scarie:study_on_the_scarie_numbers

This is a very common question in the SCARIe meeting and discussion…what are the number.

As this is not so easy to describe place here all the informations to this issue you may find relevant.

If possible try to use: parametric equations to describe the relationship between the numbers. contraint equations to describe what is the result of an “hardware” limitation interval of acceptable value for the parameter [x,y]

A SCARIe experiment is defined by: Number of Radio-Telescope: RT in [2, 16] Number of channels (aka frequency-band or subband) per telescope: NC in [1, 64] Number of bit per sample: BPS in [1, 2] Number of sample_rate : SR in [4000000, 32000000] (with only power of 2 stepping)

Data rate:

data_rate == NC*BPS*SR.

Constraints of the current hardware correlator:

data_rate == recording_speed with recording_speed is < 1Gbps.

Question: how those number are related to “polarization” etc…

The amount of data arriving from one radio-telescope is:

data_rate.

As channels can be correlated independently. The data_rate per telescope per channel is:

BPS*SR

During the correlation, the data is integrated over a certain time, called the integrationTime. The integrationTime is typical in the range of 0.1 to 5 seconds. Hence, the size of an integration bloc is bounded to be:

BPS*SR*integrationTime

Input nodes: The amount of data received on an input node is:

NC*BPS*SR

The amount of data emitted by an input node is also:

NC*BPS*SR

Correlation nodes we currently don't do baseline distribution). The amount of data that have to be received for a single integration step is:

BPS*SR*integration_step*RT

Constraints related to das-3 hardware:

2*NC*BPS*SR < 10Gbps (the pressure on the input node)

Other constraints can be calculated but are related to how the application is deployed.

work in progres…

scarie/study_on_the_scarie_numbers.txt · Last modified: 2007/08/20 11:08 (external edit)