Second wind of multimode

    With the increase in the number of devices connected to networks and the amount of data generated by them, the bandwidth requirements for network infrastructures are also growing. The main burden of traffic transmission in almost all networks today is fiber-optic systems. Moreover, the solutions based on multi-water fiber (MMW) remain the most economically attractive, especially for communication over short distances. Recently, along with the development of MMB broadband and SWDM technology, fundamentally new opportunities have appeared to increase the throughput of systems based on MMB.

    Recall that a light-carrying core in a multimode fiber has a diameter about six times larger than in a single-mode fiber (OMW). This makes fiber alignment and alignment easier — an important task for developers of connectors, as well as light sources and receivers. In many ways, this is why MMB became the first type of fiber that began to be used in communication networks - back in the early 80s of the last century. And only in the late 80s, when it became possible to provide alignment with an accuracy of the order of a micron and laser diodes appeared, single-mode fiber became widely used in communication networks.


    The structure of a typical optical fiber

    But, despite the advantages of single-mode technology in range and bandwidth, MMV has remained the main type of fiber for most networks, including LANs, data center networks, etc. This is due to the price advantages of multimode technology, due to the already mentioned simpler fiber alignment, widespread availability of low-cost radiation sources, and other reasons.

    The multimode has come a long way in improving throughput. It all started with LED emitters (LEDs) and megabit speeds. In the 90s, when higher speeds were required, LEDs began to give way to new inexpensive light sources - VCSEL lasers with a wavelength of 850 nm, which are able to modulate the signal much faster. This, in turn, led to the transition from MMW with a core diameter of 62.5 μm (cable systems of class OM1) to fibers with a core of 50 μm (class OM2).
    In the late 90s, the era of gigabit speeds came. A further increase in bandwidth was required. It was provided by new multimode fibers, which were originally developed optimized for laser transmission (LOMMF). The first standard LOMMF fibers provided a bandwidth of about four times that of OM2 fibers. So there was a new class of fibers - OM3, which opened the door to 10-gigabit systems in the early 2000s.

    The next stage is the development by the end of the decade of OM4 fiber with an even wider broadband coefficient. This fiber provided a linear speed of 25 Gbit / s. In addition to increasing the linear velocity, in order to achieve ever higher speeds, an increase in the number of fibers forming the communication channel was required. So, when using a linear speed of 10 Gbit / s, four fibers are required to form a 40G channel (eight for duplex), for a 100G channel, 10 fibers (20 for duplex). When switching to a linear speed of 25 Gbit / s, the number of fibers for a 100G channel is proportionally reduced (up to eight for a duplex), however, for the implementation of promising 400G systems, 32 fibers are already needed.

    To simplify the organization and maintenance of multi-fiber systems, group MPO connectors are increasingly being used. The compact MPO design allows you to terminate 8, 12, 16 and more fibers in the space corresponding to the duplex LC connector. The high density MPO makes it possible to deploy a pre-terminated cable system with a large number of fibers, eliminating the lengthy process of installing connectors in the field.

    However, increasing the throughput of multimode systems by increasing the number of fibers is a way that has many disadvantages. An increase in the number of fibers increases the complexity of the system, imposes increased requirements on cable channels, means of laying fibers in the space of the switching field, etc. However, until recently, multimode technologies did not involve an elegant way to increase throughput, long known in the world of single-mode technology. We are talking about spectral multiplexing (WDM), when many spectral channels at different wavelengths are formed in one fiber. Accordingly, the fiber throughput is multiplied by the number of such channels. In single-mode systems, dozens of spectral channels can be used.


    Spectral Compression Principles (WDM)

    Why was this seal not used in multimode technology? Everything is very simple. OM3 and OM4 fibers are optimized for laser transmission at the same wavelength - 850 nm. “Step to the left, step to the right” (on a spectral scale) - the throughput of such a fiber drops sharply, and it is no longer suitable for transmitting high-speed flows. Therefore, the implementation of spectral multiplexing in the MMM required the development of a new fiber capable of providing effective throughput in a relatively wide “window” of wavelengths.

    The first samples of the new fiber appeared several years ago. In the spring of 2015, at an Optical Fiber Communications (OFC) conference, Finisar and CommScope demonstrated the WDM technology on the new MMB, dubbed Broadband (ШМ-ММВ). The transmission of four spectral channels was shown (at 850, 880, 910, and 940 nm), each of which provided a bandwidth of 25 Gbit / s, and in aggregate - 100 Gbit / s. The corresponding spectral multiplexing technology is called SWDM - Short Wavelength Division Multiplexing.

    Earlier, in the fall of 2014, CommScope, together with the same Finisar and a number of other companies, initiated a project to develop a standard for a new fiber at the TIA Association. In June 2016, the TR-42.12 subcommittee, responsible for optical fibers and cables at the TIA Association, approved the ANSI / TIA-492AAAE standard, which defines the WB-MMF (WB-MMF). The document describes a laser-optimized fiber for transmitting signals at a single wavelength or at several wavelengths in the range from 850 to 953 nm. And a little later, in October 2016, at a joint meeting of relevant committees of ISO and IEC organizations, it was decided to classify ШП-ММВ as OM5.

    The efforts of cable manufacturers were picked up by manufacturers of active network equipment. The SWDM Alliance Industrial Consortium was jointly formed to develop specifications and promote Shortwave Wavelength Division Multiplexing technology. (The founders of the SWDM Alliance are Commscope, Corning, Dell, Finisar, H3C, Huawei, Juniper, Lumentum, and OFS.) The SWDM has already published two specifications for 40- and 100-Gigabit Ethernet transmission (40 GE SWDM4 and 100 GE SWDM4, respectively) . SWDM technology enables high-speed 40G and 100G channels using just a couple of OM5 fibers. It also opens up the possibility of efficient implementation of 200G, 400G and 800G Ethernet channels based on multimode fiber.


    Various transmission options for high-speed flows, including the use of ШП-ММВ and spectral multiplexing (WDM)

    So, the advent of ШП-ММВ, or OM5 fiber, marks a significant breakthrough in the development of multimode technology. Now, already high transmission rates (for example, 100G) can be realized with a significantly smaller number of fibers. In addition, it became possible to switch to higher speeds without the need to use additional fibers. In general, technical progress is impressive, due to which the throughput of multimode fiber has increased 160,000 times, from 10 Mbps to potentially 1600 Gbit / s, while maintaining the main advantages of the multimode - low cost.

    Also popular now: