We use load balance technique, Linux Virtual Server. Cluster server system construct the efficient structure own high scalability and availability through connecting with high performance network or LAN. In addition, applications provide transmitting real-time video and audio are increasingly needed with network bandwidth increase. To fit the current trend and test the multimedia streaming support of LVS, we add streaming service into web cluster system. But the playing of stream is not smooth, when the system is running. Streaming have serious delay and so we propose transmitting stream directly. When user had chose stream media from web then the client connect with streaming server directly and play the media continually, so we improve the performance of streaming.

本論文研究探討了在IEEE 802.11無線區域網路下支援服務品質的負載控制方法。我們所提出的ELB方法用動態的調整AP之間的網路負載分佈以達到負載平衡的目標。我們根據每個客戶端的統計特性進行負載平衡，並以允入控制來避免流量雍塞的情況發生。透過將使用者區分為三個等級，並控制每個使用者的使用頻寬來達到維持服務品質的目的。而進行漫遊的使用者，在我們的機制下，經由在新的AP上的頻寬預先保留，也可以維持一定的服務品質。除此之外，我們的ELB不需要修改任何的硬體機制，就可以運作在現存的802.11b無線區域網路中。最後，我們也對我們的機制做了模擬與實作，並量測、比較了我們的機制的表現。結果指出我們的結果可以有效的平衡AP間的負載，讓頻寬達到更大的使用效率，也能維持令人滿意的服務品質。 / This thesis presents and evaluates a mechanism for the load control with QoS supported in IEEE 802.11b Wireless LANs. Our mechanism named Enhanced Load Balance (ELB) dynamically adapts load distribution over APs to achieve load balance. The ELB mechanism balances the load by STAs’ statistical traffic load. This mechanism also performs admission control to avoid congestion. The ELB mechanism maintains QoS by classifying STAs into three classes and control the traffic flow of every STA. The roaming STAs can get enough bandwidth to maintain the QoS in the new AP by the bandwidth reservation mechanism of ELB. ELB can be used on top of the standard 802.11b access mechanism without requiring any modification or additional hardware. The performance of the IEEE 802.11 protocol with or without the ELB mechanism is investigated in the paper via simulation and implementation. The results indicate that our mechanism can balance the load effectively and the bandwidth can be fully utilized. Therefore, QoS can also be maintained.

PIM ¡]Processor-in-Memory¡^ architectures have been proposed in recent years for the purpose of reducing performance gap between processor and memory. This new class of computer architectures attempts to integrate processor and memory on a single one chip¡CWe proposed a new transformation and parallelizing system named SAGE ¡]Statement Analysis Group Evaluation¡^to fully utilize the host processor and memory processors in PIM systems. In this thesis, we focus on designing a load-balance optimization mechanism for the job scheduling. The experimental results of this mechanism are also discussed.

none

In this thesis, one of the most important issues in the field of networks communication is tackled and addressed. This issue is represented by QoS, where the increasing demand on highquality applications together with the fast increase in the rates of Internet users have led to massive traffic being transmitted on the Internet. This thesis proposes new ideas to manage the flow of this huge traffic in a manner that contributes in improving the communication QoS. This can be achieved by replacing the conventional application-insensitive routing schemes by others which take into account the type of applications when making the routing decision. As a first contribution, the effect on the potential development in the quality of experience on the loading of Basra optical network has been investigated. Furthermore, the traffic due to each application was dealt with in different ways according to their delay and loss sensitivities. Load rate distributions over the various links due to the different applications were deployed to investigate the places of possible congestions in the network and the dominant applications that cause such congestions. In addition, OpenFlow and Optica Burst Switching (OBS) techniques were used to provide a wider range of network controllability and management. A centralised routing protocol that takes into account the available bandwidth, delay, and security as three important QoS parameters, when forwarding traffics of different types, was proposed and implemented using OMNeT++ networks simulator. As a novel idea, security has been incorporated in our QoS requirements by incorporating Oyster Optics Technology (OOT) to secure some of the optical links aiming to supply the network with some secure paths for those applications that have high privacy requirements. A particular type of traffic is to be routed according to the importance of these three QoS parameters for such a traffic type. The link utilisation, end to end delays and securities due to the different applications were recorded to prove the feasibility of our proposed system. In order to decrease the amount of traffic overhead, the same QoS constraints were implemented on a distributed Ant colony based routing. The traditional Ant routing protocol was improved by adopting the idea of Red-Green-Blue (RGB) pheromones routing to incorporate these QoS constraints. Improvements of 11% load balancing, and 9% security for private data was achieved compared to the conventional Ant routing techniques. In addition, this Ant based routing was utilised to propose an improved solution for the routing and wavelength assignment problem in the WDM optical computer networks.

