I am trying to compute the value of bit error rate by using erfc ,but the following error occurs,how I can solve this?
effSNR = cfh.effectiveSNR(txSymbols, rxSymbols, one, reduce_axis)
BER=(2.0/np.log2(param.M))*(1.0-(1/math.sqrt(param.M)))* erfc(math.sqrt((2.0*np.log2(param.M)*effSNR)/(2.0*((param.M)-1.0))))
Type Error: must be real number, not Tensor