Three‐dimensional quota matching‐based latency‐sensitive task offloading for multi‐mode green IoT in smart buildings

Abstract

The green internet of things with heterogeneous communication technologies can provide data transmission and computing services for low-carbon operation of smart buildings. However, latency-sensitive task offloading in smart buildings for multi-mode green internet of things still faces several challenges such as coupling between multi-mode channel and multiple gateway selection, diversified quality of service requirement guarantee, and contradiction of long-term performance guarantee and short-term optimisation objectives. To address these challenges, a three-dimensional quota matching-based latency-sensitive task offloading algorithm is proposed to minimise the weighted difference between energy consumption and throughput under the long-term queuing delay constraints. Specifically, the minimisation problem is decoupled by Lyapunov optimisation. The three-dimensional quota matching among devices, gateways, and channels is employed to solve the conflicts between gateway selection and channel selection. Finally, the three-dimensional quota matching is converted to a two-side quota matching to further reduce complexity and solved iteratively. Numerical results demonstrate that compared with H3CG and MMCS, the proposed algorithm improves the weighted difference between energy consumption and throughput by 21.85% and 27.91%, respectively, and reduces the sensor-side average queuing delay by 30.82% and 16.83%, and gateway-side average queuing delay by 16.57% and 26.71%, respectively.