Originally developed at Cambridge University, the Cambridge Ring composes the basis for many of the commercially marketed ring networks available in the UK. As the name suggests it permits communication between a number of computers, terminals or intelligent device controllers, connected in such a way as to form a closed loop. The actual physical connection is normally twisted wire pair and is used to carry 16 bits of data. It should be pointed out that each network station regenerates the data packet before forwarding it to the next station. Below is an illustration of a typical station and the means by which it connects to the ring:
Firstly, repeaters are used to regenerate the digital signals transmitted round the ring and are actually powered from the ring itself rather than the local power supply. Although they usually form a component part of a station they can occasionally be required between stations if the distance is greater than 100 metres. The repeater decodes the phase modulated signals on the ring into clock and data signals for the Access Logic. However, if the repeater is not part of a station it simply allows packets to pass through.
As can be seen from the diagram the Access Logic is connected to the ring via a repeater. In addition to modulating and demodulating the clock and data signals from the repeater into a parallel interface to the host the other functions carried out are as follows:
Every ring network requires a monitor station and the Cambridge Ring is no exception. In this case it sets up and maintains circulating packets and detects and corrects various error conditions. Moreover, it sets the number of packets to be used and the size of the gaps between packets. Lastly, it monitors the number of times packets pass it - more than twice and they are deleted.
The structure of a packet in the Cambridge Ring is shown below:
The function of the various bits in the ring packet are as follows:
Bits Function 1 Marker bit to indicate start of packet 2 Indicates whether packet is full or empty 3 Indicates whether packet has passed the monitor station 4 - 11 Destination address 12 - 19 Source address 20 - 35 Data (16 bits) 36 - 37 Response bits completed by receiver (accepted, rejected, busy ignored) 38 Parity bit
Initially, when the ring is concepted, the monitor station places at least one packet in transit. It should be noted that data travelling around the ring is delayed both at the stations and on the cable - the delay provides temporary data storage. Therefore, the number of packets which can be established on the ring is determined by the length of the cable and the number of stations.
In order for a station to transmit data it must wait for an empty packet, indicated by the full/empty bit being set to zero. Next, it stores the data in the packet, sets the F/E bit to one, indicating it is full, and places the packet on the ring. The packet circles the ring until it arrives at the receiving station, determined by the destination address. If the receiver accepts it, it takes a copy of the data and sets the response bits to accepted, otherwise it sets the response bits accordingly. The packet then continues to the original sender, where it is checked for consistency with the packet originally sent. If all is well, the F/E bit is set back to zero thus returning the packet to an empty status. However, this empty packet cannot be reused immediately, as it must be passed around the ring. This helps to prevent one station from employing an unfair proportion of the network's resources.
Only one packet from each station is allowed on the network at any one time. However, since the ring operates at a high speed, any delay in waiting to transmit another packet should be slight. Although the ring has a channel speed of 10 Mbits/sec, the actual data transfer rate is much slower. In fact only 16 of the 38 bits in a packet are available for data, effectively reducing the data rate to 4 Mbits/sec.
A Cambridge Ring can have a variety of servers attached to it in order to provide a variety of functions. Although the diagram below shows a possible configuration, it is unlikely that all the servers shown would be attached to any one network.
