> 2) I am using my custom board in very noisy environment in which standalone magnetometer is not capable to calculate proper north heading after applying hard iron and soft iron. so is it possible to calculate proper north heading using accelerometer and magnetometer fusion using e-Compass library or any other library in noisy environment.
"Proper north" is defined as a magnetic pole, so an accelerometer will be of no use in determining it.
The accelerometer is only useful in determining movement (actually, the differential).
> Initially I tried to get north heading through magnetometer but due to external noise some time I am getting proper north heading but some time getting wrong orientation.
This can be internal noise due to inadequate layout or other nearby components, or due to the external environment.
If the latter is the case, your magnetometer signal is swamped by noise, and your project is very likely to fail.
But since you mentioned a custom board, I think you can try to improve the hardware design.
