In their CMOS VLSI Design, Weste and Harris seem to use a naming convention for logic gates which I cannot quite seem to define in my head. Ill give the two examples they use and hopefully someone knows about it or can point me to standard source:
(1) \$Y = \overline{AB + CD}\$. They call this an AND-OR-Invert-22, or AOI22 gate because "it performs the NOR of a pair of 2-input ANDs"
(2) \$Y = \overline{(A+B + C)D}\$. They call this an OAI31 gate.
What is the crux of this naming convention with respect to the numbers used in particular? I had initially thought that the numbers were (A) the number of inputs to each "sub-gate". For example, in case (1) with the AOI22 gate we have two inputs to the AND gates at the first level, and 2 inputs to the OR at the second level. However, this fails for case (2). Next, I thought (B) that the first number was for the number of inputs to the gate at the first level (2 in AOI22 for the 2-input ANDs, 3 for the the three-input OR) and the second number is for the number of gates at the first level (2 ANDs and 1 OR). This seems to work for both cases. However, it seems to fail to uniquely name a gate (consider e.g. \$Y = \overline{(A+B + C)DE}\$ which would also seem to get the OAI31 moniker under this rubric) and so I wasn't sure what the deal was.