This chapter deals with the issues of numerical representation of data in a digital signal processing. As all digital devices operate using bits, numerical values must be converted into a value representable using bits. Fixed point representation is most common, where numerical values are mapped into a field of bits covering a range of integer values, or alternately, a set of fractional values from â1 to 1. A more complex alternative is floating point that is similar to scientific notation. In this case, a numerical value is represented using a mantissa and exponent, along with a sign bit. This has the advantage of being able to cover a much wider range of numbers, both very large and very small.