KENDALI LAJU ADAPTIF MENGGUNAKAN KURVA BEBAN HILANG UNTUK TRAFIK ABR PADA JARINGAN ATM

• Main Author: F.S., Christoffel
• Format: Theses
• Language: Indonesia
• Subjects: Teknik (Rekayasa, enjinering dan kegiatan berkaitan)
• Online Access: https://digilib.itb.ac.id/gdl/view/4767
• Institution: Institut Teknologi Bandung

<b>Abstract :</b><p align=\"justify\">In 1988, ATM (Asynchronous Transfer Mode) was selected as the universal transfer mode for multimedia traffic in the context of B-ISDN (Broadband Integrated Services Digital Network). ATM is a connection oriented packet switching and transports data in 53-byte cells. In order to accomodate different types of connections, four services are currently defined in ATM network: CBR (Constant Bit Rate), VBR (Variable Bit Rate), ABR (Available Bit Rate) and UBR (Unspecified Bit Rate). <p align=\"justify\"><br /> This thesis is concerned about ABR services, which is devised to handle LAN traffic across ATM network. The bandwidth left unallocated to CBR and VBR sources is made available to ABR sources for sharing. In order to share the available bandwidth, an ABR source is required to regulate its transmission rate according to load conditions in the network. Adaptive rate control using load loss curve provides a direct mathematical relationship between offered load and the level of packet loss within the network. Switches compute network load and provide explicit feedback through parameters in resources management (RM) cells to the sources. With this feedback information, traffic sources can increase or decrease their transmission rates. Sources can use this mechanism to maximise throughput or to control packet loss within the network. <p align=\"justify\"><br /> Simulation results show that adaptive rate control using loss load curve gives link utililization equals to one for offered load more than link capacity, fairness in bandwidth allocation to all sources, nominal buffering and exact and fast rate transition. However, it requires large source rates database storage to work in realtime environment.