100VG-Any LAN

The 100VG-AnyLAN is a new high-speed network technology, currently being defined by IEEE as IEEE 802.12 standard that provides a data rate of 100Mbps on 4-pair unshielded twisted-pair (UTP) cable. Future implementations also support 2-pair UTP, 2-pair shielded twisted-pair (STP), and fiber-optic cabling. The 100VG-AnyLAN technology supports all of the network design rules and topologies of 10 Base T Ethernet and token ring networks. These features allow organizations to upgrade their existing network and cable infrastructures to higher transmission speeds.

100VG-AnyLAN uses a centrally controlled access method referred to as demand priority. This access method is a simple, deterministic request method that maximizes network efficiency by eliminating network collisions and token rotation delays. Also, the demand priority protocol uses two levels of priority for each user request to guarantee support for emerging time-critical multimedia applications such as real-time video and audio for video conferencing or interactive video.

100VG-AnyLAN also offers message-frame compatibility with 802.3 Ethernet and 802.5 token ring networks. Frame compatibility allows the user to transparently migrate existing network operating systems and user software applications to a 100VG-AnyLAN network. Also, the frame type compatibility allows 100VG-AnyLAN to connect to existing Ethernet and token ring networks via a simple bridge. 100VG-AnyLAN may also route to FDDI and ATM backbones and wide area network (WAN) connections.

The 100VG-AnyLAN is compatible with the ISO/ IEC and IEEE architectural models, which separate the network functions into sublayers. As illustrated in Figure 100VG-AnyLAN is composed of a Media Access Control (MAC) sublayer, a Physical Medium Independent (PM!) sublayer, and a Physical Medium Dependent (PMD) sublayer. Nodes of a 100VG-AnyLAN connected to a hub.

The Media Access Control (MAC) Sublayer

The 100VG-AnyLAN media access control layer performs the following functions. They are:

• Demand priority protocol control
• Link training
• MAC frame preparation

Demand priority is an access method used by the MAC layer, in which each node requests or demands the hub to send a packet with a particular priority label. The priority label in a packet can be a standard priority label or higher priority label. The application software of the higher-order layer usually does the labeling of packets. Higher priority requests always considered before normal priority ones.

The hub performs a round robin cycle with all the nodes connected to it. Hence, it allows each node to send one packet if the other nodes have pending requests. A hub connected to 'n' nodes can transmit a maximum number of 'n' packets during the round robin process if no higher priority packets are pending. Each hub maintains a separate list of standard and higher priority packets. Regular priority requests are serviced in the increasing order of port number until any higher priority request arrives. As soon as a higher priority packet arrives, the hub completes the current transmission and start transmitting the higher priority ones. After completing all higher priority requests, it resumes the transmission of the standard priority packets.

Link training is performed to inform the hub to know about the nodes connected to it, its type (hub node, PC, network-monitoring equipment, router, bridge, etc.), its operating mode (standard, a monitor), and its address in the network. Link training is usually performed at the beginning by the node when the hub and the node powered. The node initiates the training by sending a test packet to the hub. Then the Hub and the hub exchange a couple of test packets to test the link, whether the wire connections appropriately arranged, and the data is transmitted without errors.

The MAC frame preparation completed after receiving the packet from the Logical Link Control sublayer. The MAC sublayer adds the node source address, and stuffs any extra bits (pad), if required, to complete the missing data field. A Frame Check Sequence (FCS) is then calculated and appended to the end of the packet. The FCS used by the receiving hub and node to determine if the packet received without errors.

Physical Medium Independent (PMI) Sublayer
The PMI layer performs the following three functions.
• Quartet channeling
• Scrambling
• 5-Bit to 6-Bit encoding (5B6B)

PMI also adds start frame and end frame delimiters to prepare the packet for transmission by the Physical Medium Dependent (PMD) sublayer. The figure illustrates the functions of the PMI layer.

Quartet channeling is the process of sequentially dividing the MAC frame data octets into 5-bit data quintets and sequentially distributing them among four channels. Each of the four channels represents a transmission pair of the 4 pair Unshielded Twisted-pair Cable. Channel-O data transmitted on twisted-pair wires 1 and 2, channel-l data transmit on wires 3 and 6, channel-2 data transmitted on wires 4 and 5, and channel-3 data transmitted on wires 7 and 8. 2-pair and fiber optic 100VG-AnyLAN networks use a multiplexing scheme, implemented in the PMD sublayer, to convert the four channels into either two channels or one channel, respectively. The figure illustrates the quartet channeling process.

Data scrambling is the process of scrambling the 5-bit data quintets, using a different scrambling mechanism for each channel to randomize the bit patterns on each transmission pair. Scrambling each channel eliminates repetitious data patterns such as all 0’s or all 1’s. 5B6BEncoding is the process of encoding or mapping the scrambled 5-bit data quintets into predetermined 6-bit symbols. This process creates a balanced data pattern, containing equal numbers of as and Is to provide guaranteed clock transition synchronization for receiver circuitry.

