FDDI ve FDDI şebekeleri performans analizi
Başlık çevirisi mevcut değil.
- Tez No: 46204
- Danışmanlar: PROF.DR. GÜNSEL DURUSOY
- Tez Türü: Yüksek Lisans
- Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1995
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 90
Özet
ÖZET Dağılmış Veri Fiber Arabağlaşımı ( FDDI ), 100Mbit/s iletişim kapasitesinde bir yerel alan şebekesidir. İletişim ortamı fiber optiktir ve halka şebekede bayrak geçirme protokolünü kullanır. FDDI, ANSI ( American National Standards Institute ) X3T9 Teknik Komitesi tarafından X3T9.5 olarak geliştirilmiştir. OSI modelinin tanımlanması ve tabakalı veri için gerekli arabirimlerin sunulması, hazır şebekeler için FDDI standartlarının gelişmesine yardımcı olmuştur. FFDPın hızlı gelişmesindeki en önemli faktör yüksek performanslı, çeşitli uygulamalara yönelik video iş istasyonlarının artan kullanımıdır. FDDI-II yapısının gerçekleştirilmesi de devre-bağlaşmalı veri trafiği alanında yeni bir dönem açmıştır. FDDI-II 1980'lerin sayısal PBX'lerinin yeni bir nesli olarak görülebilir. Ortam olarak optik fiberin kullanılması, yüksek band genişliğine sahip, güvenli, gizli ve elktromagnetik etkiden etkilenmeyen optik fiber LAN teknonolojisini ortaya koymuştur. Ayrıca ikili-halka yapısı, kullanılan veri kodlama tekniği, asenkron, senkron, devre bağlaşmalı veri trafiğinin yüksek kapasitede iletimi, FDDI' m avantajlarının bir parçasını oluşturur. FDDI, son zamanlarda Ethernet ve Bayraklı Halka gibi düşük hızlı“alt”şebekeleri birbirine bağlamada önem kazandığı gibi, fiyatların ucuzlamasıyla masaüstü bilgisayar ağlarında da kullanılması olası olacaktır. FDDI, Elektronik Posta (E-Mail) grafik veri tabanları, ses ve video ile yürütülen sunuşlar gibi çok ortamlı uygulamalarda kullanılabilecek bir standarttır. Bu çalışmada FDDFın temel birimleri tartışılmış, performans analizleri çıkarılmış, çıkış veriminin arttırılması için kullanılan protokol farkları (Timed token protokol) anlatılmış ve rakip teknolojilerle teknik olarak ve kullanılabilirlik açısın dan karşılaştırılmıştır.
Özet (Çeviri)
SUMMARY FDDI AND PERFORMANCE ANALYSIS OF FDDI NETWORKS The Fiber Distributed Data Interfaces (FDDI) specifications of the ANSI committee X3T9.5 (ISO 9314) describe a 100-Mbps, token ring, fiber based LAN. The LAN has two counter-rotating fiber optic rings (usually implemented in the same fiber sheath). The standard allows up to 500 stations to communicate via fiber optic cables using a timed-token accessprotokol which was finally standardized as the Media Access Control (MAC) PROTOKOL for the IEEE 802.4 Token Bus and for FDDI. It belongs to the class of Token Passing Protokols and allows the integration of time- critical (synchronous) and nontime-critical (asynchronous) traffic. Normal data traffic as well as time constrained traffic such as voice, video, and real-time applications are supported. All major computer vendors, communications vendors, and integrated circuit manufacturers are offering products supporting this standard. The Institude of Electrical and Electronics Engineers (IEEE) P802 standards project was developing local area network (LAN) standards with data rates up to 20 Mbit/s. FDDI followed the packet data architectural concepts of IEEE P802 and chose the emerging 4Mbit/s token ring protokol of IEEE 802.5 as the starting point for the FDDI protokol. These choices placed FDDI in an ideal position to be both the backbone network and the follow-on network to the IEEE P802 LAN's, In another standards arena, the Open Systems Interconnection (OSI) model had been put in place. This layered the design of computer interconnections, allowing the development of separate standards for development of a set of FDDI standards.Another factor of significance in the development of FDDI was on increased use of high-performance video workstations for a variety of applications. This brought an emphasis on facilitating low-cost implementations. Driven by these forces, FDDI grew to satisfy the needs of many applications, including the back-end (I/O channel) interface, LAN bacbone,and front-end high performance LAN applications. As a result, the set of services offered by FDDI is broad enough to allow individual optimization of FDDI networks to satisfy the needs of diverse environments. One enhancement, FDDI-II, will offer significantly increased services by integrating circuit-switched data traffic capabilities into what had originally been strictly a packet LAN. The impetus for FDDI-II came from the new generation of digital PBX's of the early 1980's. Their needs were similar to those of many real-time applications including digital voice and video networks as well as sensor and control data streams. All of these disciplines contributed to the emerging FDDI-II design definition which began in late 1984. Two of the most mature and widely used of these toolsare Hewlett-Packard's remote procedure call service, Network Computing System (NCS), and SUN Microsystem's file service, Network File System. (NFS). With the