Traffic Adaptation and Energy Efficiency for Small Cell Networks with Dynamic TDD

The traffic in current wireless networks exhibits large variations in uplink (UL) and downlink (DL), which brings huge challenges to network operators in efficiently allocating radio resources. Dynamic time-division duplex (TDD) is considered as a promising scheme that is able to adjust the resource allocation to the instantaneous UL and DL traffic conditions, also known as traffic adaptation. In this work, we study how traffic adaptation and energy harvesting can improve the energy efficiency (EE) in multi-antenna small cell networks operating dynamic TDD. Given the queue length distribution of small cell access points (SAPs) and mobile users (MUs), we derive the optimal UL/DL configuration to minimize the service time of a typical small cell, and show that the UL/DL configuration that minimizes the service time also results in an optimal network EE, but does not necessarily achieve the optimal EE for SAP or MU individually. To further enhance the network EE, we provide SAPs with energy harvesting capabilities, and model the status of harvested energy at each SAP using a Markov chain. We derive the availability of the rechargeable battery under several battery utilization strategies, and observe that energy harvesting can significantly improve the network EE in the low traffic load regime. In summary, the proposed analytical framework allows us to elucidate the relationship between traffic adaptation and network EE in future dense networks with dynamic TDD. With this work, we quantify the potential benefits of traffic adaptation and energy harvesting in terms of service time and EE