ATM technology is a telecommunications concept defined by international standards for the transmission of a full range of user traffic, including voice, data and video signals. It was designed to meet the needs of a digital broadband service network and was originally designed to integrate telecommunication networks. The abbreviation ATM stands for Asynchonous Transfer Mode and is translated into Russian as "asynchronous data transfer."
The technology was created for networks that must handle both traditional high-performance data traffic (for example, file transfer) and real-time content with low latency (such as voice and video). The reference model for ATM is roughly comparable to the three lowest levels of ISO-OSI: network, data link, and physical. ATM is the main protocol used over the main channel SONET / SDH (Public Switched Telephone Network) and the Integrated Services Digital Network (ISDN).
What it is?
What does ATM mean for network connectivity? It provides functionality similar to switching channels and packet switching networks: the technology uses asynchronous time division multiplexing and encodes the data into small fixed-size packets (ISO-OSI frames) called cells. This differs from approaches such as the Internet Protocol or Ethernet, which use variable-size packets and frames.
The basic principles of ATM technology are as follows. It uses a connection-oriented model in which a virtual circuit must be installed between two endpoints before the actual exchange of data begins. These virtual circuits can be “permanent”, that is, dedicated connections that are usually pre-configured by the service provider, or “switched”, that is, customizable for each call.
Asynchonous Transfer Mode (ATM decryption from English) is known as the communication method used in ATMs and payment terminals. However, this application is gradually reduced. The use of technology in ATMs has been largely replaced by Internet Protocol (IP). In the ISO-OSI reference channel (layer 2), basic transmitters are commonly referred to as frames. In ATM, they have a fixed length (53 octets or bytes) and are specifically called “cells”.
Cell size
As already noted above, ATM decryption is an asynchronous data transfer carried out by dividing them into cells of a certain size.
If the speech signal is reduced to packets, and they are forced to be transmitted by reference with intense data traffic, then no matter what their size, they will encounter voluminous full-scale packets. Under normal waiting conditions, they may experience maximum delays. To avoid this problem, all ATM packets or cells are the same small size. In addition, the structure of fixed cells means that data can be easily transmitted by hardware without the inherent delays introduced by software dial-up and routed frames.
Thus, ATM developers used small data cells to reduce jitter (in this case, delay variance) in multiplexing data streams. This is especially important when transferring voice traffic, since the conversion of digitized voice into an analog audio signal is an integral part of the real-time process. This helps the decoder (codec), which requires a uniformly distributed (over time) stream of data elements. If the next in line is not available when necessary, the codec has no choice but to pause. In the future, the information is lost, because the period of time when it should have been converted into a signal has already passed.
How was the development of ATM?
During the development of ATM, a 155 Mbps synchronous digital hierarchy (SDH) with a payload of 135 Mbps was considered a fast optical network, and many channels of the plesiochronous digital hierarchy (PDH) on the network were significantly slower (no more than 45 Mbps). At this speed, a typical full-size 1500-byte (12,000-bit) data packet should load at a speed of 77.42 microseconds. In a low-speed channel, such as the T1 line of 1.544 Mbps, the transmission of such a packet took up to 7.8 milliseconds.
The delay in loading caused by several such packets in the queue can exceed the number of 7.8 ms by several times. This is unacceptable for speech traffic, which must have low jitter in the data stream supplied to the codec in order to produce good quality sound.
A packet voice system can do this in several ways, for example, such as using a playback buffer between the network and the codec. This allows you to smooth out the jitter, but the delay that occurs when passing through the buffer requires an echo canceller even on local networks. At that time, it was considered too expensive. In addition, it increased the channel delay and made interaction more difficult.
ATM networking technology inherently provides low jitter (and minimal overall latency) for traffic.
How does this help in a network connection?
