Arithmetic with binary-encoded balanced ternary numbers

Ternary number representation and arithmetic, based on the radix-3 digit set {-1, 0, ;1}, has been studied at various times in the history of digital computing. Some such studies concluded that we should abandon ternary in favor of binary computation. Others, demonstrated promise and potential advantages, but, for various reasons, including inertia, did not lead to widespread use. By proposing an efficient binary encoding for balanced ternary numbers, along with the corresponding arithmetic circuits, we argue that a reexamination of the decision against using ternary arithmetic might be in order.