Recently, with the advent of Internet of things, mobile devices are expected to carry on more and more computation-intensive applications, and mobile cloud computing also comes in to help mitigate the burden of mobile devices. However, due to the rapidly growing density of mobile devices, the current single cloud environment will still encounter resource depletion, which will lead to application offloading failure and user dissatisfaction. In order to enhance the scalability and reliability of application offloading service in mobile cloud computing, we study the heterogeneous mobile cloud, which integrates mobile cloudlet with infrastructure-based cloudlet. In such a scenario, when the infrastructure-based cloudlet becomes unavailable due to insufficient resource, mobile cloudlet can help to provide computation offloading service. We develop a centralized task scheduling algorithm to select reliable worker nodes in mobile cloudlet for the offloaded tasks. Simulation results show that heterogeneity of cloud resource can enhance scalability of offloading service efficiently. Also, the proposed scheduling algorithm can extend the network lifetime and keep balance of energy consumption in the mobile cloudlet.