Research

                                       Optical Fiber Communications

                                              -Code divisions multiple access

                                            –Coding and signal processing

                                     Mobile wireless communications

                                                               -MIMO

                                    Software defined radio prototyping

____________________________________________________________________________________________________________________________________

Coded Communications over Optical Fiber

Demands for fiber-optic communication are rapidly approaching limits posed by available bandwidths.  The most efficiently useable fiber-optic spectrum is already in sufficiently high demand that the cost per unit bandwidth could increase as more applications are introduced. Meanwhile, fiber-optic bandwidth efficiency continues to hover near 1.0 bps/Hz, an order of magnitude less than that achieved by wireless technology.

Multiple access optical waveforms are encoded with signature sequences that permit the receiver to separate them and select one (or more) for extraction of information. This permits many user signals to share channel resources (spectrum, time, and physical space) on a mostly non-interfering basis.  In fact, however, a small amount of energy from one user’s signal usually interferes with another’s.  This multiple access interference (MAI) effectively reduces the signal­-to-noise ratio of the desired user’s signal, increasing the bit error rate.

Physical impairments such as dispersion or nonlinearity often increase the MAI and must be dealt with or prevented.  Many legacy multiple access systems leave the MAI in a “scrambled” state in the hopes that it will not degrade the intended user, but the MAI remains present and generates beat noise in the optical detector circuits.

Currently, we are investigating two optical signal schemes called SPOT and SLIP that cancel the MAI in the receiver before it reaches the detector circuits.  (Write me if you want an explanation of the acronyms!)

An early and comprehensive exposition of SPOT can be found in this paper: A.B. Cooper III, J.B. Khurgin, S. Xu, and J.U. Kang, “Minimizing Multiple Access Interference in OCDMA with Phase and Polarization Diversity,” IEEE Journal of Selected Topics in Quantum Electronics, special issue on “Code in Optical Communications and Networks,” v. 13, no. 5, Sep/Oct 2007, pp 1386-1395.

 

MIMO Wireless

Wireless networks constructed from transceivers with several antennas afford many powerful configurations to improve network performance. Such multiple-input multiple-output (MIMO) schemes can perform signal processing and beamforming functions or can create multiple spatial antennas between two points, depending upon the system design requirements.

Over more than a decade, cognitive radio algorithms have emerged to solve the scarcity of radio spectrum resources via adaptive spectrum sharing methods that recognize asymmetric rights to use the spectrum while achieving throughputs that are not possible in fixed, non-adaptive schemes. In addition, in applications such as public safety and tactical networks, highly capable multichannel radios are being widely deployed to facilitate, among other things, rapid and efficient network reconfiguration.

It is natural to consider, therefore, merging these two capabilities into a cognitive MIMO network in which network nodes can adapt their parameters to communications requirements and signal/network environments in order that the system users can communicate effectively in spite of interference, changing network topology, and hostile electromagnetic activity.

The component subject areas of cognitive MIMO networking have been widely investigated and are the objects of very active research. We are interested in studying tradeoffs among multiple antennas, multiple channels, transmission schedules, and radio link configurations and for considering the mutual impacts of network reconfigurability and resource allocation.

To that end, we are creating a small network testbed to study experimentally MIMO and multichannel transmission in a controlled environment.  This testbed is expected to support a larger modeling and simulation effort to study the joint problems of reconfiguration and resource allocation.