Your new problems on coding

Question 18. Is it possible to achieve in practice those actual MTD advantages with respect to other methods on efficiency and throughput at some decimal powers (100-1000 times!), which ones formally follow from estimations for a hardware decoder project? Yes, it is. The unconditional regularity of our approach to estimations of hardware decoding efficiency is doubtless. It consists simply that as contrasted to all other methods our MTD contains as though only of shift registers, which ones are the most fast elements of any LSI circuit. Besides it is a lot of such registers working in parallel mode in MTD: from 4 up to 20 and more. And since customary speeds of shift registers of PLIS are tens Mbit/s, MTD decoders really are devices, which ones easily process flows at speeds in many hundreds Mbit/s. Certainly, the problem is what do other devises of the MTD decoder do? You see it is necessary to make the solutions on which symbols should be changed. Process of decoding-is always a very difficult computing problem. Why then: "MTD is the only shift registers!"? Yes, because really active schemes of MTD are only threshold switches (TS). They are-extremely simple devices for integers addition and comparison with threshold values. Their quantity is very small as contrasted to tens and hundreds kbits of register memory. And it means, that in MTD ideal full multisequencing of operations is probable: creation of large number separately working TS. And the last major moment: all this set of TSs can easily be made by the schemes, which one will be looked outwardly as processors with the instantaneous answer: have shifted the data in the register-and here have received reacting of TSs, answer, whether it is necessary to change a controlled symbol. It is possible and even it is simple enough, as all TSs simply summarize some small integers. And for other algorithms of any such capabilities for superfast decoders organization in general are absent. That's all about throughput. And if to return to a problem of efficiency, it is possible simply to remind, that MTD-iterative algorithm, which one at all changes of controlled