Linux ortamında ATM (asenkron iletim metodu) uygulaması

ATM (asynchronous transfer mode) application on linux

Tez No: 66436
Yazar: METEHAN SARAÇ
Danışmanlar: YRD. DOÇ. DR. A. COŞKUN SÖNMEZ
Tez Türü: Yüksek Lisans
Konular: Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrol, Computer Engineering and Computer Science and Control
Anahtar Kelimeler: Belirtilmemiş.
Yıl: 1997
Dil: Türkçe
Üniversite: İstanbul Teknik Üniversitesi
Enstitü: Fen Bilimleri Enstitüsü
Ana Bilim Dalı: Kontrol ve Bilgisayar Mühendisliği Ana Bilim Dalı
Bilim Dalı: Belirtilmemiş.
Sayfa Sayısı: 154

Özet

ÖZET Bilgisayar ağlan günümüzde çok yaygın olarak kullanılmakta ve teknolojileri de hızla gelişmektedir. Mevcut servise özel ağlardan genişbandlı ağlara geçişte önemli adımlar atılmıştır. Servise özel ağlar (telefon, data gibi) tek bir servisin ihtiyacını karşılamak üzere tasarlanmıştır. Dolayısıyla diğer uygulamalara destek verme konusunda son derece yetersizdirler. Son yıllardaki gelişmeler pekçok servise ortak olarak hizmet verecek, hızlı bir ağın gerekliliğini ortaya çıkarmıştır. Böylece genişbandlı haberleşme ağlan için Asenkron İletim Modu (Asynchronous Transfer Mode, ATM),CCnT tarafından (bugünkü adıyla ITU-T) en ideal çözüm olarak kabul edilmiştir. Genişbandlı haberleşme ağlan için ATM'e geleceğin teknolojisi gözüyle bakılmaktadır. ATM'de farklı istekleri olan çok değişik servisler tek bir ağ üzerinden sunulabilir. ATM, mevcut servise özel ağlara göre bir çok esnekliklere sahiptir. Ancak yapısı gereği ATM 'in de çözüm bekleyen önemli problemleri ortaya çıkmıştır. ATM bağlantı tabanlı bir iletim şeklidir. Yani haberleşecek olan iki sistem önce aralarında bir görüntü bağlantı (virtual connection) kurar. Bu bağlantı kurulurken sistemde servis için gerekli kaynaklatın mevcut olup olmadığı kontrol edilir ve mevcudiyeti durumunda bağlantıya izin verilir. Ayrıca ATM üzerindeki servisler, minimum bir servis kalitesi (Quality of Service, QoS) gerektirirler. Bu servis kalitesinin sağlanması da ATM 'in önemli konularından biridir. ATM'de haberleşme birimi olarak 53 byte uzunluğunda hücre (celi) adı verilen birimler seçilmiştir. Bu şekilde kısa bir paket boyu seçilmesini temel amacı işlem hızını artırmaktır. ATM'de hata kontrolü da minimuma indirilmiştir. ATM ile çok yüksek iletim hızlarına çıkalabilmektedir. Günümüzde ATM kullanımı hızla yaygınlaşmaktadır. Ancak henüz ATM ağlan azınlıkta olduğundan ATM'in mevcut ağlarla özellikle de İP ağlan ile entegrasyonu gerekmektedir. Bu üzerinde çok çalışılan konulardan biridir. Şu an piyasada PC'ler için de ATM sürücü kartları yaygınlaşmaya başlamıştır. Bu kartlar için değişik işletim sistemlerini destekleyen sürücü programlan geliştirilmiştir. Bunlar genellikle UNIX ve Microsoft tabanlı işletim sistemleri içindir. Bu çalışmada LINUX işletim sistemi için geliştirilmekte olan bir sürücü programı incelenmiş ve bu program kullanılarak LINUX işletim sistemi koşturan, birbirine ATM anahtarı üzerinden bağlı iki PC üzerinde basit bir istemci-sunucu ATM SVC uygulaması yapılmıştır.. Bu çalışmada Efficient Networks firmasınca üretilen 155 Mbps'lik ENI155p isimli ATM kartları kullanılmıştır. ATM anahtarına doğrudan bağlı bu iki PC, anahtar üzerinden hem PVC hem de SVC olarak haberleştirilmiştir. Kullanılan bu paket UNIX soketleri kullanılarak gerçeklenmiştir. Mevcut soket yapısına yapılan bazı eklerle ATM için gerekli fonksiyonlar sağlanmıştır. Hem LAN emülasyonu hem de ATM üzerinde İP olarak bu iki makina çalıştırılmıştır. Daha sonra da SVC kullanılarak basit bir istemci-sunucu programı gerçeklenmiştir. Bu çalışmada bir ATM kartı için bir cihaz sürücü programının nasıl yazılacağının ve ATM'in genel fonksiyonlarının nasıl sağlanacağının öğrenilmesi amaçlanmıştır. xu

