• Refine Query
• Source
• Publication year
• to
• Language
• 140
• 41
• 19
• 11
• 7
• 6
• 3
• 2
• 2
• 2
• 2
• 1
• 1
• 1
• 1
• Tagged with
• 279
• 279
• 59
• 58
• 51
• 51
• 47
• 45
• 38
• 34
• 33
• 33
• 33
• 31
• 31
• The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Visibility-based microcells for dynamic load balancing in MMO games

SUMILA, ALEXEI 29 September 2011 (has links)
Massively multiplayer games allow hundreds of players to play and interact with each other simultaneously. Due to the increasing need to provide a greater degree of interaction to more players, load balancing is critical on the servers that host the game. A common approach is to divide the world into microcells (small regions of the game terrain) and to allocate the microcells dynamically across multiple servers. We describe a visibility--based technique that guides the creation of microcells and their dynamic allocation. This technique is designed to reduce the amount of cross--server communication, in the hope of providing better load balancing than other load--balancing strategies. We hypothesize that reduction in expensive cross-server traffic will reduce the overall load on the system. We employ horizon counts map to create visibility based microcells, in order to emphasize primary occluders in the terrain. In our testing we consider traffic over a given quality of service threshold as the primary metric for minimization. As result of our testing we find that dynamic load balancing produces significant improvement in the frequency of quality of service failures. We find that our visibility-based micro cells do not outperform basic rectangular microcells discussed in earlier research. We also find that cross-server traffic makes up a much smaller portion of overall message load than we had anticipated, reducing the potential overall benefit from cross server message optimisation. / Thesis (Master, Computing) -- Queen's University, 2011-09-28 14:15:32.173
2

Stochastic scheduling in networks

Dacre, Marcus James January 1999 (has links)
No description available.
3

Improved Prediction-based Dynamic Load Balancing Systems for HLA-Based Distributed Simulations

Alkharboush, Raed January 2015 (has links)
4

A resource aware distributed LSI algorithm for scalable information retrieval

Liu, Yang January 2011 (has links)
5

Chou, Yu-chieh 02 September 2009 (has links)
6

Generation and properties of random graphs and analysis of randomized algorithms

Gao, Pu January 2010 (has links)
We study a new method of generating random $d$-regular graphs by repeatedly applying an operation called pegging. The pegging algorithm, which applies the pegging operation in each step, is a method of generating large random regular graphs beginning with small ones. We prove that the limiting joint distribution of the numbers of short cycles in the resulting graph is independent Poisson. We use the coupling method to bound the total variation distance between the joint distribution of short cycle counts and its limit and thereby show that $O(\epsilon^{-1})$ is an upper bound of the $\eps$-mixing time. The coupling involves two different, though quite similar, Markov chains that are not time-homogeneous. We also show that the $\epsilon$-mixing time is not $o(\epsilon^{-1})$. This demonstrates that the upper bound is essentially tight. We study also the connectivity of random $d$-regular graphs generated by the pegging algorithm. We show that these graphs are asymptotically almost surely $d$-connected for any even constant $d\ge 4$. The problem of orientation of random hypergraphs is motivated by the classical load balancing problem. Let $h>w>0$ be two fixed integers. Let $\orH$ be a hypergraph whose hyperedges are uniformly of size $h$. To {\em $w$-orient} a hyperedge, we assign exactly $w$ of its vertices positive signs with respect to this hyperedge, and the rest negative. A $(w,k)$-orientation of $\orH$ consists of a $w$-orientation of all hyperedges of $\orH$, such that each vertex receives at most $k$ positive signs from its incident hyperedges. When $k$ is large enough, we determine the threshold of the existence of a $(w,k)$-orientation of a random hypergraph. The $(w,k)$-orientation of hypergraphs is strongly related to a general version of the off-line load balancing problem. The other topic we discuss is computing the probability of induced subgraphs in a random regular graph. Let $0<s<n$ and $H$ be a graph on $s$ vertices. For any $S\subset [n]$ with $|S|=s$, we compute the probability that the subgraph of $\mathcal{G}_{n,d}$ induced by $S$ is $H$. The result holds for any $d=o(n^{1/3})$ and is further extended to $\mathcal{G}_{n,{\bf d}}$, the probability space of random graphs with given degree sequence $\bf d$. This result provides a basic tool for studying properties, for instance the existence or the counts, of certain types of induced subgraphs.
7

