遠隔制御に適した圧縮センシング手法の検討

Abstract

無線通信路のように帯域制限された通信路を介した遠隔制御系では，制御信号をいかに効果的に表現し送信するかが重要な問題となる．この問題に対して，圧縮センシングの枠組みでの新しい手法が提案され， 1- 2 最適化による送信信号のスパース表現が有効であることが示された．しかし，圧縮センシングにおいては， 1- 2 最適化のみならず，様々な手法が存在する．本研究では，数値シミュレーションを用いて，遠隔制御系に適した手法を比較検討し，その優劣を考察する.

In remote control systems, efficient representation of control signals is one of the crucial issues because of bandwidth-limitedness of the communication channel, such as a wireless communication link, between the controller and the controlled object. Recently, a new method based on compressed sensing has been proposed, in which control signals are sparsely representation based on l^1-l^2 optimization. There exist however so many methods other than l^1-l^2 optimization for compressed sensing. In this study, we perform a comparative study of sparsity-promoting methods in compressed sensing, and reveal their advantages and disadvantages by simulation in view of remote control over rate-limited networks.