This encoding provides an error-checking capability. Invalid symbols and invalid data patterns of continuous Is or Oscan be easily detected. Preamble, Start Frame Delimiter, and End Frame Delimiter are added to each channel to package the data into a format ready for transmission on the network.

Physical Medium Dependent (PMD) Layer

The physical medium dependent sublayer functions include channel multiplexing (for the 2-pair and fiber-optic implementations only), NRZ encoding, link ordinary operation, and link-status control. 100VG-AnyLAN technology support four types of cabling media. They are:

1. 4-pair Unshielded Twisted-pair
2. 2-pair Unshielded Twisted-pair
3. 2-pair Shielded Twisted-pair
4. Single or Multimode Optical Fiber

A 100VG-AnyLAN network using 4-pair unshielded twisted-pair cabling uses a 30 MHz clock, can transmit 30 Mbps of data on each of the four pairs and can transmit data at the rate of 120 Mbps per second. At the receiving end, the 30 Megabits of encoded data are received and decoded into 25
Megabits of the original data, resulting in an effective data rate of 100 Mbps. Full duplex operations are required to communicate link-status control information between the hub and the node, and half duplex operation is used by all four channels to transmit data from the node to the hub or receive data from the hub to the node.

The data flow in a 100VG-AnyLAN network can explained in the following manner. The upper layer software called logical link control sublayer of the end node informs the media access Sublayer that it has a packet to send on the network. After receiving the packet, the media access control sublayer adds the source address and any required pad bits to complete the data field. The source node sends a standard priority request (NPR) to the hub, requesting to send a standard priority packet on the network.

The hub, during its round-robin scan procedure, selects the source node and alerts all potential destination nodes on the network segment that a packet may destined for them by sending an Incoming (INC) message. The potential destination nodes stop sending control tones, clearing the link to allow the node to receive the packet on all four channels.

Meanwhile, the source node detects that the link is clear and forwards the message packet from the media access control sublayer to the physical medium independent sublayer to prepare the data for transmission. The physical medium independent sublayer separates the data into four channels, scrambles the five-bit data quintets, and encodes the quintets into six-bit symbols. The preamble, start frame delimiter, and end frame delimiter added to each channel. The physical medium dependent sublayer begins to send the packet to the hub, using NRZ encoding. As the hub receives the packet, it decodes the destination address. The packet then routed to the various nodes with the matching destination address. At the same time, the hub stops sending INC and begins to send 'Idle tone' to the other nodes. All other nodes then resume sending requests, or idle, to the hub.

Structure of a 100VG-AnyLAN

The structure of a 100VG-AnyLAN shown in Figure. The Hub is an intelligent central controller that manages the network access by continually performing a scan of its network port requests. It checks for service requests from the attached nodes. The hub receives the incoming data packet and directs it only to the port with a matching destination address providing underlying network data security.

Each hub may be configured to support either 802.3 Ethernet or 802.5 Token Ring frame formats. All hubs located in the same network segment must configure for the same frame format. A bridge may be used either to connect a 100VG-AnyLAN network, using an 802.3 frame type to an Ethernet network, or a 100VG-AnyLAN network, using an 802.5 frame type to a token ring network. A router may be used to connect a 100VG-AnyLAN network to FODI and ATM networks or WAN connections.

Each hub includes one uplink port and "n" number of downlink ports. The uplink port functions as a node port but reserved for connecting the hub (as a node) to an upper-level hub. The "n" downlink ports are used to connect to 100VG-AnyLAN nodes. Each hub port may be configured to operate in either a normal mode or a monitor mode. Under normal mode, a port forwards all packets addressed to the nodes that are attached to it. Ports configured to operate in a monitor mode, forwards all packets that the hub receives. The standard and monitor mode configuration may be automatically learned in case if the hubs are connected through cascaded ports (uplink or downlink to another hub) or manually configured for a port connected to network-monitoring equipment.

A node may be a client or server computer, workstation, or other 100VG-AnyLAN devices such as a bridge, router, switch, or hub. Hubs connected as nodes are referred to as lower level, such as level 2, or level 3 hub devices. (Up to three levels of cascading may use on a 100VG-AnyLAN network.) A node issues requests to the hub to initiate link training and to send a packet onto the network. The 100VG-AnyLAN nodes also respond to incoming message commands from the hub.

The link connecting the hub and the node may be 4-pair UTP cable (Category 3, 4, or 5), 2-pair UTP cable (Category 5),2-pair STP cable, or fiber-optic cable. The maximum length of the cable from the hub to each node is 100 meters for Category 3 and 4 UTP, 150metres for Category 5 UTP and STP and 2000 meters for fiber-optic cable.