In a P2P (Peer-to-Peer) system, every user node, i.e., the peer, may dynamically join and leave the system. In general, peers can exchange information and contribute portions of their resources to the community in a P2P system. They are treated functionally identical. Therefore, it is very important to efficiently locate the peer that stores a particular data item and make the system load balance in P2P systems. Chord is a structured P2P system which has a ring architecture, where a structured P2P system means that peers maintain information about what resources neighbor peers offer. It provides support for just one operation: to assign the data key to the peer by hashing. Therefore, we can efficiently locate the peer that stores a particular data key. However, in the Chord system, most of data keys may be assigned to the same peer by using the static hashing scheme, which results in the case that the load of the system not be balanced. Therefore, in this thesis, we propose a strategy which uses the dynamic hashing scheme to locate the data key based on the Chord architecture, and to maintain the load balance. A dynamic hashing allows the address space allocated to the file to be increased and reduced without reorganizing the whole file. The basic idea of a dynamic hashing approach is to split the current overflow bucket into two new buckets by using the next level hashing function without reorganizing the other buckets, and our proposed strategy uses such an approach. In our strategy, we use two data structures for a peer, one stores the data hashed to the current peer and the other one stores the data from its predecessor. When an overflow occurs in the bucket after insertion of a data key, we use the one hashing function to split data keys stored in the data bucket. If the capacity of the current peer is larger than that of its successor, we forward some data keys to the successor. Similarly, we also consider the case of an underflow occurs in the bucket after deletion of a data key. Therefore, the unbalanced condition of the load (even distribution of items to nodes) of the system can be improved based on our strategy. From our simulation results, we show that the load of the P2P system based on our strategy is much more balanced than that used in the Chord system, when there are few peers and a lot of data keys in the P2P system. We also show that the load based on our strategy is still more balanced than that used in the Chord system, when the data distribution becomes skew.

A Peer-to-Peer system is a distributed system whose component nodes participate in similar roles. Every user node (the peer) can exchange and contribute its resources to another one in the system. Similar to the case that peers may dynamically join and leave the system, the data will also be inserted into and removed from the system dynamically. Given a certain range, a range query will find any data item whose value within the range. For example, a range query can find all the Beatle's works between 1961 and 1968 for us. However, once the range data is distributed over a P2P system through the hash function which has been used largely in many P2P systems, the continuity of the range data is not guaranteed to exist. Therefore, finding the scattered data whose value within a certain range costs much in a P2P system. The Distributed Segment Tree method (DST) preserves the local continuity of the range data at each node by using a segment tree and can break any given range into minimum number of node intervals whose union constitutes the whole requested range. The DST method works based on the Distributed Hash Table (DHT) logic; therefore, it can be applied in any DHT-based P2P system. However, data distribution of the DST method may cause overlapping. When searching a data range, the DST method sends more number of requests than what is really needed. Although the DST method designs the Downward Load Stripping Mechanism, the load on peers still may not be balanced. The main reason of these problems is that the DST method applies the DHT logic to the P2P systems. Therefore, in this thesis, we propose a method called Structured Segment Tree (SST) that does not use the DHT logic but embeds the structure of the segment tree into the P2P systems. In fact, the P2P network topology of an SST is the structure of a segment tree. Unlike a DST, an SST can fully reflect the properties of the original segment tree. Each peer in our proposed P2P system represents a node of a segment tree. Data intervals at the same level are continuous and will not overlap with each other. The union of data intervals at a level with full nodes is totally the whole data range which the P2P system can support. When searching a data range, the SST method sends as many number of requests as needed. In addition, we add sibling links to preserve the spatial locality and speed up the search efficiency. For the issue of load balance, our SST method also performs better than the DST method. From our simulation, we show that the SST method routes less number of peers to locate the requested range data than the DST method. We also show that the load based on our method is more balanced than that based on the DST method.

