For a blockwise coding of convolutional codes, we need to generate convolutional codes in a block unit. Two methods are well known for this application. The one is LCC(Linear Convolutional Code) which is based on linear convolution for a block of information bits and generator polynomials of convolutional code, i.e., zero-tail convolutional code. The other is CCC(Circular Convolutional Code) which is based on circular convolution for a block of information bits and generator polynomials. The LCC has essentially a rate loss due to the additional tail bits, but the CCC does not have it. The rate loss causes to expand transmission bandwidth or increase transmission power. The circular convolutional code(CCC) has advantage of no rate loss, but the maximum likelihood decoding requires much computational complexity.
In this thesis, we propose new decoding algorithms for the CCC of which computational complexity and performance approach to those for the zero-tail convolutional codes(ZTCC). The first algorithm is based on the application of the survivor path correction technique to the Viterbi algorithm. The decoding performance of this algorithm is similar to that of CVA(Circular Viterbi Algorithm) algorithm, but its decoding complexity is nearly half to that of CVA The second algorithm is based on the estimation of the tail symbols from the received symbol streams. The decoding performance of this algorithm is similar to that of the first algorithm, but the complexity is very much reduced to the range for the ZTCC. The third algorithm is a modification of the second algorithm. The decoding performance of this algorithm approaches to that of the ZTCC with a small increase in complexity.
The property of tail bit saving of CCC is applied in a CDMA mobile communications system design, effectively reducing decoder complexity. It shows that the system is further simplified by using a new Multilevel Logic Operation(MLO). MLO is an extention of conventional binary logic operation.