Distributed algorithms for QoS load balancing

Abstract

We consider a dynamic load balancing scenario in which users allocate resources in a non-cooperative and selfish fashion. The perceived performance of a resource for a user decreases with the number of users that allocate the resource. In our dynamic, concurrent model, users may reallocate resources in a round-based fashion. As opposed to various settings analyzed in the literature, we assume that users have quality of service (QoS) demands. A user has zero utility when falling short of a certain minimum performance threshold and having positive utility otherwise. Whereas various load-balancing protocols have been proposed for the setting without quality of service requirements, we consider protocols that satisfy an additional locality constraint: The behavior of a user depends merely on the state of the resource it currently allocates. This property is particularly useful in scenarios where the state of other resources is not readily accessible. For instance, if resources represent channels in a mobile network, then accessing channel information may require time-intensive measurements. We consider several variants of the model, where the quality of service demands may depend on the user, the resource, or both. For all cases we present protocols for which the dynamics converge to a state in which all users are satisfied. More importantly, the time to reach such a state scales nicely. It is only logarithmic in the number of users, which makes our protocols applicable in large-scale systems

Distributed Selfish Load Balancing on Networks

“…More directly related are concurrent protocols for selfish load balancing in different contexts that allow convergence results similar to ours. Whereas some papers consider protocols that use some form of global information [Even-Dar and Mansour 2005] or coordinated migration [Fotakis et al 2010], others consider infinitesimal or splittable tasks [Fischer and Vöcking 2008;Awerbuch et al 2008] or work without rationality assumptions Ackermann et al 2011;Hoefer and Skopalik 2013]. The machine models in these cases range from identical, uniformly related (linear with speeds) to unrelated machines.…”

“…Some distributed algorithmic approaches for load balancing have been proposed in algorithmic game theory (see, e.g., Berenbrink et al [2007], Even-Dar and Mansour [2005], Ackermann et al [2011], and Fischer et al [2010]). In this context, tasks are considered as selfish agents that act unilaterally and migrate concurrently between machines without global coordination.…”

“…The main challenge in the design of concurrent protocols such as ours is to carefully choose appropriate migration probabilities in order to guarantee rapid convergence while avoiding oscillation effects. Our scenario represents a significant extension of the existing literature on selfish load balancing, as concurrent protocols have been considered essentially only for complete graphs [Awerbuch et al 2008;Ackermann et al 2011;Even-Dar and Mansour 2005;Berenbrink et al 2007Berenbrink et al , 2012]. …”

IDM: An Indirect Dissemination Mechanism for Spatial Voice Interaction in Networked Virtual Environments

“…Since the idea of using a potential function to measure the closeness to a Nash equilibrium was introduced by Eyal Even-Dar and Mansour [11], a number of protocols (e.g., [12], [13]) have been proposed to balance the loads in a weakly distributed manner, where the migration events take place one at a time, and costs on users are updated immediately. More recently, concurrent protocols (e.g., [14], [15], [16]) for load balancing have been proposed to improve the convergence time by allowing users to change their strategies concurrently. All the protocols above only address load balancing and do not consider the impact of heterogeneous peer capacities on the performance of the system.…”

Brief announcement

“…In particular, previous works [1,7] have only addressed the case when every machine is available to every user throughout the whole balancing process (i.e., when a "complete network" exists among resources). In this paper, we advance the understanding of these protocols in scenarios with locality restrictions to user migration.…”