研究実績の概要 |
We are progressing according to the plan in 2020, and achieved the following results. The first achievement is construction of signature code in algebraic. We propose a coding scheme for a noisy MAAC. In our scheme, given a k-ary code A, with code length n and a (2k-1)-ary code B, with code length n, by a Hadamard matrix of order q we obtain a k-ary code C with code length qn. The main idea behind our coding scheme is to introduce the (2k-1)-ary code B, as well as A for constructing the k-ary code C, thus providing code C with a higher sum rate than A. This is an improvement of the sum rate compared to conventional coding schemes for a noisy MAAC, where the sum rates of C and A are always the same. The second achievement is considering user identification and channel estimation of binary signature code by DNN-based decoder on multiple-access channel. In the previous works, the signature code was used over a noisy multiple-access adder channel and only the status of uses are decoded by the signature decoder. By considering the communication model as a compressed sensing process, it is possible to estimate the channel coefficients while identifying users. To improve the efficiency of the decoding process, we proposed a iterative deep-neural-network-based decoder. Our simulation results show that for the binary signature code, our proposed DNN-based decoder requires less computing time to achieve higher active user detection accuracy and channel estimation accuracy than the classical signal recovery algorithm used in compressed sensing.
|