Faster algorithm for designing optimal prefix-free codes with unequal letter costs

Summary form only given. A new approach to solving the optimal prefix-free coding problem for the case of binary encoding alphabet has been proposed. The new approach, however, is limited to the binary case because the techniques employed seemed not to extend to the r-ary case. This paper proves that, on the contrary, the approach can be generalized to the r-ary case as well, thus leading to an O(nc) time solution for general r