ATM design designed for low jitter network interface. Nevertheless, “cells” were introduced into the project in order to provide short delays in the queues, while continuing to support datagram traffic. ATM technology split all packets, data, and voice streams into 48-byte fragments, adding a 5-byte routing header to each so that they could be reassembled later.
This size choice was political, not technical. When CCITT (currently ITU-T) standardized ATM, US officials wanted a 64-byte payload because it was considered a good compromise between the large amounts of information optimized for data transfer and shorter payloads designed for real-time applications. . In turn, developers from European countries wanted to receive 32-byte packets, because the small size (and, consequently, the short transmission time) simplifies voice applications with respect to echo cancellation.
A size of 48 bytes was chosen as a compromise between the two sides (plus header size = 53). 5-byte headers were chosen because it was believed that 10% of the payload was the maximum price to pay for routing information. ATM technology multiplexed 53-byte cells, which reduced data corruption and latency by almost 30 times, which reduced the need for echo cancellers.
ATM cell structure
ATM defines two different cell formats: user network interface (UNI) and network interface (NNI). Most ATM network links use UNI. The structure of each such package consists of the following elements:
- The Generic Flow Control (GFC) field is a 4-bit field that was originally added to support ATM interconnection in a public network. By topology, it is represented as a ring with a double bus distributed queue (DQDB). The GFC field was designed to provide 4 bits of User-Network Interface (UNI) for multiplex negotiation and flow control among cells of various ATM connections. However, its use and exact values ​​have not been standardized, and the field is always set to 0000.
- VPI - virtual path identifier (8 bits UNI or 12 bits NNI).
- VCI - virtual channel identifier (16 bits).
- PT - payload type (3 bits).
- MSB is a network management cell. If its value is 0, the user data packet is used, and in its structure 2 bits are an explicit indication of direct congestion (EFCI), and 1 is the experience of network congestion. In addition, 1 more bit is allocated for the user (AAU). It is used by AAL5 to indicate packet boundaries.
- CLP - cell loss priority (1 bit).
- HEC - Header Error Management (8-bit CRC).
The ATM network uses the PT field to denote various special cells for operations, administration, and management (OAM), as well as to determine packet boundaries at some adaptation layers (AALs). If the MSB value of the PT field is 0, this is a user data cell, and the other two bits are used to indicate network congestion and as a general-purpose header bit available for adaptation layers. If the MSB is 1, this is a management pack, and the other two bits indicate its type.
Some ATM communication protocols (asynchronous data transfer methods) use the HEC field to control the CRC-based framing algorithm, which allows you to find cells at no extra cost. 8-bit CRC is used to correct single-bit header errors and detect multi-bit. If the latter are detected, the current and subsequent cells are discarded until a cell without header errors is found.
The UNI package reserves a GFC field for a local flow control system or sub-multiplexing between users. This was intended so that several terminals could share the same network connection. Also, this technology was used with the aim that two telephones of a digital network with integrated service (ISDN) could use one basic ISDN connection at a certain speed. All four GFC bits must be zero by default.
The NNI cell format replicates the UNI format in much the same way, except that the 4-bit GFC field is redistributed into the VPI field, expanding it to 12 bits. Thus, one NNI ATM connection can process almost 216 VCs each time.
Cells and transmission in practice
What does ATM mean in practice? It supports various types of services through AAL. Standardized AALs include AAL1, AAL2 and AAL5, as well as the rarely used AAC3 and AAL4. The first type is used for constant bit rate (CBR) services and circuit emulation. Synchronization is also supported in AAL1.
The second and fourth types are used for services with variable bitrate (VBR), AAL5 - for data. Information about which AAL is used for this cell is not encoded in it. Instead, it is negotiated or configured at the endpoints for each virtual connection.
After the initial design of this technology, networks began to work much faster. A 1,500-byte (12,000 bit) full-size Ethernet frame requires only 1.2 ÎĽs for transmission on a 10 Gb / s network, which reduces the need for small cells to reduce latency.
What are the strengths and weaknesses of such a connection?
