Quality Of Service Support Mechanisms Research Articles

Toward OS-Level and Device-Level Cooperative Scheduling for Multitasking GPUs

As one of the most popular accelerators, the graphics processing unit (GPU) has been extensively adopted throughout the world. With the burst of new applications and the growing scale of data, co-running applications on limited GPU resources has become increasingly important due to its dramatic improvement in overall system efficiency. Quality of service (QoS) support among concurrent general-purpose GPU (GPGPU) applications is currently one of the most trending research topics. Prior efforts have been focused on providing QoS support either with OS-level or device-level scheduling methods. Each of these scheduling methods possesses pros and cons and may be unable to independently cover all the scheduling cases. In this paper, we propose a cooperative QoS scheduling scheme (C-QoS) that consists of operating-system-level (OS-level) scheduling and device-level scheduling. Our proposed scheme can control the progress of a kernel and provide thorough QoS support for concurrent applications in multitasking GPUs. Due to the accurate resource management of the copy engine and execution engine, C-QoS achieves QoS goals 23.33% more often than state-of-the-art QoS support mechanisms. The results demonstrate that cooperative methods achieve 17.27% higher system utilization than uncooperative methods.

Enhanced HCCA mechanism for multimedia traffics with QoS support in IEEE 802.11e networks

In broadband wireless networks, providing Quality of Service (QoS) support for multimedia traffics is no doubt crucial to many application requirements such as real-time audio and video streaming communications. IEEE 802.11e Medium Access Control (MAC) standard proposes Hybrid Coordination Function (HCF), which is composed of HCF Controlled Channel Access (HCCA) and Enhanced Distributed Channel Access (EDCA) mechanisms for QoS support. HCCA performs a polling procedure with parameterized channel. However, a station may not always have pending data to transmit. That is, polling all the stations may waste time and therefore deteriorate the transmission performance. In this paper, we propose a modified HCCA mechanism using the beacon frame broadcasted by the AP to determine which station can transmit its pending data so as to improve the transmission performance. The simulation results show that the performance improvement of our proposed mechanism is significant for QoS support particularly under light load condition.

Cross-layer bandwidth management and optimization in TDMA based wireless mesh networks using network coding

Wireless Mesh Networks (WMNs) provide a novel network architecture to extend broadband network coverage with low costs. Additionally, we see an increased interest in supporting demanding multimedia applications in next-generation wireless mesh networks. Provision of Quality of Service (QoS) in wireless mesh networks requires end-to-end support for routing packets via a suitable multihop route to the destination. However, in wireless mesh networks, if the medium access control layer does not support mechanisms for QoS support at a per-link level, all efforts for providing end-to-end QoS are futile. Hence, we see a trend towards standards for wireless mesh networks which support QoS at the Medium Access Control (MAC) level on a per-link basis.

단말의 마이크로 이동성을 고려한 자원예약 메커니즘의 설계

이동 컴퓨팅 환경에서 사용자의 이동 특성은 실시간 멀티미디어 서비스를 제공하기 위해서 요구되는 QoS에 심각한 영향을 미친다. 기존의 유선환경에서 실시간 서비스를 제공하기 위해서 제안된 RSVP 프로토콜을 이동 사용자의 네트워크 접속점이 수시로 변화하는 이동 컴퓨팅 환경에 적용하기에는 부적합하다. 사용자의 이동이 QoS에 미치는 영향을 최소화하기 위해서 사용자가 이동할 것으로 예상되는 영역에 대한 자원을 미리 예약하는 프로토콜이 MRSVP(Mobile RSVP)이다. 본 논문은 기존의 MRSVP를 확장하여 RSVP Tunnel 및 Mobile IP의 지역 등록(Regional Registration) 프로토콜이 결합된 동적 이중 앵커노드(DDAN : Dynamic Dual Anchor Node) 구조를 제안한다. 이는 기존의 MRSVP 프로토콜에서 미리 예약되는 자원을 지역적으로 제한함으로써 자원예약을 최소화시키는 반면에, 기존의 MRSVP와 유사한 수준의 QoS 보장을 목적으로 한다. It has been known that the host mobility feature has very significant impact on the QoS (Quality of Service), which is usually required to a real-time multimedia application. The existing QoS support mechanisms to provide the real-time services to fixed network environment, like as RSVP, are inadequate to accommodate the mobile hosts which can frequently change their point of attachments to the fixed network. So, MRSVP (Mobile RSVP) protocol has been proposed to reduce the impacts of host mobility on QoS guarantees, in which a mobile host needs to make advance resource reservations at multiple locations it may possibly visit during the lifetime of the connection. This paper proposes a dynamic dual anchor node (DDAN) architecture which integrates the MRSVP and RSVP tunnel, in addition to the Mobile IP Regional Registration protocol. By limiting the resource reserved in local area, it preserves the lower level of resource reservation, but provides approximately the same degree of QoS support as the existing MRSVP.

Quality-of-service architecture: monitoring and control of multimedia communications

For distributed multimedia applications the provision of system-wide quality of service (QoS) is essential. The quality-of-service architecture (QoS-A) developed at Lancaster University offers a framework to specify and implement the required performance properties of multimedia applications over integrated-service networks in a heterogeneous environment. The work addresses the problem of lower-level QoS as well as QoS support for multipeer multimedia applications operating in an open communication system. QoS filters have been developed to deal with the problems of heterogeneity in group communication environments. New challenges are also encountered in emerging mobile networks and user mobility; and recent developments in the Internet community highlight the need for appropriate QoS support mechanisms for multimedia applications. The paper describes how the QoS-A research at Lancaster University provides a flexible framework for the monitoring and control of multimedia communications across the broad range of computing systems and networks.

Filters: QoS support mechanisms for multipeer communications

The nature of distributed multimedia applications is such that they require multipeer communication support mechanisms. The multimedia traffic needs to be delivered to end-systems, networks, and end-users in a form that they can handle while satisfying the constraints imposed by the multimedia application. Quality-of-service (QoS) mechanisms that can ensure full quality media playout at high-performance workstations, while at the same time providing appropriately filtered lower quality media for playout at other end-systems, are required. Existing multicast support mechanisms are deficient for this purpose, in a heterogeneous environment, because they work on a common denominator premise where the quality delivered depends on the least capable link or node involved in the multicast session. This paper begins by discussing video compression; it proposes and analyzes the use of filtering mechanisms as means of supporting disparate receiver capabilities and QoS requirements. The paper describes the implementation of a number of filtering mechanisms and highlights the communications architecture within which these mechanisms are built. This architecture constitutes a specific network topology and a new protocol family developed within a UNIX-like operating system.