What is intersymbol interference? What types of dispersion affect optical fiber? How can dispersion be minimised? And what performance standards are required of multimode optical fiber solutions today?
Intersymbol Interference (ISI) is the spreading of a ‘bit’ into the adjacent ‘bit’ periods. Ideally, a bit would consist of a square wave that instantly reaches peak output at the beginning of a bit period, and drops instantly to zero at the end of the period.
In the real world, the pulse is dispersed, or spread out, so that it overlaps into adjacent ‘bit’ periods. If this overlap is bad enough, the receiver cannot accurately detect the presence or absence of a ‘bit’. This dispersion increases as the link distance increases.
Within optical fiber, bits of data are represented by pulses of light. Each pulse of light will spread, or disperse, over time as it travels along the length of the fiber. When these spreading pulses overlap, intersymbol interference results. The less intersymbol interference, the greater the fiber’s capacity to transmit information.
ISI was not a critical issue before gigabit Ethernet. Bit periods, the amount of time occupied by one bit, were so large that pulse spreading was not an issue. With the introduction of gigabit Ethernet, bit periods decreased to 800 picoseconds (10,000,000,000 bits/1second).
The extremely short bit period makes high bandwidth fiber essential and the bandwidth of a fiber is ultimately determined by its pulse spreading, or optical fiber dispersion characteristics.
Chromatic dispersion describes the tendency for different wavelengths to travel at different speeds in a fiber. A light source is rarely pure, in that it is actually a composition of wavelengths of various intensities surrounding a central wavelength e.g. 850nm. If operated at wavelengths where chromatic optical fiber dispersion is high, optical pulses tend to broaden as a function of time or distance and cause intersymbol interference.
To minimize this, it is desirable to operate at wavelengths where the optical fiber’s chromatic dispersion is small or use a light source, such as laser technology, which consists of only a few wavelengths. Although multimode fiber exhibits relatively high chromatic dispersion at the 850 nm wavelength, the use of controlled launch lasers in gigabit networks and the distances considered in the LAN minimize the effects.
The majority of the dispersion in optical fiber systems is caused by modal dispersion. Modal dispersion exists because the different light rays (modes) have a different optical path length along the fiber, therefore rays entering at the same time will not leave the far end of the fiber at the same time.
This dispersion effect limits the bandwidth as shown in the illustration above. Modal bandwidth is strongly influenced by the fiber core size and the shape and smoothness of the refractive index profile across the fiber core.
In multimode fiber design, fiber cabling types must meet minimum performance levels based, in broad terms, on bandwidth and attenuation. Fiber cabling systems with certain performance levels are referenced in applications standards such as IEEE 802.3 for Ethernet.
The performance needed to effectively support the most cost effective implementations of systems such as 10Gbps Ethernet is attributed to careful attention to the light propagation properties of the fiber. With LOMMF or OM3/4 fiber, the intermodal delay characteristics of the fiber are characterized by measuring what is known as its differential modal delay (DMD).
The cost of the pluggable optics continues to limit the implementation of singlemode fiber (SMF) in buildings and data centers. Although new technologies and manufacturing efficiencies are helping to reduce prices for SMF, it is still not enough to justify the high cost of singlemode optics. Two areas where use of SMF is increasing are from the entrance facility to the main distribution area and for extreme scale in mega-data center designs.
For the enterprise, multimode fiber (MMF) continues to offer a more attractive balance of performance, density and cost. The challenge for MMF is distance. As data traffic grows and interconnectivity speeds increase, the maximum distance for a communication link tends to decrease. But emerging higher quality components and engineered links can provide the link capacity to.
The recent introduction of OM5 may eventually provide the optimum solution for fiber migration. Introduced by CommScope in 2015, OM5 was recently approved under ANSI/TIA-492AAAE and is expected to be recommended by ANSI/TIA-942-B. The new fiber enhances the ability of short-wavelength division multiplexing (SWDM) technology to provide at least a four-fold increase in usable bandwidth. It also supports all legacy multimode applications by maintaining compatibility with OM3 and OM4 fiber. By multiplexing four wavelengths spaced in the 850–950 nm region, one strand of WBMMF can increase the data capacity by a factor of four.
If you have an interest in fiber optics; be that in communication networks, specification, termination, cabling in or between buildings in any capacity and need to understand the principals, then our SP4420 Fiber Optic Infrastructure course could be of interest to you.
Of course, Fiber Optic Infrastructure is just one of multiple aspects of passive infrastructure.
Facilitating ever growing and expanding networks, passive infrastructure is set to increase as demand for global telecommunications infrastructure advances. To stay ahead of demand, you can read more in our eBook: Understand the Passive Infrastructure that Underpins Your Network.