explosion of distributed applications based on these tools, network performance of FDDI speeds will be critical. FDDI has the potential to provide the bandwidth requirements for these applications, but to truly deliver this potential and achieve high performance distributed computing in FDDI networks, several factors will need to be considired. Among these factors are FDDI topologies, network protokol and the design of distributed processing tools to utilize the dual ring (200Mb/s) bandwidth potential of FDDI. The FDDI standard specifies four items:. Physical Media Dependent (PMD) -specifies the optical transceivers, connectors and media characteristics.. Physical Layer (PHY) - specifies data encoding, decoding, and clocking. VI. Media Access Control (MAC) - specifies the frame format, addressing, token passing mechanism. ® Station Management (SMT) - specifies the management of the ring, including the interoperability of stations, token management recovery, etc. FDDI's basic topology is characterized by their attachment types and number of Media Access Controllers (MACs). A Dual Attachment Station (DAS) has two physical attachments, one to each ring of the dual trunk ring, and one or two MACs (single or dual MAC). High end workstations, servers, and routers will generally be DAS stations. Stations that have only one physical attachment (Single Attachment Station-SAS) are connected to one of the dual rings by stations called concentrators. Lower and desktop workstations and Personel Computers (PCs) are likely to be FDDI SAS stations. Advantages of a ring design: A ring can be shown to offer superior reliability, availability, and serviceability, even in the face of physical damage to the network. A ring topology can be designed to be capable of continued operation despite any projected failure. Other advantages include the interconnect simplicity of the physical hardware at the interface level. Ring topologies offer advantages in the ease of initial configuration and reconfiguration as the network requirements change. Failing stations or fiber links can be isolated through the use of appropriate protocols. These protocols also provide for the logical addition and deletion of stations without detrimental effects on existing ring traffic. Actual physical addition or removal of stations from the network is also facilitated because ring initialization, failure isolation, recovery,and reconfiguration mechanisms can provide for continued operation even while the cables are being rearranged. VIIRing topologies inherently impose no restrictive logical limit on the length of ring links, the number of stations, or the total extent of the network that can be accommodated. Ring topologies and the protocols supported by them,offer significant performance advantages. These include insensitivity to load distribution, the ease of fairly allocating the available bandwidth, low arbitration times, bounded access delay, and no requirement for long preambles. FDD1-II Concepts: FDDI-II is an upward-compatible enhancement of the basic FDDI that adds a circuit- switched service to the existing packet capability. Circuit-switched service provides a continues connection between two or more stations. Instead of using addresses, the connection is established based upon some prior agreement, which may have been negotiated using packet messages or established by some other suitable convention known to the stations involved. This prior agreement typically takes the form of knowing the location of a time slot, or slots, that occur regularly relative to a readily recognizable timing marker. A common timing marker used in North America is the Basic System Reference Frequency (BSRF), a 125us clock used by the public networks. Use of this clock is assumed for FDDI-II. In local FDDI usage, this is referred to as the cycle clock and is signaled by the JK starting delimiter of the FDDI-II cycle format. In FDDI-II, a circuit-switched connection is described as N bits beginning at byte M after the cycle clock marker in Wideband Channel (WBC) number X. The last descriptor is necessary because FDDI-II has 16 WBC's that may be independently assigned to either packet-switched or circuit-switched data. This definition allows connections at data rates of all multiples of 8 kbits/s (i.e.,N=l) up to the 6.144Mbit/s data rate of a WBC. If need be, multiple WBC's may be used to accommodate higher data rates. The data transferred in a circuit-switched mode is best described as a stream of data. VIIIThe data rate is appropriate to the service being provided with, for example, 64 kbits/s being used for a digital voice data stream rates, even up to many Mbit/s in the case of video, are used for other applications. Once a connection is established, the data rate remains constant. Most packet data traffic occurs in random quantities and at random times. This is referred to as asynchronous traffic, more regular in nature and occuring in relatively predictable quantities on a regular time basis, is