Cant we use 2s complement or sign bit in biasing the exponent in 32 bit binary floating point representation?
