I. Introduction
The future wireless network, such as Smart City and Industry 4.0 enable numerous innovative services by massively deploying distributed nodes to observe diverse phenomena, i.e., temperature, humidity, and gases [1]–[3]. There is no doubt that huge numbers of observations for phenomena are generated at distributed nodes and require rapid transmissions to support the above services. It is a great challenge for orthogonal multi-access (OMA) schemes [4] to transmit such massive observations in time due to the low spectrum efficiency [5]. To tackle this challenge, the compressed sensing (CS) technique is widely regarded as a promising method by exploiting the sparsity of observations, which realizes the CS measurement of observations, rather than collecting enormous redundant observations [6]–[8].