Özet (Çeviri)

SUMMARY ATM (ASYNCHRONOUS TRANSFER MODE) APPLICATION ON LINUX In the evolution from the current telecommunication networks towards the Integrated Broadband Communication Network (IBCN) also called Full Service Network (FSN), some important directions and guidelines have recently been made. IBCN is often referred to as the Broadband Integrated Services Digital Network (BISDN) since it is considered as a logical extension of of the ISDN. The recent directions taken by the BISDN are influenced by a number of parameters, the most important being the emergence of a large number of teleservices with different, sometimes yet unknown requirements. The most famous teleservices to appear in the future are video-on-demand, video conferencing, high speed data transfer, videophony, video library, home education, home shopping, teleworking and HTV (High Definition TV). Each of these services will generate other requirements for the BISDN. This large span of requirements introduces the need for one universal network which is flexible enough to provide all of these services in the same way. Two other parameters are influencing the directions taken by the BISDN : they are the fast evolution of the semi-conductor and optical technology and the evolution in system concept ideas, e.g. the shift of superfluous transport functions to the edge of the network. Both the need for a flexible network and the progress in technology and system concepts led to the definition of the Asynchronous Transfer Mode (ATM) principle. This ATM concept is now accepted as the ultimate solution for the BISDN by CCITT (the International Consultative Committee for Telecommunications and Telecommunication). ATM is also accepted as the technology to inter-connect computers over ATM LANs by the computer industry in the ATM Forum. Todays telecommunication networks are characterized by specialization. This means that for every individual telecommunication service at least one network exists that transports this service. A few examples of existing public networks are described below : POTS (plain old telephone service) is transported via the public switched telephone network (PSTN). Packet switched data network (PSDN) based on X.25 protocols, circuit switched data network (CSDN) based on X.21 protocol. Broadcast via radio waves using ground antennae, coaxial tree netwoek of the committy antenna tv Xlll(CATV) network, direct broadcast system (DBS). For LANs Ethernet, token ring, token bus (IEEE 802 series). When designing the future BISDN network, one must take account all possible existing and future services. The networks of today are very specialized and suffer from a large number of disadvantages. For example service dependence, inflexibility, inefficiency. It is very important in the future that only a single network exists and that this network of the future (BISDN) is service-independent. A single service-independent network will not suffer from the disadvantages described above, but it will have these main advantages : flexible and future-safe, efficient in the use of its available resources, less expensive. The asynchronous transfer mode (ATM) is the transfer mode for implementing BISDN. The term transfer comprises both transmission and switching aspects, so a transfer mode is a specific way of transmitting and switching information in a network. In ATM, all information to be transferred is packed into fixed-size slots called cells. These cells have a 48 octet information field and a 5 octet header. In the flexibility of the ATM-baesd BISDN network access resulting from the cell transport concept strongly supports the idea of a unique interface which can be employed by a variety of customers with quite different service needs. However, the ATM concept requires many new problems to be solved. For example, the impacts of possible cell loss, cell transmission delay and cell delay variation on service quality need to be determined. ATM have the following common characteristics: no error protection or flow control on a link-by-link basis, ATM operates in a connection-oriented mode, the header functionality is reduced, the information field length is relatively small. The definition of ATM has specific consequences for the performance of an overall ATM based network, especially with respect to time and information transparency. Time Transparency Time transparency can be defined as the absence of delay and delay jitter (i.e. different parts of the information arrive at the destination with a different delay). Due to the high speeds used in a broadband network and the small information field of ATM packets, the delay characteristics of an ATM network are very different from those of classical packet switching networks. In ATM, all information must be tarnsfer to the destination in an acceptable delay. Semantic Transparency Information transparency determines the capability of the network to transport information accurately from the source to destination, i.e. with a limited (acceptable) number of errors. As in any packet switching system, the errors encountered in an ATM network are caused by transmission and switching / multiplexing systems. However, due to the properties of ATM (e.g. no error or flow control) the error behavior is different from that of the classical packet switching systems. xivThe BISDN protocol model for ATM contains 3 planes: a user plane to transport user information, a control plane mainly composed of signalling information and a management plane, used to maintain the network and to perform operational functions (Figure 1). In addition, a third dimension is added to the protocol referans model, called the plane management, which is responsible for the management of the different planes. This model has 3 layers