This paper describes the architecture of a low-latency symmetric multiprocessing optical soft memory system to
cluster computing inside the core of an adaptive optical signal processor with the aid of soft decision algebraic
polynomial algorithms. The optical system hardware is shown to evolve along with the iterator instantiations of the
soft algorithm that forms the core of the memory map. The system enables efficient cache coherence protocols used
in unit multiprocessors to be run across a homogeneous cluster in optical soft memory systems. We define a
structure called the Optical Generalized Viterbi Algorithm Data Structure (Optical GVA DS) that makes up the
system map for adaptive optical signal processing. The system executes transforms where algorithms for handling
the entire data vector is processed, shortening the computational complexity effectively. Thus the optical soft
memory system as described by the evolving Optical GVA DS iterator instantiates enables the design of parallel
processors to handle modulated data in the optical domain. This is of importance in the realization of distributed netcentric
architectures and forms the basis of large-scale real-time data processing and acquisition in m2m units.
This paper discusses about the extension of Gabor Expansions to the optical domain and the design of an efficient filter bank to provide adaptive equalization in the light of Optical Signal Processing. The isomorphism between this localized linear operator and the filter design fundamentals are examined in the framework of image sequence compression. A new and efficient technique to perform Gabor expansion of Optical signals is introduced. The multi-resolution representation of data is considered in particular. A new approach to filter bank design in optical domain, using matrix formulation is introduced. Using this approach, an efficient optical filter bank with low complexity and good frequency response is designed. It is interesting to note that this design is a mathematical model of the quincunx filter bank. The characteristics of this optical filter bank are compared with that of other commonly used short kernel filter banks, for video compression applications. The approach is based on multi-resolution representation of data, which is generated by the filter bank proposed in this work. The use of multi-resolution data structure in conjunction with other components of the system allows a simple and efficient implementation. Simulations on typical image sequences show that it is possible to perform generic coding with reduced complexity and good efficiency.
KEYWORDS: Signal processing, Optical signal processing, Transform theory, Computer architecture, Control systems design, Signals intelligence, Digital signal processing, Filtering (signal processing), Radar, Optical components
This paper delves into the fundamentals of hardware-software partitioning between an intelligent, generalized signal processor and reconfigurable hardware. A novel architecture is proposed for a RADAR control system and a design specification is illustrated with simulation results.
When Light propagates through optical elements (such as lenses), it undergoes a transform. The input and the output data take the form of light and optical elements that perform different mathematical operations on light represent the linear transform. The transform is performed not on the discrete elements of the data but on the whole vector at once, and most significantly, at the speed of light. The great advantages offered by optical processing are that it offers enormous parallelism, operating on all data points simultaneously, very low latency, a high transform rate and low power dissipation. The outcome is enormously increased speed and a reduction in the amount of associated cooling required. The Optical Signal Processor (OSP) increases the speed of processing transforms by many orders of magnitude. The Signal Processor is also reconfigurable and can be dynamically tailored to the required transform type. One advantage of an optical processor is that it allows software designers to work at a much higher level of abstraction. This is because the device executes transforms instead of the ordinary MACs in the case of DSPs. Instead of handling algorithms at individual data points, algorithms for handling the entire vector could be processed, shortening the computational complexity and speeding the time-to-market for new products. An optical filter can be represented as a generic function, the most fundamental of the optical processor. The impulse response of this filter is defined with respect to frequency of light. Any transform on light can be represented as a combination of linear transforms. This is fundamentally the law of optical signal processing. The most important application of an OSP in Optical Networking is Pattern recognition, and this can easily be done by the usual cross-correlation technique that is common in digital signal processing. The OSP can be programmed to autocorrelate against specific temporal reference waveforms, viz. Data. The decoding is done without electronic processing. And of
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.