<p>Integrating IoT resources including edge devices and services in the IoT computing platforms under wireless network is a mainstream task. However, the large number of heterogeneous edge devices providing diverse mobile services and applications brings challenges for the service selection and governance for IoT computing platforms. Therefore, we consider to introduce the recommendation technique to facilitate mobile service selection and governance. The existing works are generally limited by high-order feature interaction modeling and complex feature integration. To alleviate this issue, this paper proposes a Mamba-Transformer based framework called MamTRec for IoT service recommendation. MamTRec effectively fuses the features of multiple types by leveraging the self-attention mechanism and position-wise feed-forward network. Directly fusing multiple features might lead to the feature redundancy, MamTRec incorporates the Mamba layer to make a data-dependent feature selection with the structured state space models (SSMs) for the next feature fusion. MamTRec ingeniously integrates the distinct advantages of both the mamba layer and Transformer architecture. Comprehensive experiments on the real-world dataset demonstrate the effectiveness of MamTRec in service recommendation task.</p>