defined by ITU-T (International Telecommunications Union-Telecommunication). These are physical layer, ATM Management Plane A Plane Figure 1 The BISDN ATM Protocol Reference Model layer, ATM adaptation layer (AAL). The physical layer of the BISDN is further composed of 2 sublayers : the Physical Medium (PM) sublayer supports pure medium dependent bit functions, the Transmission Convergence (TC) sublayer converts the ATM cell stream into bits to be transported over the physical medium. The ATM layer is fully independent of the physical medium used to transport the ATM cells and thus of the physical layer. Cell multiplexing / demultiplexing, VPI / VCI translation, cell header generation / extraction, implementation of a flow control mechanism on the user-network interface are the main functions performed by this layer. ATM Adaptation Layer (AAL) enhances the service provided by the ATM layer to a level required by the next higher layer. It performs functions for the user, control and management planes and supports the mapping between the ATM layer and the next higher layer. The AAL layer is subdivided into 2 sublayers : the segmantation and reassembly sublayer (SAR) and the convergence sublayer (CS), The main purpose of the SAR sublayer is segmentation of the higher layer information into a size suitable for the payload of the consecutive ATM cells of a virtual connection and the inverse operation. The convergence sublayer performs functions like message identificaiton, xvtime / clock recovery, etc. For some AAL types, the convergence sublayer has been further subdivided in a Common Part Convergence Sublayer (CPCS), and a Service Specific Convergence Sublayer (SSCS). The services which will be transported over the ATM layer are classified in 4 classes, each of which has its own specific requirements towards the AAL. To obtain these 4 classes, the services are classified according to 3 basic parameters : time relation between source and destination, bit rate, connection mode. Four types of AAL protocols are defined. These are AAL1, AAL2, AAL3/4 and AAL5. AAL1 is for constant bit rate (CBR) services. AAL2 offers a transfer of information with a variable bit rate. AAL3/4 is for transfer of data which is sensitive to loss, but not to delay. AAL5 offers a service with less overhead and better error detection below the CPCS layer according to AAL3/4. AAL5 is the simple way of AAL3/4. As explained previously, ATM is connection oriented, and thus requires a signalling protocol, able to set up, modify and release the connectios. The parameters used for each connection (such as bit rate, VPI/VCI,...) have to be agreed between the originating and receiving party as well as with the network, since it has to be checked by the 2 parties and the network whether both parties are able to operate in a compatible mode. The protocol between the user and the network (UNI) and the protocol between the network nodes (NNI) are quite comparable, but on the NNI often additional requirements are defined, such as higher reliability and additional multiplexing since more connections are involved. In the ISDN network of today, already such protocols exist which negotiate on parameters. These protocols are called Q.931 for the UNI and ISUP (ISDN User Part) for the NNI. They can be extented to replace the N-ISDN parameters by ATM parameters. In addition, in ATM other requirements such as point-to-multipoint exist, resulting in further requirements for the enhancement of the N-ISDN protocol by using additional messages. The BISDN signalling protocol is standardized as Q.2931 by ITU-T and under ATM User to Network Interface by ATM Forum, at the UNI. At the NNI it is standardized as B-ISUP by ITU, whereas ATM Forum has no specific standart for this interface. At the UNI, both Q.2931 and the ATM Forum protocols are derived from Q.93 1 (USDN). At the NNI, it is derived from ISUP. UNI Signalling Protocol This protocol is based on a subset of the broadband signalling protocol standarts that are currently under development (and currently identified as Q.2931). Additions to this subset have been made where necessary to support capabilities identified by the ATM Forum as important for early deployement and interoperability of ATM equipment. The primary areas where the standard has been supplemented are to support point-to-multipoint connections, additional traffic descriptos, and private network addressing. The procedures included in this protocol apply to the interface between terminal or endpoint equipment and a public network, referred to as Public UNI, and terminal or endpoint equipment connected to a private network, referred to as Private UNI. The basic capabilities supported by this protocol are listed below : xvi. Demand (switched) channel connections.. Point-to-point and point-to-multipoint switched channel connections.. Connections with symmetric or asymmetric bandwith requirements.. Single-connection (point-to-point or point-to-multipoint) calls.. Basic signalling functions via protocol messages, information elements and procedure.. Request and Indication of signalling parameters.. VPCI/VPI/VCI assignment. A single, statically defined out-of-band channel for all signalling messages.. Public UNI and Private UNI addressing formats for unique identification of ATM endpoints.. A client registration mechanism for exchange of addressing information across a UNI.. End-to-end Compatibility Parameter Identification. For UNI signalling, messages are used. Some of these are CALL PROCEEDING, CONNECT, CONNECT ACKNOWLEDGE SETUP, RELEASE, RELEASE COMPLETE, STATUS and STATUS ENQUIRY. A message is sum of information elements. Every message shall consist of the following parts : a) protocol discriminator, b) call reference, c) message type, d) message