referred to as synchronous (packet) traffic. In contrast, isochronous data occur in precise amounts on a precise time basis. They typically represent a sequence of digital samples from a sensor (e.g., voice or video). More importantly, isochronous data must by synchronized with clock information to ensure the accurate regeneration of the sampling clock (as distinct from the bit clock) to minimize distortion in data reconstruction. Isochronous data are more easily transferred in a circuit-switched network. Networks that carry isochronous data must maintain precise synchronism with the cycle clock. For the FDDI ring, this means that one station (called the cycle master) must insert a delay for all isochronous data so that the ring appears to be an exact multiple of 125us in length.FDDI incorporates this delay in the cycle master in such a way that if does not cause any delay in the packet traffic. This is essential in providing an integrated services network with acceptable packet service. Data Encoding: FDDI uses a 4 of 5 code, that is a code with 4 data bits for every 5 code bits. Such a code has an efficiency of 80 percent, in contrast to the“Manchester”codes used in many other networks, which are 1 of 2 codes with an efficiency of 50 percent. There is, however, no perfectly balanced non-adaptive 4 of 5 code. This means that for some code values there are more ones than zeroes. The code chosen for FDDI has a maximum plus or minus 10 percent DC unbalance. Experimental work done in the development of the standard indicates that this causes about a 1 dB link power penalty for equivalent error rates when a long sequence of worst case code values is transmitted. Since the balanced alternatives either have lower code efficiencies, or, in the case of an 8 of 10 code,require a significantly more complex decoder and encoder, this seems a good tradeoff. IXIt will require the use of worst case code value patterns to test receiver sensitivity and must be allowed for in the sensitivity specification. In the 32-member symbol set, 16 symbols are data symbols, each representing four bits of ordered binary data. Three symbols are used for line-stated signaling which is recognized by the physical layer hardware, two are used as control indicators, and four are used for starting and ending delimiters. The remaining symbols of the symbol set are not to be transmitted since they violate code run length and dc balance requirements. The focus on developing FDDI as a standard for a highly reliable network was a good one.FDDI has won acceptance in a number of military arenas normally closed to commercial standards, with SAFENET and aircraft notable examples. FDDI's reputation as a highly reliable network has proved an important factor in its acceptance. The 1990's have brought increasing emphasis on multimedia, e.g., video and services. This has stirred interest in the FDDI-II enhancement. While some support its use, others have shown that FDDI synchronous service is satisfactory for some bounded multimedia applications. Others have analyzed the performance tradeoffs between the use of FDDI isochronous and synchronous services for multimedia applications. In any case, it is evident that FDDI, including FDDI-II, can satisify a variety of multimedia applications. Although implementers will in all likelihood choose different solutions, no problem is foreseen, since both alternatives may coexist in the same FDDI ring. Even more important, adopting new ideas improved the FDDI standards.Indeed, as the standards progressed, the overall capability and applicability of FDDI grew considerably. By the time FDDI matured, it had become a family of standards with much broader appeal than was imagined İn the early days. It is the premier high-performance LAN of today. FDDI and the follow-up work to it now under way in X3T9.5 can be expected to dominate high-speed LAN networking into the next millennium. X
Benzer Tezler
- ATM şebekelerde trafik analizi ve güvenlik
Başlık çevirisi yok
HALİL AYDIN
Yüksek Lisans
Türkçe
1998
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektronik ve Haberleşme Mühendisliği Ana Bilim Dalı
PROF. DR. GÜNSEL DURUSOY
- Firma şebekelerinde ATM
Başlık çevirisi yok
BEHÇET KURBAN
Yüksek Lisans
Türkçe
1995
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiPROF. DR. GÜNSEL DURUSOY
- SDH şebekeler ve SDH şebekelerde yönetim
Başlık çevirisi yok
ZAFER GEDİK
Yüksek Lisans
Türkçe
1998
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektrik-Elektronik Mühendisliği Ana Bilim Dalı
PROF. DR. GÜNSEL DURUSOY
- Asenkron transfer modu ve çerçeve aktarma
Asynchronous transfer mode and frame relay
HALİT DÖNMEZ
Yüksek Lisans
Türkçe
1997
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektronik ve Haberleşme Mühendisliği Ana Bilim Dalı
PROF. DR. GÜNSEL DURUSOY
- ATM'de lan hizmetleri
ATM-lan services
İRFAN UĞUR
Yüksek Lisans
Türkçe
1997
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektronik ve Haberleşme Mühendisliği Ana Bilim Dalı
PROF. DR. GÜNSEL DURUSOY