Exploiting the bandwidth potential of multimode optical fibres

Historically the optical fibre was perceived to provide "unlimited" bandwidth, however, the capacity of current communications systems based on single mode optical fibre technology is very close to the limits (within a factor of 2) imposed by the physical transmission properties of single mode fibres. The major challenge facing optical communication systems is to increase the transmission capacity in order to meet the growing demand (40% increase year-on-year) whilst reducing the cost and energy consumption per bit transmitted. If new technologies are not developed to overcome the capacity limitations inherent in single mode fibres and unlock the fibre bandwidth then the growth in the digital services, applications and the economy that these drive is likely to be curtailed. The need for increased capacity in the core and metro areas of the network and within computing data centres is likely to become even more acute as optical access technologies, providing far greater bandwidths directly to the users, take hold and services such as ubiquitous cloud computing are adopted.
Multimode optical fibres (MMF) offer the potential to increase the capacity beyond that of current technologies by exploiting the spatial modes of the MMF as additional transmission paths. To fully exploit this available capacity it is necessary to use coherent optical (CO) reception and multiple-input multiple-output (MIMO) digital signal processing techniques analogous to those already used in wireless communication systems such as WiFi. This project aims to develop the technologies and sub-systems required to implement a CO-MIMO system over MMF that exceeds the capacity of current single mode fibre systems and reduces the cost and energy consumption per bit transmitted. To achieve this goal the project addresses the following key engineering challenges necessary to realise a complete system demonstrator.
Engineer the channel: The multimode optical fibre MIMO channel, unlike its wireless counterpart, presents the opportunity to engineer the optical channel to optimise its performance for MIMO operation by designing and fabricating new optical fibres, using proven solid core technology, to maximise the MIMO capacity of the fibre.
Dynamically control the channel: The transmission characteristic of the multimode optical fibre channel varies with time. We will exploit both the flexible and fast adaptive nature of digital signal processing, and the less energy intensive and slower adaptation of liquid crystal spatial light modulator based optical signal processing to compensate for the channel variation and recover the spatially multiplexed data channels.
Employ energy efficient optical amplification: In order to reduce both the energy consumption and cost per bit and to extend the propagation distance into the hundreds of kilometres region it is essential to develop optical fibre amplification technologies that provide amplification to multiple spatial and wavelength channels and thus share the cost.
Coherently detect the optical signal to exploit the wavelength and spatial domains: The developed system will combine spatial multiplexing with existing dense wavelength division multiplexing, polarisation multiplexing and multilevel modulation techniques to maximise the capacity. The key to achieving this is the use of coherent optical detection and digital signal processing, which is essential not only to fully exploit the spatial capacity of the MMF channel, but also facilitates the use of existing multiplexing techniques that are difficult to realise in conventional multimode transmission systems.
The technologies and systems developed within this project will find applications, ranging from capacity upgrades of existing MMF data networks in data and computer processing centres, through to the installation of new high capacity metro and long haul fibre transmission systems using the MIMO optimised fibres and technologies developed in this project.

All members of society will benefit from improved access to the communications infrastructure that this research will enable. In particular the continuing growth in the digital economy with its requirement for universal access and increased bandwidth to support services such as cloud computing requires the development of new communications technologies, such as those proposed here, to provide increased capacity at lower cost and with reduced energy consumption. Cloud computing in particular has the potential to significantly decrease the energy footprint of computing, but its successful adoption is entirely dependent on a low cost, high capacity and reliable communications infrastructure between the users and the computing centre. It also requires energy efficient, high capacity links for interconnection of computing resources within the data centre. Of particular importance here is low energy dissipation and small form factor as the data centre environment is space constrained. The use of high capacity MMF transmission technology developed within this proposal will be particularly attractive in this application particularly as these centres already have installed MMF links and this represents an attractive capacity upgrade path. Moreover, simply adding more fibre to meet the capacity in metro and long haul communications does not reduce the cost or energy consumption.
To ensure that economic and societal benefits are realised from this work it is essential to facilitate the commercialisation of the technologies and subsystems that will be developed in this proposal. To achieve this we will contribute to the standardisation activities in this area to ensure that the systems developed are able to be widely deployed and work closely with our industrial project partners to commercialise the subsystem technologies. The industrial project partners and supporters have been carefully chosen to have interests and capabilities that are closely aligned across the full range of outcomes from this project. This includes: RSoft an optical systems simulation software provider with whom we will work to distribute the models developed; Oclaro an optical components manufacturer with a particular interest in integrated optical transmitters and optical amplifiers; SPI Lasers who in collaboration with Southampton develop and manufacture specialised optical amplifiers and have devices that may be well suited to multimode amplification; PMC Sierra a semiconductor chip designer and manufacturer for communications applications with interest in the implementation of advanced DSP for coherent optical receivers.