length, e) variable length information elements, as required. Information elements a), b), c) and d) are common to all the messages and shall always be present, while information element e) is specific to each message type. Some variable length information elements are call state, cause, endpoint reference, endpoint state, AAL parameters, ATM traffic descriptor, Quality of service parameter, broadband high layer information, broadband low layer information,calling party number, calling party subaddress, called party number, called party subaddress, restart indication. Signalling protocol uses the virtual channel VPI=0, VCI=5. For a call/connection to be established it must satisfy the following general criteria determined by the network and end systems :. Basic service support,. VC availability,. Physical and virtual network resource availability to provide quality service requested,. End system resource availability to provide quality of service requested,. End to end compatibility. xvnPoint-to-multipoint connections are supported in this protocol. A point-to-multipoint connection is a collection of associated ATM VC or VP links, with associated endpoint nodes, with the follwing properties :. One ATM link, called the Root Link, serves as the root in a simple tree topology. When the Root node sends information, all of the remaining nodes on the connection, called Leaf Nodes, receive copies of the information.. Only zero return bandwidth (i.e., from the Leaves to the Root) is supported.. The Leaf Nodes can not communicate directly to each other with this connection type.. A distributed implementation can be used to connect leaves to the tree. ADD PARTY, ADD PARTY ACKNOWLEDGE, DROP PARTY, DROP PARTY ACKNOWLEDGE are the messages used for adding and deleting a node to a call. In this protocol an address registration mechanism is used to exchange address information across the UNI. In order establish an ATM connection at the UNI, both the user and the network must know the ATM address(es) which are in effect at that UNI. The address registration procedures in this section provide the means for the dynamic exchange of addressing information between the user and the network at the UNI, at initialization and at other times as required. Through this dynamic exchange the user and network can agree on the ATM address(es) in effect. A Signalling ATM Adaptation Layer (SAAL) resides between the ATM layer and Q.2931. The purpose of the SAAL is to provide reliable transport of Q.2931 messages between peer Q.2931 entities (e.g., ATM Switch and host) over the ATM layer. The SAAL is composed of two sublayers, a common part and a service specific part. The service specific part is further subdivided into a Service Specific Coordination Function (SSCF), and a Service Specific Conncetion Oriented Protocol (SSCOP). UNIX Sockets ATM API, used for ATM on LINUX package is depend on the UNIX sockets. The two most prevalent communication APIs for Unix systems are Berkeley sockets and the System V Transport Layer Interface (TLI). ATM API is depend on Berkeley sockets.The typical client-server relationship is not symmetrical. To initiate a network connection requires that the program know which role (client or server) it is to play. A network connection can be connection-oriented or connectionless. For connection-oriented protocols, server uses socket( ), bind( ), listen( ), accept( ), read( ), write( ) and client uses socket ( ), connect( ), write( ), read( ) system calls. For connectionless protocol, server uses socket( ), bind( ), recvfrom( ),sendto( ) and client uses socket( ), bind( ), sendto( ), recvfrom( ) system calls. ATM on LINUX The first step in bringing ATM to LINUX was to find ATM adapters that offered sufficient performance, that were available on the market, and for which xvinprogramming information was openly available. In spring 1995, a driver fot the Efficient Networks ENI155p adapter was written. In order to send data over even only PVCs, a device driver alone isn't enough, but also an API is needed. Although ATM Forum is defining a semantic API, this description is far too general for any concrete implementation. Early tests revealed that throughput left much to be desired : instead of the theoretical maximum of 135.6 Mbps for user data with raw ATM, only a throughput of approximately 100 Mbps was obtained under ideal conditions. The results for IP over ATM were much worse. The reason was easily found : because PCs tend to have a slow memory interface, the comparably large number of copy operations in the kernel created a bottleneck. The problem was resolved using a concept called“single-copy”, where data is copied directly between user space and the device driver, without additional copying to kernel buffers. With single-copy, transfer rates of up to 130 Mbps are possible on Linux PCs with native ATM when using sufficiently large datagrams. Signaling protocols are implemented in a demon in user mode. The Linux kernel only implements a very simple protocol to support ATM signaling. All the complexity of ATM signaling is delegated to a user-mode demon process. The internal signaling protocol is much simpler than the signaling protocol that is used on the network, because the following assumptions can be made :. Communication is well-behaved (reliable, preserves sequence, etc.). The communication parties agree on every protocol detail, including the version or revision of the protocol.. The communicating parties always cooperate. Both parties share the same architecture. The protocol is designed for the use with Berkeley-style socket. All the general operations, such as binding to a local address, requesting an outgoing call (either blocking or non-blocking), accepting of incoming connections, etc. are supported. xix

