This paper studies the IoT relaying networks, where one decode-and-forward (DF) relay equipped with a buffer helps determine whether to receive the data from the source or send the data to the destination. Moderate shadowing environments are considered and the source can also communicate with the destination through the direct link. Although the opportunistic selection (OS) technique can select one better branch among the direct and buffer-aided relay branches, it requires to estimate the channel parameters of two branches, and the branch switching may occur frequently. To deal with these disadvantages, we employ the distributed switch-and-stay combining (DSSC) technique for the buffer-aided relay network, in which the same branch goes on being employed until the current branch cannot support the data transmission no longer. In this protocol, the branch switching depends on the buffer state and the state transition in the buffer is also affected by the branch switching. For this communication scenario where the buffer meets the DSSC, we study the system performance by deriving the analytical outage probability and the asymptotic expression with high regime of transmit power. From the asymptotic expression, we find that DSSC can effectively exploit the system spatio-temporal resources and it obtains the same performance of the OS with ease of implementation. Moreover, the system performance improves with the buffer size. We finally provide some simulation and numerical results to demonstrate the merits of the DSSC-based buffer-aided IoT relaying networks.