Synchronous Optical Networks, or SONET, and Synchronous Digital
Hierarchy, or SDH, standards were introduced in the mid 1980’s. These
technologies set new industry standards for transport with payloads
larger than a T3. Prior to these standards, higher capacity transports
used proprietary protocols and equipment. Standardizing protocols would
allow for equipment from various vendors to interface, as well as allow
different carriers to interface with each other.(1)
The American National Standards Institute (ANSI) and the International
Telecommunication Union-Telecommunication Standardization Sector (ITU-T)
had settled on different schemes for their versions of high capacity
optical transport. ANSI had considered different payloads and frame sizes
before settling on 50.688 Mbits/sec transfer rate in known as Synchronous
Transport Signal Number One or STS-1 for short. European groups preferred
a base rate of 155.52 Mbits/sec which was proposed by ITU-T. Through the
disagreement, ANSI altered the STS-1 structure to 51.84 Mbits/sec. Multiplexing
the STS-1 by an integer of 3 would equal the ITU-T standard of 155.52 Mbits/sec
and allow for interfacing between the 2 different standards.(1)
Once SONET/SDH standards were set, carriers began implementing networks
across the North America and around the world. As the internet and computer
networking needs grew, so did the need for higher bandwidth WAN links. The
development of Packet over SONET/SDH (PoS) standards made use of the vast
existing SONET/SDH networks that carriers already had in place.
PoS encapsulates IP traffic with a Point-to-Point Protocol (PPP) header. The encapsulated traffic is then placed in a High-Level Data Link Control (HDLC)-delimited SONET Synchronous Payload Envelope (SPE) then transported across the SONET(2). Figure 1 below illustrates an IP frame encapsulation into a PoS frame.
A PPP frame contains 3 fields:
Figure 2 illustrates a PPP frame within an HDLC frame.
Figure 3 illustrates the frame information contained in a PoS Frame. Point-to-point connections only contain 2 end points; therefore the destination is always the other end point of the connection. For this reason, all frames are broadcast and the address field of a PoS frame will always be 0xFF. A control field value of 0x33 indicates an HDLC frame. The PPP frame is inserted in the Information field of the frame and is a variable size due to the variable MRU size. The trailer of the frame is a 16 or 32 bit Frame Check Sequence (FCS). The beginning and ending of the frame are denoted with a 0x7E value delimiter. (2)
Typical ethernet line rates operate at 10, 100, 1000, or 10000Mbit/sec. SONET/SDH bandwidth is based on the frame standards set by ANSI and ITU-T. Table 1 shows data rates for different SONET/SDH designations.
Using standard SONET/SDH designations can provide some issues when trying to
build circuits with certain line rates. Here are a few examples:
10Mbit/sec line rate would require an STS-1 or STM-0 circuit at 51.84Mbit/sec
100Mbit/sec line rate would require an STS-3 or STM-1 circuit at 155.52Mbit/sec
1,000Mbit/sec line rate would an STS-24 or STM-8 circuit at 1,244.16Mbit/sec
As you can see from these examples, this is inefficient use of bandwidth. To
improve the data transport efficiency, Virtual Concatenation (VCAT) is used.
VCAT allows grouping of Synchronous Payload Envelopes (SPE) into VCAT groups.
The SPEs become VCAT members and do not have to be consecutive. These members
act as individual circuits along the SONET/SDH paths. Table 2 below shows how
VCAT circuits are a more efficient use of bandwidth for certain line rates.(7)
The above information references high order VCAT. Low order VCAT can be used
sub STS SPEs known as Virtual Tributaries (VT). There are four different VT sizes.(6)
VT1.5 = 1.728 Mbits/sec
VT2 = 2.304 Mbits/sec
VT3 = 3.456 Mbits/sec
VT6 = 6.912 Mbits/sec
Utilizing low order VCAT enables smaller sized circuits that can be provisioned much
closer to the requested bandwidth. An example of this would be a VT1.5-7v circuit
giving a line rate of approximately 10 Mbits/sec when accounting for overhead. This
is much more efficient than an provisioning an STS-1 as shown above.
SONET is typically deployed in a ring architecture. In a ring architecture there are 2 optical paths between any 2 nodes on the ring. Figure 4 shows a simple ring diagram with 4 nodes. All nodes are connected to the adjacent nodes via 2 fibers, 1 transmit and 1 receive. The diagram shows the 2 paths traffic can take between node A and node C.(3)
SONET uses Automatic Protection Switching (APS) to route traffic based on ring conditions. The transmitting node transmits signals in both directions on the ring, one direction is considered the working path while the other is considered the protect path. The receiving node accepts the signals from the working path and ignores the signals from the protect path. If a fiber cut or transmitter failure happens on the working path, the receiving node switches the circuit and accepts signal from the protect path.(3) There are different protection schemes that can be configured on SONET rings such as Bidirectional Line Switched Rings (BLSR) and Unidirectional Path Switched Rings (UPSR). While the protection schemes work differently, both provide protection in the event of a fiber cut. Figure 5 shows the working and protect UPSR selectors for a circuit.
Another form of reliability for PoS circuits is Link Capacity Adjustment Scheme (LCAS). LCAS comes into play on circuits built with virtual concatenation (see PoS Bandwidth). LCAS can be applied to Virtual Concatenation Groups (VCG) to provide fault tolerance protection. Without LCAS, one tributary of a VCG taking errors causes the entire data stream to have errors. LCAS detects the tributary that has errors, and temporarily removes it from the VCG. This allows the VCG to run error free at a reduced bandwidth.(4)