Benzer Tezler

Tez No
613332
Implementation of a packet classifier for a router on FPGA
Bir ağ yönlendiricisi için FPGA üzerinde paket sınıflandırıcı gerçeklemesi
OĞUZHAN ÇİK
Yüksek Lisans
İngilizce
2019
Elektrik ve Elektronik Mühendisliği İstanbul Teknik Üniversitesi
Elektronik ve Haberleşme Mühendisliği Ana Bilim Dalı
PROF. DR. MÜŞTAK ERHAN YALÇIN
Tez No
129347
Assessing the quality of random varieties generated with the Unuran Library
Unuran Kütüphanesi ile üretilen rastsal değişkenlerin kalite analizi
SERKAN BARIŞ ORAL
Yüksek Lisans
İngilizce
2002
Endüstri ve Endüstri Mühendisliği Boğaziçi Üniversitesi
Endüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. WOLFGANG HÖRMANN
Tez No
524615
Grafik işleme ünitesi (GPU) tabanlı öğrenme kullanarak otonom araçlar için algılama sisteminin geliştirilmesi
Development of perception system for autonomous vehicles by using GPU-based learning
MEHMET SAFA BİNGÖL
Yüksek Lisans
Türkçe
2018
Mekatronik Mühendisliği Fırat Üniversitesi
Mekatronik Mühendisliği Ana Bilim Dalı
DOÇ. DR. AYŞEGÜL UÇAR
Tez No
183460
Construction and control of a desktop earthquake simulator
Bir masaüstü deprem simülatörünün imalatı ve kontrolü
GÖKÇE KINAY
Yüksek Lisans
İngilizce
2006
Deprem Mühendisliği İzmir Yüksek Teknoloji Enstitüsü
İnşaat Mühendisliği Ana Bilim Dalı
Y.DOÇ.DR. GÜRSOY TURAN
Tez No
620440
Shared object building makefiles with applications for ip protection
Sınai hakların korunması için paylaşımlı nesne derleyen makefıle tasarımı ve uygulamaları
ÖNDER GÖRMEZ
Yüksek Lisans
İngilizce
2020
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrol Gebze Teknik Üniversitesi
Elektronik Mühendisliği Ana Bilim Dalı
DR. ÖĞR. ÜYESİ ÖNDER ŞUVAK

Geri Dön