On Load Balancing and Routing in Peer-to-peer Systems

Giakkoupis, George 15 July 2009 (has links)
A peer-to-peer (P2P) system is a networked system characterized by the lack of centralized control, in which all or most communication is symmetric. Also, a P2P system is supposed to handle frequent arrivals and departures of nodes, and is expected to scale to very large network sizes. These requirements make the design of P2P systems particularly challenging. We investigate two central issues pertaining to the design of P2P systems: load balancing and routing. In the first part of this thesis, we study the problem of load balancing in the context of Distributed Hash Tables (DHTs). Briefly, a DHT is a giant hash table that is maintained in a P2P fashion: Keys are mapped to a hash space I --- typically the interval [0,1), which is partitioned into blocks among the nodes, and each node stores the keys that are mapped to its block. Based on the position of their blocks in I, the nodes also set up connections among themselves, forming a routing network, which facilitates efficient key location. Typically, in a DHT it is desirable that the nodes' blocks are roughly of equal size, since this usually implies a balanced distribution of the load of storing keys among nodes, and it also simplifies the design of the routing network. We propose and analyze a simple distributed scheme for partitioning I, inspired by the multiple random choices paradigm. This scheme guarantees that, with high probability, the ratio between the largest and smallest blocks remains bounded by a small constant. It is also message efficient, and the arrival or departure of a node perturbs the current partition of I minimally. A unique feature of this scheme is that it tolerates adversarial arrivals and departures of nodes. In the second part of the thesis, we investigate the complexity of a natural decentralized routing protocol, in a broad family of randomized networks. The network family and routing protocol in question are inspired by a framework proposed by Kleinberg to model small-world phenomena in social networks, and they capture many designs that have been proposed for P2P systems. For this model we establish a general lower bound on the expected message complexity of routing, in terms of the average node degree. This lower bound almost matches the corresponding known upper bound.
8

Generation and properties of random graphs and analysis of randomized algorithms

Gao, Pu January 2010 (has links)
We study a new method of generating random $d$-regular graphs by repeatedly applying an operation called pegging. The pegging algorithm, which applies the pegging operation in each step, is a method of generating large random regular graphs beginning with small ones. We prove that the limiting joint distribution of the numbers of short cycles in the resulting graph is independent Poisson. We use the coupling method to bound the total variation distance between the joint distribution of short cycle counts and its limit and thereby show that $O(\epsilon^{-1})$ is an upper bound of the $\eps$-mixing time. The coupling involves two different, though quite similar, Markov chains that are not time-homogeneous. We also show that the $\epsilon$-mixing time is not $o(\epsilon^{-1})$. This demonstrates that the upper bound is essentially tight. We study also the connectivity of random $d$-regular graphs generated by the pegging algorithm. We show that these graphs are asymptotically almost surely $d$-connected for any even constant $d\ge 4$. The problem of orientation of random hypergraphs is motivated by the classical load balancing problem. Let $h>w>0$ be two fixed integers. Let $\orH$ be a hypergraph whose hyperedges are uniformly of size $h$. To {\em $w$-orient} a hyperedge, we assign exactly $w$ of its vertices positive signs with respect to this hyperedge, and the rest negative. A $(w,k)$-orientation of $\orH$ consists of a $w$-orientation of all hyperedges of $\orH$, such that each vertex receives at most $k$ positive signs from its incident hyperedges. When $k$ is large enough, we determine the threshold of the existence of a $(w,k)$-orientation of a random hypergraph. The $(w,k)$-orientation of hypergraphs is strongly related to a general version of the off-line load balancing problem. The other topic we discuss is computing the probability of induced subgraphs in a random regular graph. Let $0<s<n$ and $H$ be a graph on $s$ vertices. For any $S\subset [n]$ with $|S|=s$, we compute the probability that the subgraph of $\mathcal{G}_{n,d}$ induced by $S$ is $H$. The result holds for any $d=o(n^{1/3})$ and is further extended to $\mathcal{G}_{n,{\bf d}}$, the probability space of random graphs with given degree sequence $\bf d$. This result provides a basic tool for studying properties, for instance the existence or the counts, of certain types of induced subgraphs.
9

On Load Balancing and Routing in Peer-to-peer Systems

Giakkoupis, George 15 July 2009 (has links)
A peer-to-peer (P2P) system is a networked system characterized by the lack of centralized control, in which all or most communication is symmetric. Also, a P2P system is supposed to handle frequent arrivals and departures of nodes, and is expected to scale to very large network sizes. These requirements make the design of P2P systems particularly challenging. We investigate two central issues pertaining to the design of P2P systems: load balancing and routing. In the first part of this thesis, we study the problem of load balancing in the context of Distributed Hash Tables (DHTs). Briefly, a DHT is a giant hash table that is maintained in a P2P fashion: Keys are mapped to a hash space I --- typically the interval [0,1), which is partitioned into blocks among the nodes, and each node stores the keys that are mapped to its block. Based on the position of their blocks in I, the nodes also set up connections among themselves, forming a routing network, which facilitates efficient key location. Typically, in a DHT it is desirable that the nodes' blocks are roughly of equal size, since this usually implies a balanced distribution of the load of storing keys among nodes, and it also simplifies the design of the routing network. We propose and analyze a simple distributed scheme for partitioning I, inspired by the multiple random choices paradigm. This scheme guarantees that, with high probability, the ratio between the largest and smallest blocks remains bounded by a small constant. It is also message efficient, and the arrival or departure of a node perturbs the current partition of I minimally. A unique feature of this scheme is that it tolerates adversarial arrivals and departures of nodes. In the second part of the thesis, we investigate the complexity of a natural decentralized routing protocol, in a broad family of randomized networks. The network family and routing protocol in question are inspired by a framework proposed by Kleinberg to model small-world phenomena in social networks, and they capture many designs that have been proposed for P2P systems. For this model we establish a general lower bound on the expected message complexity of routing, in terms of the average node degree. This lower bound almost matches the corresponding known upper bound.
10