we show that in fact, Foreground–Background and First-Come-First-Served are optimal if and only if the service time distribution is of type Decreasing Hazard Rate and New Better than Used in Expectation, respectively. For the multi-class case, where jobs of different classes have different service distributions, we obtain new results that characterize the optimal policy under various assumptions on the service time distributions. We also investigate distributions whose hazard rate and mean residual lifetime are not monotonic.