Optimization problems of location-assignment correspond to a wide range of real situations, such as factory network design. However most of the previous works seek in most cases at minimizing a cost function. Traffic incidents routinely impact the performance and the safety of the supply. These incidents can not be totally avoided and must be regarded. A way to consider these incidents is to design a network on which multiple assignments are performed. Precisely, the problem we focus on deals with power supplying that has become a more and more complex and crucial question. Many international companies have customers who are located all around the world; usually one customer per country. At the other side of the scale, power extraction or production is done in several sites that are spread on several continents and seas. A strong willing of becoming less energetically-dependent has lead many governments to increase the diversity of supply locations. For each kind of energy, many countries expect to deal ideally with 2 or 3 location sites. As a decrease in power supply can have serious consequences for the economic performance of a whole country, companies prefer to balance equally the production rate among all sites as the reliability of all the sites is considered to be very similar. Sharing equally the demand between the 2 or 3 sites assigned to a given area is the most common way. Despite the cost of the network has an importance, it is also crucial to balance the loading between the sites to guarantee that no site would take more importance than the others for a given area. In case an accident happens in a site or in case technical problems do not permit to satisfy the demand assigned to the site, the overall power supply of this site is still likely to be ensured by the one or two available remaining site(s). It is common to assign a cost per open power plant and another cost that depends on the distance between the factory or power extraction point and the customer. On the whole, such companies who are concerned in the quality service of power supply have to find a good trade-off between this factor and their overall functioning cost. This situation exists also for companies who supplies power at the national scale. The expected number of areas as well that of potential sites, can reach 100. However the targeted size of problem to be solved is 50. This thesis focuses on devising an efficient methodology to provide all the solutions of this bi-objective problem. This proposal is an investigation of close problems to delimit the most relevant approaches to this untypical problem. All this work permits us to present one exact method and an evolutionary algorithm that might provide a good answer to this problem. / Master of Science

The parallelization of complex, irregular scientific applications with various computational requirements often results in severe load imbalance. Load balancing increases the efficient utilization of available resources in parallel and distributed applications, thereby reducing the overall processor completion times. Loops are a rich source of parallelism in data parallel applications. In recent years, several loop scheduling schemes that balance processor workloads have been proposed and have been successfully implemented in data parallel applications. If the workload on processors is balanced, then the overall efficiency of a computation increases, and that, in turn reduces the computation run-time. Therefore, loop scheduling routines are incorporated into applications to insure that the workload is balanced for all the available processors. Significant research effort has been made towards embedding the most competitive loop scheduling algorithms into specific scientific applications. The application developer has to rewrite the algorithm to be incor-porated into a different application, each time a new one is developed. Certain compilers take advantage of loops present in the application and perform automatic parallelization on them. However, the automatic parallelization doesn?t address all sources of algorithmic and systemic variances in heterogeneous environments. These limitations raise a compelling need for building an application programmable interface (API) for adaptive loop scheduling algorithms that can be incorporated into any scientific application. This thesis presents an API for various adaptive loop scheduling strategies for data parallel applications in a shared address space architecture, which allows for parallelization as well as adaptive load balancing of a scientific application. This API has been incorporated into a few scientific applications in order to evaluate the performance of each application using the adaptive loop scheduling routines on shared address space parallel machines against the automatic loop scheduling offered by present parallelizing compiler technology.

The IEEE 802.11 standard has been widely adopted as a Wireless LAN (WLAN) technology. This widespread proliferation of the technology has lead to an increase in the number of users taking advantage of so-called "hot-spots" which leads to an increased demand on bandwidth provided by Access Points (APs) in the hot-spot. The logical solution is to deploy more overlapping access points in the same coverage area, thus increasing the capacity of the system by providing load balancing services. However, when a hot-spot is located in an outdoor environment, it becomes difficult to provide the AP with power which is traditionally carried over wired links thus causing the service provider to incur additional costs, not to mention the impossibility in some cases of delivering power to the AP. This problem can be overcome by using solar-panel powered APs which we will refer to as solar nodes (SNs). In this thesis we examine the load-balancing problem that arises when two or more SNs are co-located in the same coverage area. We propose and evaluate two algorithms for efficiently distributing the load among them (transferring stations (STAs) from SN to neighboring SNs) and increasing their lifetime by using power saving schemes that co-ordinate the wake/sleep patterns of the SNs based on traffic load. Finally, a Connection Admission Control (CAC) function is proposed that the SN should use in order to provide controlled access to services. We demonstrate through simulations that our proposals can significantly reduce the hardware requirements and cost of SNs and improve the service perceived by STAs in terms of transmission delay. / Thesis / Master of Applied Science (MASc)