The advantages and disadvantages of ATM network technology are as follows. Some believe that increasing the speed of communication will replace it with Ethernet in the backbone network. However, it should be noted that an increase in speed does not in itself reduce jitter due to queuing. In addition, the hardware for implementing service adaptation for IP packets is expensive.
At the same time, due to the fixed payload of 48 bytes, ATM is not suitable as a data channel directly under IP, since the OSI level on which IP operates must provide a maximum transmission unit (MTU) of at least 576 bytes.
For slower or congested connections (622 Mbit / s and below), the use of an ATM network makes sense, and for this reason most asymmetric digital subscriber line (ADSL) systems use this technology as an intermediate layer between the physical link layer and the Layer 2 protocol, such like PPP or Ethernet.
At these lower speeds, ATM provides a useful ability to transfer multiple logical circuits on the same physical or virtual media, although there are other methods, such as multi-channel PPP and Ethernet VLAN, which are optional in VDSL implementations.
DSL can be used as a way to access an ATM network, allowing you to connect to many Internet service providers through a network of broadband ATMs.
Thus, the disadvantages of the technology are that in modern high-speed connections it loses its effectiveness. The advantages of such a network are that it significantly increases the bandwidth, since it provides a direct connection between various peripheral devices.
In addition, if there is one physical connection using ATM, several different virtual channels with different characteristics can function simultaneously.
This technology uses quite powerful tools designed for traffic management, which continue to develop today. Thanks to this, it becomes possible to simultaneously transmit data of various types, even if they present completely different requirements for sending and receiving them. So, you can create traffic through various protocols on one channel.
The basics of virtual circuits
Asynchonous Transfer Mode (abbreviation ATM) acts as a channel-based transport layer using virtual circuits (VCs). This is due to the concept of virtual paths (VP) and channels. Each ATM cell has an 8- or 12-bit virtual path identifier (VPI) and a 16-bit virtual channel identifier (VCI) defined in its header.
VCI, along with VPI, is used to identify the next destination of a packet when it passes through a series of ATM switches on its way to its destination. The length of the VPI varies depending on whether the cell is sent over the user or network interface.
As these packets go through the ATM network, switching occurs by changing the VPI / VCI values ​​(replacing the shortcuts). Although they are not necessarily consistent with the ends of the connection, the concept of the scheme is consistent (unlike IP, where any packet can reach its destination by a different route). ATM switches use the VPI / VCI fields to identify the virtual channel (VCL) of the next network, which the cell must transit on its way to the final destination. The VCI function is similar to the Data Link Connection Identifier (DLCI) function in the frame relay and logical channel group number in X.25.
Another advantage of using virtual circuits is the ability to use them as a multiplexing layer, allowing you to use various services (such as voice and frame relay). VPI is useful for reducing the switching table of some virtual circuits that share common paths.
Using cells and virtual circuits to organize traffic
ATM technology includes additional traffic movement. When a circuit is configured, each circuit switch is informed of the connection class.
ATM traffic contracts are part of the Quality of Service (QoS) mechanism. There are four main types (and several options), each of which has a set of parameters that describe the connection:
- CBR - constant data rate. Peak speed (PCR) is indicated, which is constant.
- VBR - variable data rate. Its average or stable value (SCR) is indicated, which can reach a peak at a certain level, at the maximum interval before problems arise.
- ABR is the available data rate. The minimum guaranteed value is indicated.
- UBR - Undefined data rate. Traffic is distributed over the remaining bandwidth.
VBR has options in real time, and in other modes it serves for “situational” traffic. Incorrect time is sometimes reduced to vbr-nrt.
Most traffic classes also use the Cell Tolerance Variation (CDVT) concept, which defines their “congestion” over time.
Data transfer management
What does ATM mean, given the above? In order to maintain network performance, traffic rules for virtual networks can be applied, limiting the amount of data transferred at the entry points to the connection.
