Real time signal processing

Bickel, H.J., and Rothschild, R.S., Real-Time Signal Processing in the Frequency Domain, Federal Scientific Monograph 3, March 1973. 

Floating Point. Integrated floating point units first arrived as separate coprocessors under the direct control of the microprocessor. However, these processors performed arithmetic with numerous sequential operations, resulting in performance too slow for real-time signal processing. 

Most of the current design models are not suitable for the real-time signal processing domain and do not deal with the above requirements. At the start of the Ascis project, there was a clear need to fill this gap. This has led to the Ascis data flow-graph (DFG) model described in chapter 2. It differs... 

A second important architecture style for real-time signal processing systems is the multiplexed processor style, characterized by a set of application-specific, time-multiplexed data-paths steered by a hierarchically organized controller. This style is tuned to irregular applications, requiring a medium to high sample rate (10 kHz - 10 MHz). Many applications at these rates require a combination of computation-intensive arithmetic and complex decision-making operations. For these applications, the array style (as described in section 5) is unsuitable. 

S. Hansen and J. G. Nash. Modified Faddeev algorithm for matrix manipulation. In Real-time Signal Processing VII, pages 39-46. Society of Photo-Optical Instrumentation Engineers, 1984. 

Many design methods have been proposed to synthesize RAAs. The majority of these methods (see chapter 1 for an overview) are based on an affine transformation (first described by Quinton [15] and Moldovan [12]) to map the index space of the application description to time and processor space. The use of such a transformation method simplifies the design task considerably and requires only a few parameters to characterize a design completely. Unfortunately, most of these methods start from a relatively low level specification, using sets of UREs (uniform recurrence equations) [3] or CUREs (conditional uniform recurrence equations) [17] to describe an application. Moreover, especially in the case of real-time signal processing applications, the resulting architecture is usually unnecessarily fast or too slow. 

The synthesis objective that will be used throughout this chapter is the minimization of the total area consumed by all resources, i.e., memories, operators, and interconnect, under a user-specified throughput constraint (e.g., the sample rate). This is compatible with the requirements for real-time signal processing systems that are targeted to customized architectures but not fully power-dominated. Extensions to other optimization objectives, based on cost factors like low power, are feasible but will not be discussed here. 

M. van Swaaij, J. Rosseel, F. Catthoor, and H. De Man. High-level synthesis of ASIC regular arrays for real-time signal processing systems. In Proc. Int. Workshop on Algorithms and Parallel VLSI Architectures, Pont-a-Mousson, Prance, June 1990. 

The system uses a library of FUs similar to the application-specific units (ASUs) used in Cathedral-3, which is described in chapter 7 [11]. In Cathedral-3, the ASUs are extracted automatically from the behavioral description. This is important for high-throughput real-time signal processing applications that require complex application-specific data-paths, as targeted by Cathedral-3. Within Amical, the FUs are provided by the user. Even if this approach is less automatic, it provides more flexibility when using existing hardware. 

In PARR-1 a new application of real-time signal processing has been used. Statistical analysis of signals from reactor instrumentation channels is done in real-time for evaluation of instrumentation performance. The computer calculates mean value, standard deviation errors, and probability distribution function of the signals and compares these errors with reference errors of nuclear detection phenomena. In case of a malfunction in any part of instrumentation, the signal error exceeds the reference error and the computer generates an alarm. In this way a faulty instrument channel is identified. 

Nijnieijer, M.C, Boer, M.A. Correction of lens-distortion for real-time image processing systems. VLSI Signal Processing VI... 

With regard to instrumentation design, WT was applied to process real-time signals from the mass spectrometer. Shew [51] invented a new procedure for determining the relative ion abundances in ion cyclotron resonance mass spectrometry, by utilizing WT to isolate the intensity of a particular ion frequency as a function of position or time within the transient ion cyclotron resonance signal. In 1995, this new method was patented in the U.S. Shew explained that the WT intensity corresponding to the frequency of each ion species as a function of time can be fitted by an exponential decay curve. By... 

State-of-the-art real-time signal and data processing applications typically involve a rapidly increasing arithmetic complexity. In many cases, this is combined with a need for fiexible and powerful decision-making. Furthermore, they... 

The properties of real-time signal and data processing applications have to be exploited to arrive at fully efficient application-specific architectures. This requires extensions to the basic linear space-time mapping methods [44] as proposed in chapters 3 and 6. In addition, an alternative affine space-time transformation method based on the existence of independent subsets in the index space is introduced in chapter 5. This method has led to efficient arrays with high hardware utilization [24]. 

In latency-limited applications, the important criterion is the execution time T. This is in contrast to the real-time signal and data processing applications, where the period P is more important. Solutions for the latter category are the focus of chapters 5 and 6. In this chapter, the latency-limited case will be addressed. Even though a large part of the underlying theory can be shared, important distinctions will be pointed out. 

Low-pass filtering can also be performed prior to computer sampling by conventional analog filtering. This allows a designer to be less concerned about the speed of computer processing needed for real-time signals. 

Real-Time Clocks. Process control systems must respond to events in a timely manner and should have the capabihty of real-time control. Some DCSs are connected to atomic clock signals to maintain accuracy. 

Processing Speed DSP processors have a processing speed that is superior to others that are fi equently used for real-time signal analysis. 

Digital Signal Processor board fPSP) it is hosted into the PC and processes in real time the binary sequences stored into the acquisition board FIFO memories. The board processes arrival times and extracts the correlated AT generated by AE events. The PC picks up the data stored into the DSP memories and calculates the position of the AE sources. 

---