Cheetah - SKA - PSS - Prototype Time Domain Search Pipeline
FLDO Configuration Guide

The folding portion behavior can be changed by means of some input parameters. In the following discussion the configurable parameter names are in italic . The complete list of SKA DDD input parameters for PSS can be found in PSS_Parameters.md

FLDO receives multi-channel time-series data (a stream of filterbank data) and produces dedispersed, binned (folded) profiles for each candidate in input, at the given dispersion measure, period and period derivative.

Data are organized as a succession of measures, each comprising freq_channels samples each referring to a different frequency. Samples are integer numbers bit_per_sample bits long. The measures are taken every time_resolution microseconds. The number of measures is integration_time / time_resolution .

The folding algorithm consists in a synchronous summation of input data in order to improve the Signal/Noise. Input data is processed in blocks of subints sub-integrations. Each sub-integration is then divided in sub_bands bands of frequencies and corrected for dispersion. Data are summed up coherently, following the pulsar period of the current candidate in an array of maximum length of phases phases.

subint_structure.png
Schema of data of a scan