The reference model approved for UPC and NPC is the common cell rate algorithm (GCRA). As a rule, VBR traffic is usually controlled using a controller, unlike other types.
If the amount of data exceeds the traffic determined by the GCRA, the network can either discard cells or mark the cell loss priority (CLP) bit (to identify the packet as potentially redundant). , ( ). , Partial Packet Discard (PPD) Early Packet Discard (EPD), , . .
EPD and PPD work with AAL5 connections because they use the end of the packet token: the ATM user interface indication (AUU) bit in the Payload Type field of the header that is set in the last SAR-SDU.
Traffic shaping
The basics of ATM technology in this part can be represented as follows. Traffic shaping typically occurs on a network interface card (NIC) in user equipment. At the same time, an attempt is made to ensure conditions where the cell stream on the VC will correspond to its traffic contract, that is, units will not be dropped or reduced in priority order in the UNI. Since the reference model defined for traffic control in the network is GCRA, this algorithm is usually used to generate and direct data.
Types of Virtual Circuits and Paths
ATM technology can create virtual circuits and paths, both statically and dynamically. Static circuits (PVA) or paths (PVP) require that the circuit consist of a series of segments, one for each pair of interfaces through which it passes.
PVP and PVC, although conceptually simple, require considerable effort in large networks. They also do not support service re-routing in the event of a failure. In contrast, dynamically constructed PVPs (SPVPs) and PVCs (SPVCs) are constructed by specifying the characteristics of the circuit (service “contract”) and two endpoints.
Finally, ATM networks create and remove switched virtual circuits (SVCs) at the request of the end piece of equipment. One of the applications for SVC is the transfer of individual telephone calls when the switch network is interconnected via ATM. SVCs were also used when trying to replace ATM local area networks.
Virtual routing scheme
Most ATM networks supporting SPVP, SPVC, and SVC use the Private Network Node or Private Network-to-Network Interface (PNNI). PNNI uses the same shortest path algorithm that OSPF and IS-IS use to route IP packets to exchange topological information between switches and select a route through the network. PNNI also includes a powerful summarization mechanism for creating very large networks, as well as a Call Access Control (CAC) algorithm that determines the availability of sufficient bandwidth on a proposed route through a network to meet VC or VP service requirements.
Reception and connection to calls
The network must establish a connection before both sides can send cells to each other. In ATM, this is called a virtual circuit (VC). This can be a permanent virtual circuit (PVC), which is created administratively at the endpoints, or a switched virtual circuit (SVC), created as necessary by the transmitting parties. The creation of the SVC is controlled by an alarm in which the requesting party indicates the address of the receiving party, the type of service requested and any traffic parameters that may be applicable to the selected service. Then, “Network” will confirm that the requested resources are available, and that the route exists for the connection.
ATM technology defines the following three levels:
- ATM Adaptations (AAL);
- 2 ATM, roughly equivalent to the OSI data link layer;
- physical equivalent to a similar OSI layer.
Deployment and Distribution
ATM technology became popular among telephone companies and many computer manufacturers in the 1990s. However, even towards the end of this decade, the best price and performance of Internet-based products began to compete with ATM for real-time integration and packet network traffic.
Some companies today focus on ATM products, while others provide them as an option.
Mobile technology
Wireless technology consists of an ATM core network with a wireless access network. Cells here are transmitted from base stations to mobile terminals. Mobility functions are performed on an ATM switch in a core network known as a crossover network, which is similar to MSC (Mobile Switching Center) GSM networks. The advantage of ATM wireless communication is its high throughput and high level 2 handover.
In the early 1990s, several research laboratories actively worked in this area. An ATM forum was created to standardize wireless technology. He was supported by several telecommunications companies, including NEC, Fujitsu and AT&T. Mobile ATM technology aims to provide high-speed multimedia communication technologies capable of providing broadband mobile communications in addition to GSM and WLAN networks.