Speaker: Yun Li
Title: Delay and Energy Aware Duty Cycle Control for Machine to Machine Networks
A hybrid network architecture has been proposed for machine-to-machine (M2M) communications, where M2M gateways connect the capillary networks and cellular networks. In capillary M2M network, a large number of non-cellular M2M devices using short-range, low-cost, low-energy consumption radio inter- faces, such as IEEE 802.15.4 to connect to the base station via M2M gateways. As the massive access of M2M devices in the capillary networks to LTE net- works is ensured via M2M gateways, novel access control algorithms for the M2M gateways and coordinators of capillary networks are necessary to ensure the performance of network throughput, delay and energy efficiency.
Our work focus on developing novel delay and energy aware duty cycle con- trol for the gateway and the capillary networks coordinator. The low duty cycle in IEEE 802.15.4 improves the energy efficiency. However, it increases the end- to-end delay as devices may switch into inactive periods to save energy. In addition, the uniformed duty cycle control in current standard may not meet various requirements of different applications. To tackle these challenges, we formulate the duty cycle optimisation problem with joint energy efficiency and end-to-end delay consideration as an inventory control problem. Then, the opti- mal solution to the problem is derived by applying dynamic programming (DP). Furthermore, the rollout algorithm based low complexity duty cycle control is proposed to reduce the computational complexity of implementing the control on computation limited sensor devices.
Date: 12th Nov, 2014.
Time: 14.00-13.00 hrs